Base_JSON/prefixC/json/constraint/2022.constraint-1.1.json

{
    "paper_id": "2022",
    "header": {
        "generated_with": "S2ORC 1.0.0",
        "date_generated": "2023-01-19T13:12:33.875946Z"
    },
    "title": "Findings of the CONSTRAINT 2022 Shared Task on Detecting the Hero, the Villain, and the Victim in Memes",
    "authors": [
        {
            "first": "Shivam",
            "middle": [],
            "last": "Sharma",
            "suffix": "",
            "affiliation": {},
            "email": "[email protected]"
        },
        {
            "first": "Tharun",
            "middle": [],
            "last": "Suresh",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Atharva",
            "middle": [],
            "last": "Kulkarni",
            "suffix": "",
            "affiliation": {},
            "email": "[email protected]"
        },
        {
            "first": "Himanshi",
            "middle": [],
            "last": "Mathur",
            "suffix": "",
            "affiliation": {},
            "email": "[email protected]"
        },
        {
            "first": "Preslav",
            "middle": [],
            "last": "Nakov",
            "suffix": "",
            "affiliation": {
                "laboratory": "",
                "institution": "HBKU",
                "location": {
                    "settlement": "Doha",
                    "country": "Qatar"
                }
            },
            "email": "[email protected]"
        },
        {
            "first": "Md",
            "middle": [
                "Shad"
            ],
            "last": "Akhtar",
            "suffix": "",
            "affiliation": {},
            "email": "[email protected]"
        },
        {
            "first": "Tanmoy",
            "middle": [],
            "last": "Chakraborty",
            "suffix": "",
            "affiliation": {},
            "email": "[email protected]"
        }
    ],
    "year": "",
    "venue": null,
    "identifiers": {},
    "abstract": "We present the findings of the shared task at the CONSTRAINT 2022 workshop on \"Hero, Villain, and Victim: Dissecting Harmful Memes for Semantic Role Labeling of Entities.\" The task aims to delve deeper into meme comprehension by deciphering the connotations behind the entities present in a meme. In more nuanced terms, the shared task focuses on determining the victimizing, glorifying, and vilifying intentions embedded in meme entities to explicate their connotations. To this end, we curate HVVMemes, a novel meme dataset of about 7,000 memes spanning the domains of COVID-19 and US Politics, each containing entities and their associated roles: hero, villain, victim, or other. The shared task attracted 105 registered participants, but eventually only nine of them made official submissions. The most successful systems used ensembles combining textual and multimodal models, with the best system achieving an F1-score of 58.67.",
    "pdf_parse": {
        "paper_id": "2022",
        "_pdf_hash": "",
        "abstract": [
            {
                "text": "We present the findings of the shared task at the CONSTRAINT 2022 workshop on \"Hero, Villain, and Victim: Dissecting Harmful Memes for Semantic Role Labeling of Entities.\" The task aims to delve deeper into meme comprehension by deciphering the connotations behind the entities present in a meme. In more nuanced terms, the shared task focuses on determining the victimizing, glorifying, and vilifying intentions embedded in meme entities to explicate their connotations. To this end, we curate HVVMemes, a novel meme dataset of about 7,000 memes spanning the domains of COVID-19 and US Politics, each containing entities and their associated roles: hero, villain, victim, or other. The shared task attracted 105 registered participants, but eventually only nine of them made official submissions. The most successful systems used ensembles combining textual and multimodal models, with the best system achieving an F1-score of 58.67.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Abstract",
                "sec_num": null
            }
        ],
        "body_text": [
            {
                "text": "The unwarranted spread of misinformation (Wu et al., 2019; Hardalov et al., 2022) , propaganda (Da San Martino et al., 2020a,b) , fake news (Lazer et al., 2018; Vosoughi et al., 2018) , COVID-19 infodemic Nakov et al., 2022) , hate speech (MacAvaney et al., 2019; Zampieri et al., 2019a) , and other harmful content (Nakov et al., 2021) has plagued social media. Lately, memes have emerged as a powerful multimodal means to disseminate malicious content due to their ability to circumvent censorship norms (Mina, 2014) and to their fast-spreading nature. With an aptly crafted combination of images and text, a seemingly na\u00efve meme can easily become a source of harmful information diffusion. As a result, exploring the noxious side of memes has become a pressing research topic; see also recent surveys on harmful memes (Sharma et al., 2022b ) and on multimodal disinformation detection (Alam et al., 2021a) .",
                "cite_spans": [
                    {
                        "start": 41,
                        "end": 58,
                        "text": "(Wu et al., 2019;",
                        "ref_id": "BIBREF61"
                    },
                    {
                        "start": 59,
                        "end": 81,
                        "text": "Hardalov et al., 2022)",
                        "ref_id": null
                    },
                    {
                        "start": 95,
                        "end": 127,
                        "text": "(Da San Martino et al., 2020a,b)",
                        "ref_id": null
                    },
                    {
                        "start": 140,
                        "end": 160,
                        "text": "(Lazer et al., 2018;",
                        "ref_id": "BIBREF24"
                    },
                    {
                        "start": 161,
                        "end": 183,
                        "text": "Vosoughi et al., 2018)",
                        "ref_id": "BIBREF60"
                    },
                    {
                        "start": 205,
                        "end": 224,
                        "text": "Nakov et al., 2022)",
                        "ref_id": "BIBREF40"
                    },
                    {
                        "start": 239,
                        "end": 263,
                        "text": "(MacAvaney et al., 2019;",
                        "ref_id": "BIBREF29"
                    },
                    {
                        "start": 264,
                        "end": 287,
                        "text": "Zampieri et al., 2019a)",
                        "ref_id": "BIBREF67"
                    },
                    {
                        "start": 316,
                        "end": 336,
                        "text": "(Nakov et al., 2021)",
                        "ref_id": "BIBREF50"
                    },
                    {
                        "start": 506,
                        "end": 518,
                        "text": "(Mina, 2014)",
                        "ref_id": "BIBREF31"
                    },
                    {
                        "start": 821,
                        "end": 842,
                        "text": "(Sharma et al., 2022b",
                        "ref_id": "BIBREF54"
                    },
                    {
                        "start": 888,
                        "end": 908,
                        "text": "(Alam et al., 2021a)",
                        "ref_id": "BIBREF11"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "While meme analysis has been studied in a variety of contexts, such as hate speech (Zhou et al., 2021; Kiela et al., 2020) harmfulness (Pramanick et al., 2021a,b) , emotions (Sharma et al., 2020) , misinformation (Zidani and Moran, 2021) , sarcasm (Kumar and Garg, 2019) , offensiveness (Suryawanshi et al., 2020) , and propaganda (Dimitrov et al., 2021a,b) , limited forays have been made on comprehending the role of the entities that make up a meme. This is our main focus here: on identifying the hero, the villain, and the victim entities present in a meme. Given a meme and a list of the entities it involves, the task is to identify which entity plays what role. Such categorization of the entities in the meme can help understand the entity-specific connotation and their nature, attitudes, decisions, and demeanour. For instance, when the meme creators intend to spread misinformation and hatred towards minority communities or to defame certain individuals, politicians, or organizations, they would depict the target entities as villains. Similarly, when the intent is to shed light on the deplorable state of certain entities or to glorify them, these entities would be portrayed as victims or as heroes, respectively. Fig. 1 depicts apt examples for hero, villain, and victim categorization of the entities in a meme. The meme in Fig. 1a draws a comparison between Abraham Lincoln, John F. Kennedy, Barack Obama, and Donald Trump, where the former three are portrayed as heroes, while Donald Trump is shown in negative light, as a villain. Similarly, Fig. 1b mocks Jill Stein and the Green Party as villains for allegedly getting bribed by the rich. Fig. 1c on the other hand, frames the Republican Party as a villain, for their inconsiderate views on the poor, the minorities, and women, thus making them the victims. In conclusion, through depictions of heroism, villainy, and victimization, memes act as an appealing means to propagate certain views about the targeted entities. While some previous meme studies have sought to identify harmfulness and the entities (Sharma et al., 2022a) or the categories that are being targeted, e.g., a person, a group, an organization, or society (Pramanick et al., 2021a,b) , none of them has scrutinized the entity's connotation. Our shared task aims to bridge this gap. We release HVVMemes, a meme dataset with about 7,000 memes on COVID-19 and US Politics, where each meme is annotated with a list of entities, each labeled with its role: hero, villain, victim, or other. The shared task attracted 105 teams, and nine of them made official submissions. Most teams finetuned pre-trained language and multimodal models or used ensembles, with the best system achieving an F1-score of 58.67. We discuss the submissions and their approaches in more detail in Section 5.",
                "cite_spans": [
                    {
                        "start": 83,
                        "end": 102,
                        "text": "(Zhou et al., 2021;",
                        "ref_id": "BIBREF72"
                    },
                    {
                        "start": 103,
                        "end": 122,
                        "text": "Kiela et al., 2020)",
                        "ref_id": null
                    },
                    {
                        "start": 135,
                        "end": 162,
                        "text": "(Pramanick et al., 2021a,b)",
                        "ref_id": null
                    },
                    {
                        "start": 174,
                        "end": 195,
                        "text": "(Sharma et al., 2020)",
                        "ref_id": null
                    },
                    {
                        "start": 213,
                        "end": 237,
                        "text": "(Zidani and Moran, 2021)",
                        "ref_id": "BIBREF74"
                    },
                    {
                        "start": 248,
                        "end": 270,
                        "text": "(Kumar and Garg, 2019)",
                        "ref_id": "BIBREF22"
                    },
                    {
                        "start": 287,
                        "end": 313,
                        "text": "(Suryawanshi et al., 2020)",
                        "ref_id": "BIBREF58"
                    },
                    {
                        "start": 331,
                        "end": 357,
                        "text": "(Dimitrov et al., 2021a,b)",
                        "ref_id": null
                    },
                    {
                        "start": 2081,
                        "end": 2103,
                        "text": "(Sharma et al., 2022a)",
                        "ref_id": "BIBREF53"
                    },
                    {
                        "start": 2200,
                        "end": 2227,
                        "text": "(Pramanick et al., 2021a,b)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [
                    {
                        "start": 1231,
                        "end": 1237,
                        "text": "Fig. 1",
                        "ref_id": "FIGREF0"
                    },
                    {
                        "start": 1343,
                        "end": 1350,
                        "text": "Fig. 1a",
                        "ref_id": "FIGREF0"
                    },
                    {
                        "start": 1564,
                        "end": 1571,
                        "text": "Fig. 1b",
                        "ref_id": "FIGREF0"
                    },
                    {
                        "start": 1663,
                        "end": 1670,
                        "text": "Fig. 1c",
                        "ref_id": "FIGREF0"
                    }
                ],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "Despite the growing body of research on meme analysis, understanding the connotation underlying the individual entities in the meme remains a challenging endeavour. Their camouflaged semantics, satirical outlook, and cryptic nature make their analysis a daunting task (Sabat et al., 2019) . Moreover, categorizing the entities as heroes, villains, or victims requires real-world and commonsense knowledge, which often are not present in popular pre-trained language models. Thus, it should not be surprising that, as the shared task's results show, off-the-shelf multimodal models, as well as various ensembles thereof, struggle with this task (Kiela et al., 2020) . This highlights that the current stateof-the-art visual-linguistic models are unable to grasp the veiled information present in the memes. Thus, we hope that the dataset and task will foster further research in this interesting direction.",
                "cite_spans": [
                    {
                        "start": 268,
                        "end": 288,
                        "text": "(Sabat et al., 2019)",
                        "ref_id": "BIBREF47"
                    },
                    {
                        "start": 644,
                        "end": 664,
                        "text": "(Kiela et al., 2020)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "More details about the shared task is available at http://constraint-lcs2.github.io/",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "Studies on Online Targeting. Previous work studied affective content in the context of harmful discourse in social media (Zainuddin et al., 2017 (Zainuddin et al., , 2018 Gautam et al., 2020; Ousidhoum et al., 2019) . Sarcastic content was detected by leveraging data sparseness (Zainuddin et al., 2019) towards studying aspect-based sentiment analysis. Shvets et al. (2021) established enhancements in target detection by examining generic concept extraction for hate speech detection. Targeted protected categories were characterised by harmful online engagements whilst addressing societal bias along with explainability (Sap et al., 2020; Mathew et al., 2021) . For affective target characterisation, sequence modeling was explored in a hierarchical formulation of stacked BiGRUs (Ma et al., 2018) as well as in lowresource scenarios (Mitchell et al., 2013) . Most approaches did not consider the variability in target referencing and the associated affective spectrum (Shvets et al., 2021) . Finally, (Gomez-Zara et al., 2018) discussed hero/villain/victim analysis of news text; unlike their work, here we focus on multi-modality and memes.",
                "cite_spans": [
                    {
                        "start": 121,
                        "end": 144,
                        "text": "(Zainuddin et al., 2017",
                        "ref_id": "BIBREF64"
                    },
                    {
                        "start": 145,
                        "end": 170,
                        "text": "(Zainuddin et al., , 2018",
                        "ref_id": "BIBREF65"
                    },
                    {
                        "start": 171,
                        "end": 191,
                        "text": "Gautam et al., 2020;",
                        "ref_id": null
                    },
                    {
                        "start": 192,
                        "end": 215,
                        "text": "Ousidhoum et al., 2019)",
                        "ref_id": "BIBREF42"
                    },
                    {
                        "start": 279,
                        "end": 303,
                        "text": "(Zainuddin et al., 2019)",
                        "ref_id": "BIBREF66"
                    },
                    {
                        "start": 354,
                        "end": 374,
                        "text": "Shvets et al. (2021)",
                        "ref_id": "BIBREF55"
                    },
                    {
                        "start": 624,
                        "end": 642,
                        "text": "(Sap et al., 2020;",
                        "ref_id": "BIBREF48"
                    },
                    {
                        "start": 643,
                        "end": 663,
                        "text": "Mathew et al., 2021)",
                        "ref_id": "BIBREF30"
                    },
                    {
                        "start": 784,
                        "end": 801,
                        "text": "(Ma et al., 2018)",
                        "ref_id": "BIBREF28"
                    },
                    {
                        "start": 838,
                        "end": 861,
                        "text": "(Mitchell et al., 2013)",
                        "ref_id": "BIBREF32"
                    },
                    {
                        "start": 973,
                        "end": 994,
                        "text": "(Shvets et al., 2021)",
                        "ref_id": "BIBREF55"
                    },
                    {
                        "start": 1006,
                        "end": 1031,
                        "text": "(Gomez-Zara et al., 2018)",
                        "ref_id": "BIBREF16"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Work",
                "sec_num": "2"
            },
            {
                "text": "Studies on Detecting Harmful Memes. The constant transitioning of harmful memes from unfiltered and largely anonymous communities and platforms such as 4chan, Reddit, and Gab to more mainstream social media has made the entire social media ecosystem both sensitive and vulnerable to extremism (Zannettou et al., 2018) . Research on offense (Suryawanshi et al., 2020) , hate speech (Kiela et al., 2020; Gomez et al., 2020) , and online harm detection (Pramanick et al., 2021b) has found the availability of large datasets and the use of multi-modal frameworks crucial for these tasks.",
                "cite_spans": [
                    {
                        "start": 293,
                        "end": 317,
                        "text": "(Zannettou et al., 2018)",
                        "ref_id": "BIBREF70"
                    },
                    {
                        "start": 340,
                        "end": 366,
                        "text": "(Suryawanshi et al., 2020)",
                        "ref_id": "BIBREF58"
                    },
                    {
                        "start": 381,
                        "end": 401,
                        "text": "(Kiela et al., 2020;",
                        "ref_id": null
                    },
                    {
                        "start": 402,
                        "end": 421,
                        "text": "Gomez et al., 2020)",
                        "ref_id": "BIBREF15"
                    },
                    {
                        "start": 450,
                        "end": 475,
                        "text": "(Pramanick et al., 2021b)",
                        "ref_id": "BIBREF44"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Work",
                "sec_num": "2"
            },
            {
                "text": "Additional contextual cues involving commonsense knowledge (Shang et al., 2021) , semantic entities, cues about the protected categories (Pramanick et al., 2021b; Karkkainen and Joo, 2021) , along with other meta information, have also been explored for characterising various aspects of the online harm conveyed by memes. Most such tasks address affect detection at various levels of granularity, sometimes organised in a taxonomy. Still, none of these tasks has focused on explicitly modeling the complex narrative framework of the memetic discourse surrounding the specific entities referred to in the meme. With this in mind, here we attempt to alleviate a few associated challenges by exploring the feasibility of entity-specific visual-semantic role labelling for memes.",
                "cite_spans": [
                    {
                        "start": 59,
                        "end": 79,
                        "text": "(Shang et al., 2021)",
                        "ref_id": "BIBREF51"
                    },
                    {
                        "start": 137,
                        "end": 162,
                        "text": "(Pramanick et al., 2021b;",
                        "ref_id": "BIBREF44"
                    },
                    {
                        "start": 163,
                        "end": 188,
                        "text": "Karkkainen and Joo, 2021)",
                        "ref_id": "BIBREF19"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Work",
                "sec_num": "2"
            },
            {
                "text": "Other Related Shared Tasks. Several shared tasks have targeted the broad field of harmful social media content. Some tasks investigated the characterisation of offensive language, hate speech, profanity, and associated fine-grained attributes such as implicit and explicit implications in binary, multiclass, multi-label, and hierarchical settings (Strus et al., 2019; Zampieri et al., 2019b Zampieri et al., , 2020 . Their coverage has been fairly comprehensive in terms of the languages covered including Arabic, Danish, Greek, English, Turkish, and Dravidian Languages like Tamil, Malayalam, Kannada as well as German and English/Indo-Aryan code-mixing (Zampieri et al., 2019b; Zampieri et al., 2020; Chakravarthi et al., 2021; Modha et al., 2021) . They also address harmful content dissemination, targeting various protected categories such as religious affiliation, national origin, sex, etc. . Other efforts have targeted misinformation, propaganda, and persuasiveness detection (Aly et al., 2021; Shaar et al., 2021; Da San Martino et al., 2020a) , where the goal is to detect verifiable claims, their veracity, span, and check-worthiness. Persuasive technique detection has also been explored for images besides text-based content, e.g., introduced the task of propaganda in memes.",
                "cite_spans": [
                    {
                        "start": 348,
                        "end": 368,
                        "text": "(Strus et al., 2019;",
                        "ref_id": "BIBREF57"
                    },
                    {
                        "start": 369,
                        "end": 391,
                        "text": "Zampieri et al., 2019b",
                        "ref_id": "BIBREF68"
                    },
                    {
                        "start": 392,
                        "end": 415,
                        "text": "Zampieri et al., , 2020",
                        "ref_id": "BIBREF69"
                    },
                    {
                        "start": 656,
                        "end": 680,
                        "text": "(Zampieri et al., 2019b;",
                        "ref_id": "BIBREF68"
                    },
                    {
                        "start": 681,
                        "end": 703,
                        "text": "Zampieri et al., 2020;",
                        "ref_id": "BIBREF69"
                    },
                    {
                        "start": 704,
                        "end": 730,
                        "text": "Chakravarthi et al., 2021;",
                        "ref_id": "BIBREF6"
                    },
                    {
                        "start": 731,
                        "end": 750,
                        "text": "Modha et al., 2021)",
                        "ref_id": "BIBREF33"
                    },
                    {
                        "start": 986,
                        "end": 1004,
                        "text": "(Aly et al., 2021;",
                        "ref_id": null
                    },
                    {
                        "start": 1005,
                        "end": 1024,
                        "text": "Shaar et al., 2021;",
                        "ref_id": "BIBREF50"
                    },
                    {
                        "start": 1025,
                        "end": 1054,
                        "text": "Da San Martino et al., 2020a)",
                        "ref_id": "BIBREF8"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Related Work",
                "sec_num": "2"
            },
            {
                "text": "Some tasks have attempted to address affect concerning various targets. Xu et al. (2016) focused on stance prediction for given targets, i.e., whether the comment is in favour or against the target, both in supervised and in unsupervised scenarios. Molla and Joshi (2019) COVID-19  Train  2,700  163  576  317 2,438 3,494  Val  300  19  65  40  268  392  Test  381  18  106  50  359  533  Total  3,381  200  747  407 3,065 4,419   Politics  Train  2,852  230 1,308  441 2,617 4,596  Val  350  27  166  58  317  568  Test  350  31  167  45  308  551  Total  3,552 288 1,641 544 3,242 5,715 In contrast, here we focus not only on the polarity of the target entity, but also on understanding complex connotations such as glorification, vilification, and victimisation in memes. This is both challenging and important, as memetic discourse has taken over a sizable portion of online engagement and as it requires specialised moderation given its multimodal nature.",
                "cite_spans": [
                    {
                        "start": 72,
                        "end": 88,
                        "text": "Xu et al. (2016)",
                        "ref_id": "BIBREF62"
                    },
                    {
                        "start": 249,
                        "end": 271,
                        "text": "Molla and Joshi (2019)",
                        "ref_id": "BIBREF34"
                    }
                ],
                "ref_spans": [
                    {
                        "start": 272,
                        "end": 562,
                        "text": "COVID-19  Train  2,700  163  576  317 2,438 3,494  Val  300  19  65  40  268  392  Test  381  18  106  50  359  533  Total  3,381  200  747  407 3,065 4,419   Politics  Train  2,852  230 1,308  441 2,617 4,596  Val  350  27  166  58  317  568  Test  350  31  167  45  308  551  Total  3,552",
                        "ref_id": "TABREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Related Work",
                "sec_num": "2"
            },
            {
                "text": "Towards curating a dataset that would enable the identification of hero, villain, and victim as roles in memes, we leveraged and reannotated the HarMeme dataset released in (Pramanick et al., 2021b) , and we call this new dataset HVVMemes. HarMeme includes 3,544 memes about COVID-19 and 3,552 memes about US Politics, which are annotated for harmfulness as well as for target type, in case the meme is harmful, with four categories for the latter: individual, organisation, community, and society. Table 1 gives some statistics about HVVMemes (note that for COVID-19, we filtered out some of the memes in HarMeme, keeping 3,381 of the original 3,554 memes). As a general trend for both domains, we observe a neutral reference for most of the entities mentioned in the memes (3,065 for COVID-19, and 3,242 for US Politics); for such cases, we assign a fourth category: other. We further see that villain is the second most frequent role (747 memes for COVID-19, and 1,641 for US Politics), followed by victim (407 memes for COVID-19, and 544 for US Politics), and then hero (200 memes for COVID-19, and 288 for US Politics). We believe that this is a realistic representation of social media engagement involving memes, which are mostly humorous with neutral connotations, and less frequently harmful by indulging in vilification. Victimisation can also be interpreted as a countering resistance to incessant vilification. Finally, glorification is generally the weakest voice in memetic discourse.",
                "cite_spans": [
                    {
                        "start": 173,
                        "end": 198,
                        "text": "(Pramanick et al., 2021b)",
                        "ref_id": "BIBREF44"
                    }
                ],
                "ref_spans": [
                    {
                        "start": 499,
                        "end": 506,
                        "text": "Table 1",
                        "ref_id": "TABREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Dataset Curation",
                "sec_num": "3"
            },
            {
                "text": "Meme author's perspective needs to be considered as the frame of reference, while assigning roles. 2",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Annotation Guidelines 1",
                "sec_num": null
            },
            {
                "text": "Towards complete assimilation, both visual and textual cues should be factored in. 3",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Annotation Guidelines 1",
                "sec_num": null
            },
            {
                "text": "Relevant background context should be acquired before assigning roles. 4 Ambiguous memes can be categorised as other.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Annotation Guidelines 1",
                "sec_num": null
            },
            {
                "text": "A 3-point Likert scale based mental frame of reference, implying negative, neutral and positive sentiments involved, should steer the connotation adjudication. 6",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "5",
                "sec_num": null
            },
            {
                "text": "All reasonably intelligible (without ambiguity) entities that are referred to in the meme must be considered as valid targets. 7",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "5",
                "sec_num": null
            },
            {
                "text": "Entities with multiple interpretations should be categorised as other. 8",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "5",
                "sec_num": null
            },
            {
                "text": "The role of the original speaker of a quote, as expressed within a meme, must not be presumed. ",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "5",
                "sec_num": null
            },
            {
                "text": "Since entity role labelling is complex and subjective, we formulated clear annotation guidelines, which are summarized in Table 2 . Each meme was annotated by three annotators, and the disagreements were resolved with the help of a consolidator. We asked the annotators (i) to identify the entities, and (ii) to assign roles to these entities.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 122,
                        "end": 129,
                        "text": "Table 2",
                        "ref_id": "TABREF2"
                    }
                ],
                "eq_spans": [],
                "section": "Annotation Setup",
                "sec_num": "3.1"
            },
            {
                "text": "This step requires the annotators to elicit all entities that the meme refers to. This includes persons, norp (nationalities, religious, or political groups), facilities, organizations, geopolitical entities, locations, products, and other, as defined by spaCy's label scheme for named entity recognition. 1 1 spacy.io/models/en#en_core_web_sm",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Identifying the Entities",
                "sec_num": "3.1.1"
            },
            {
                "text": "To assist the annotators, we provided them an exhaustive list of all automatically identified entities along with resolution remarks whenever needed as shown in Table 3 . Note that the annotators were not restricted to select entities from our provided list, which can be error-prone as automatic named entity recognition is not perfect; in fact, they were encouraged to add additional entities as needed, e.g., such shown in the image, but not mentioned in the textual part of the meme. Fig. 2a shows a word cloud visualization of the entities referenced in COVID-19 memes: we can see social, global, political, and economic entities such as coronavirus, China, home, Wuhan, mask, work, etc. Similarly, in Fig. 2b shows a word cloud for US Politics memes, where we see entities like Biden, party, Donald, Democratic, Obama, etc.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 161,
                        "end": 168,
                        "text": "Table 3",
                        "ref_id": "TABREF3"
                    },
                    {
                        "start": 488,
                        "end": 495,
                        "text": "Fig. 2a",
                        "ref_id": "FIGREF1"
                    },
                    {
                        "start": 707,
                        "end": 714,
                        "text": "Fig. 2b",
                        "ref_id": "FIGREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Identifying the Entities",
                "sec_num": "3.1.1"
            },
            {
                "text": "To assess the general agreement between the annotators, we considered an agreement towards entity identification if at least two annotators agreed on an entity in the meme. The number of memes with agreed entities was normalised by the total number of memes with at least one valid entity assignment by the annotators. This was done independently of the implied role category, as the emphasis in this first step is on entity identification. The highest agreement towards this was 0.98, which suggests the reliability associated with the annotator's collective understanding of the task. We followed a similar approach for the overall rolewise inter-annotator agreement; see below.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Identifying the Entities",
                "sec_num": "3.1.1"
            },
            {
                "text": "The annotation was done in three stages: (i) dryrun, (ii) complete annotation, and (iii) consolidation. As part of the dry-run, the annotators and the consolidator annotated a random subset of 250 memes, assigning the entities the roles of hero, villain, victim, and other. Then, we gave them feedback and we trained them carefully by issuing detailed guidelines that included the formal definitions of the role categories and the instructions exemplifying the edge scenarios identified as part of the dry-run disagreements. In the second stage, the annotators performed a complete annotation. This was followed by a third consolidation stage with the help of a consolidator.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Role Assignment",
                "sec_num": "3.1.2"
            },
            {
                "text": "Due to the varying annotation responses and coreferencing for each role, conventional annotation agreement measures are not suitable for our setup. We consider an agreement when at least two annotators agree on one of the candidate entities for a particular role, which we formalize as the following role-wise agreement score a:",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Role Assignment",
                "sec_num": "3.1.2"
            },
            {
                "text": "EQUATION",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [
                    {
                        "start": 0,
                        "end": 8,
                        "text": "EQUATION",
                        "ref_id": "EQREF",
                        "raw_str": "a = v agr v tot",
                        "eq_num": "(1)"
                    }
                ],
                "section": "Role Assignment",
                "sec_num": "3.1.2"
            },
            {
                "text": "We define v agr , which refers to the total number of valid agreements, and v tot , which is the total number of valid responses, as follows:",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Role Assignment",
                "sec_num": "3.1.2"
            },
            {
                "text": "v agr = N i=1 I i ; v tot = N i=1 Z i (2)",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Role Assignment",
                "sec_num": "3.1.2"
            },
            {
                "text": "where I i is a valid agreement (1, iff two or more annotators agree on an entity in example i), Z i is a valid response (1, iff at least one annotator provides a valid entity as a response in example i), and N is the total number of examples in the dataset. In the first dry-run stage of the annotation process, the annotators worked on 250 memes, and then we examined their agreement, which was 0.50, 0.35, 0.14, and 0.55, for the roles of hero, villain, victim, and other, respectively, for COVID-19 and US Politics combined. The inter-annotator agreement for stages 2 and 3 is shown in Table 4 . We can see that the average agreement scores after the completion stage (stage-2) are 0.35 and 0.48 for COVID-19 and US Politics, respectively. After the consolidation stage (stage-3), these numbers increased to 0.58 and 0.64, respectively.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 589,
                        "end": 596,
                        "text": "Table 4",
                        "ref_id": "TABREF5"
                    }
                ],
                "eq_spans": [],
                "section": "Role Assignment",
                "sec_num": "3.1.2"
            },
            {
                "text": "The distribution of the referencing entities within our HVVMemes dataset is somewhat skewed towards specific entities as well as towards specific predominant roles for these specific entities. The entities fairly emulate the prevalent trends and discourse topics that social media engagement around the period of the dataset collection reflected, which was at the onset of the COVID-19 pandemic and the surrounding political outlook within the United States of America. We observed that entities like Donald Trump and China were referenced almost equally in COVID-19 memes as a villain and other, while other entities are invariably referenced as other using humor, sarcasm, limerick, etc. For the domain of US Politics, on one hand, entities like Donald Trump, the Democratic Party, the Republican Party, and the Democrats are observed to have similar trend of pre-dominantly being referenced as a villain and other, and on the other hand, as a general trend, most of the memes have at least one vilified reference. Table 5 : Leaderboard summary for the shared task.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 1017,
                        "end": 1024,
                        "text": "Table 5",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "Role-wise Analysis of HVVMemes",
                "sec_num": "3.2"
            },
            {
                "text": "The CONSTRAINT 22 Shared Task on Detecting the Hero, Villain, and the Victim in Memes asked to predict which entities are glorified, vilified, and victimised in a given meme. We gave the participants the above-described labeled training and validation datasets, where for each meme, we had the list of corresponding entities and their labeled role. The task was, given a meme and a list of entities, to predict the role of each of these entities in the meme. We provided the data split by topic (COVID-19 and US Politics), as discussed in Section 3. For the test set, we combined and shuffled the memes from the two topics, and we provided the memes with a list of corresponding entities, but no labels. The task was organized on CodaLab, an opensource platform widely used to host machine learning and data science competitions. Our competition link 2 provided all the necessary resources for the participants including archived news, notifications, and forum posts communicated during the running of the competition. We allowed the participants a maximum of 25 submissions, and the best submission was considered for the leaderboard.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Shared Task Details",
                "sec_num": "4"
            },
            {
                "text": "The official evaluation measure was macro-F1 score, as we have an imbalanced multi-class problem. We further report precision and recall.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Shared Task Details",
                "sec_num": "4"
            },
            {
                "text": "The total of 105 teams registered for the competition, and nine of them made submissions to the leaderboard, making a total of 71 attempts to improve their scores. The teams tried a variety of approaches, and below we discuss the approaches by the six teams who also submitted a system description paper with information about their runs.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Participation and Results",
                "sec_num": "5"
            },
            {
                "text": "\u2022 shiroe/jayeshbankoti (Kun et al., 2022) achieved the best results overall. One of the distinctive approaches that the authors followed was to make use of Celebrity face detection from the input meme images using Giphy's Github. 3 In addition, a sub-image detector using YoloV5 4 leveraged the bounding boxes for memes with multiple images. This was input into an ensemble model of DeBERTa (He et al., 2021 ) + RoBERTa ) + ViLT ) + EfficientNetB7 (Tan and Le, 2019) with averaging of the predictions in the final layer. Though they incorporated a celebrity detector, the lack of other external knowledge limited their system performance. Their source code is available at https: //bitbucket.org/logicallydevs/ constraint_2022/src/master/ \u2022 c1pher (Singh et al., 2022) were ranked third.",
                "cite_spans": [
                    {
                        "start": 23,
                        "end": 41,
                        "text": "(Kun et al., 2022)",
                        "ref_id": "BIBREF23"
                    },
                    {
                        "start": 230,
                        "end": 231,
                        "text": "3",
                        "ref_id": null
                    },
                    {
                        "start": 391,
                        "end": 407,
                        "text": "(He et al., 2021",
                        "ref_id": "BIBREF18"
                    },
                    {
                        "start": 448,
                        "end": 466,
                        "text": "(Tan and Le, 2019)",
                        "ref_id": "BIBREF59"
                    },
                    {
                        "start": 748,
                        "end": 768,
                        "text": "(Singh et al., 2022)",
                        "ref_id": "BIBREF56"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Participation and Results",
                "sec_num": "5"
            },
            {
                "text": "It is remarkable that they achieved this result using just the text input. They formulated the problem as a Multiple Choice Question Answering Task (MCQA), and they used an ensemble of three modules: twitter-xlmroberta + COVID-BERT (M\u00fcller et al., 2020) + BERT-tweet (Nguyen et al., 2020) . They further added a sentiment module trained using RoBERTa, with the final classification layer comprising Support Vector Machine (SVM). A major drawback of this approach is that they ignored the image as an input altogether.",
                "cite_spans": [
                    {
                        "start": 232,
                        "end": 253,
                        "text": "(M\u00fcller et al., 2020)",
                        "ref_id": "BIBREF37"
                    },
                    {
                        "start": 256,
                        "end": 288,
                        "text": "BERT-tweet (Nguyen et al., 2020)",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Participation and Results",
                "sec_num": "5"
            },
            {
                "text": "\u2022 zhouziming/zjl123001 (Zhou et al., 2022) leveraged the Visual Commonsense Reasoning (VCR) framework in a multimodal model. They built an ensemble of VisualBERT (Li et al., 2019 ) + UNITER (Chen et al., 2020) + OSCAR ) + ERNIE-Vil (Yu et al., 2021) , combined using an SVM. To handle the disproportionately large number of Other examples, they introduced lossreweighting. The lack of sufficient external knowledge and position information about the OCR text with the image restricted their system performance. Their source code is available at https://github.com/ zjl123001/DD-TIG-Constraint System BERT R-BERT D-BERT CLIP EB7 OFA ViLT ViT VB U O E-V SVM XGB BF VADER W-P shiroe c1pher zhouziming smontariol IIITDWD rabindra.nath Table 6 : Models used by the participants as part of their system submissions. R-BERT: RoBERTa, D-BERT: DeBERTa, EB7: EfficientNetB7, OFA: Once-for-All, ViLT: Visual and Language Transformer, ViT: Visual Transformer, VB: Visual BERT, U: UNITER, O: OSCAR, E-V: ERNIE-Vil, SVM: Support Vector Machines, XGB: XGBoost, BF: Block Fusion and W-P: Wu-Palmer.",
                "cite_spans": [
                    {
                        "start": 23,
                        "end": 42,
                        "text": "(Zhou et al., 2022)",
                        "ref_id": "BIBREF73"
                    },
                    {
                        "start": 162,
                        "end": 178,
                        "text": "(Li et al., 2019",
                        "ref_id": "BIBREF25"
                    },
                    {
                        "start": 190,
                        "end": 209,
                        "text": "(Chen et al., 2020)",
                        "ref_id": "BIBREF7"
                    },
                    {
                        "start": 232,
                        "end": 249,
                        "text": "(Yu et al., 2021)",
                        "ref_id": "BIBREF63"
                    }
                ],
                "ref_spans": [
                    {
                        "start": 731,
                        "end": 738,
                        "text": "Table 6",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "Participation and Results",
                "sec_num": "5"
            },
            {
                "text": "\u2022 smontariol (Montariol et al., 2022) experimented with sampling to handle data imbalance, trying six strategies. On top of that, they used an ensemble of CLIP (Radford et al., 2021) + VisualBERT + OFA (Cai et al., 2020) with XGBoost as the final layer for classification. The potential limitations of this approach include OCR errors and issues with imagetext correspondence. Their source code is available at https://github.com/ smontariol/mmsrl_constraint",
                "cite_spans": [
                    {
                        "start": 13,
                        "end": 37,
                        "text": "(Montariol et al., 2022)",
                        "ref_id": null
                    },
                    {
                        "start": 160,
                        "end": 182,
                        "text": "(Radford et al., 2021)",
                        "ref_id": "BIBREF45"
                    },
                    {
                        "start": 202,
                        "end": 220,
                        "text": "(Cai et al., 2020)",
                        "ref_id": "BIBREF5"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Participation and Results",
                "sec_num": "5"
            },
            {
                "text": "\u2022 IIITDWD (Fharook, 2022) combined sentiment-and lexicon-based approaches to associate sentiment polarity and roles with each entity. For sentiment classification, they used VADER 5 . Moreover, to associate commonly used words for hero, villain, and victim, they developed a corpus and used Wu-Palmer similarity. 6 The way was done and its impact are described in insufficient detail. Their source code is available at https: //github.com/fharookshaik/ shared-task_constraint-2022",
                "cite_spans": [
                    {
                        "start": 313,
                        "end": 314,
                        "text": "6",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Participation and Results",
                "sec_num": "5"
            },
            {
                "text": "\u2022 rabindra.nath (Nandi et al., 2022) proposed an approach using BLOCK fusion (Benyounes et al., 2019) for combining the image with text embeddings. They used a combination of ViT (Bobicev and Sokolova, 2017) and BERT (Devlin et al., 2019) for the image and for the text, respectively, followed by SVM as the final layer for classification. The empirical approach limits their system performance despite adding several data augmentation techniques. Their source code is available at https://github.com/robi56/ harmful_memes_block_fusion 5 https://pypi.org/project/ vaderSentiment/ 6 https://arxiv.org/ftp/arxiv/papers/ 1310/1310.8059.pdf",
                "cite_spans": [
                    {
                        "start": 16,
                        "end": 36,
                        "text": "(Nandi et al., 2022)",
                        "ref_id": "BIBREF40"
                    },
                    {
                        "start": 179,
                        "end": 207,
                        "text": "(Bobicev and Sokolova, 2017)",
                        "ref_id": "BIBREF4"
                    },
                    {
                        "start": 217,
                        "end": 238,
                        "text": "(Devlin et al., 2019)",
                        "ref_id": "BIBREF10"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Participation and Results",
                "sec_num": "5"
            },
            {
                "text": "The evaluation results for the above systems are shown in Table 5 . We can see that the macro-F1 scores range between 58.67 and 23.72, with a mean of 44.31 and a median of 48.48. Table 6 further gives a summary of the most important components of the participating systems.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 58,
                        "end": 65,
                        "text": "Table 5",
                        "ref_id": null
                    },
                    {
                        "start": 179,
                        "end": 186,
                        "text": "Table 6",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "Participation and Results",
                "sec_num": "5"
            },
            {
                "text": "We can see that one commonly used architecture is BERT and its variants, including multi-modal variants, whereas SVM is the preferred way to combine the components of ensemble systems.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Participation and Results",
                "sec_num": "5"
            },
            {
                "text": "Understanding and interpreting the connotations behind the entities in a meme is a difficult problem, which we pioneered in this shared task. Given a meme and a list of entities, the task asks to detect the role of each entity as a hero, a villain, a victim, or other. We curated HVVMemes, a largescale meme dataset of 7,000 memes spanning the domains of COVID-19 and US Politics, annotated with the entities they refer to as well as with their role. The shared task attracted 105 registered participants, out of which nine made official submissions, and six submitted papers describing their systems. We hope that our dataset and task setup will enable further research towards understanding how entities are portrayed in memes.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Conclusion",
                "sec_num": "6"
            },
            {
                "text": "https://codalab.lisn.upsaclay.fr/ competitions/906",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "http://github.com/Giphy/ celeb-detection-oss 4 https://github.com/ultralytics/yolov5",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            }
        ],
        "back_matter": [
            {
                "text": "The work was partially supported by a Wipro research grant, Ramanujan Fellowship, the Infosys Centre for AI, IIIT Delhi, and ihub-Anubhuti-iiitd Foundation, set up under the NM-ICPS scheme of the Department of Science and Technology, India. It is also part of the Tanbih mega-project, which is developed at the Qatar Computing Research Institute, HBKU, and aims to limit the impact of \"fake news,\" propaganda, and media bias by making users aware of what they are reading, thus promoting media literacy and critical thinking.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Acknowledgments",
                "sec_num": null
            }
        ],
        "bib_entries": {
            "BIBREF0": {
                "ref_id": "b0",
                "title": "Hamed Firooz, and Preslav Nakov. 2021a. A survey on multimodal disinformation detection",
                "authors": [
                    {
                        "first": "Firoj",
                        "middle": [],
                        "last": "Alam",
                        "suffix": ""
                    },
                    {
                        "first": "Stefano",
                        "middle": [],
                        "last": "Cresci",
                        "suffix": ""
                    },
                    {
                        "first": "Tanmoy",
                        "middle": [],
                        "last": "Chakraborty",
                        "suffix": ""
                    },
                    {
                        "first": "Fabrizio",
                        "middle": [],
                        "last": "Silvestri",
                        "suffix": ""
                    },
                    {
                        "first": "Dimiter",
                        "middle": [],
                        "last": "Dimitrov",
                        "suffix": ""
                    },
                    {
                        "first": "Giovanni",
                        "middle": [],
                        "last": "Da San",
                        "suffix": ""
                    },
                    {
                        "first": "Shaden",
                        "middle": [],
                        "last": "Martino",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Shaar",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Firoj Alam, Stefano Cresci, Tanmoy Chakraborty, Fab- rizio Silvestri, Dimiter Dimitrov, Giovanni Da San Martino, Shaden Shaar, Hamed Firooz, and Preslav Nakov. 2021a. A survey on multimodal disinforma- tion detection. CoRR, abs/2103.12541.",
                "links": null
            },
            "BIBREF1": {
                "ref_id": "b1",
                "title": "Fighting the COVID-19 infodemic: Modeling the perspective of journalists, fact-checkers, social media platforms, policy makers, and the society",
                "authors": [
                    {
                        "first": "Firoj",
                        "middle": [],
                        "last": "Alam",
                        "suffix": ""
                    },
                    {
                        "first": "Shaden",
                        "middle": [],
                        "last": "Shaar",
                        "suffix": ""
                    },
                    {
                        "first": "Fahim",
                        "middle": [],
                        "last": "Dalvi",
                        "suffix": ""
                    },
                    {
                        "first": "Hassan",
                        "middle": [],
                        "last": "Sajjad",
                        "suffix": ""
                    },
                    {
                        "first": "Alex",
                        "middle": [],
                        "last": "Nikolov",
                        "suffix": ""
                    },
                    {
                        "first": "Hamdy",
                        "middle": [],
                        "last": "Mubarak",
                        "suffix": ""
                    },
                    {
                        "first": "Giovanni",
                        "middle": [],
                        "last": "Da San",
                        "suffix": ""
                    },
                    {
                        "first": "Ahmed",
                        "middle": [],
                        "last": "Martino",
                        "suffix": ""
                    },
                    {
                        "first": "Nadir",
                        "middle": [],
                        "last": "Abdelali",
                        "suffix": ""
                    },
                    {
                        "first": "Kareem",
                        "middle": [],
                        "last": "Durrani",
                        "suffix": ""
                    },
                    {
                        "first": "Abdulaziz",
                        "middle": [],
                        "last": "Darwish",
                        "suffix": ""
                    },
                    {
                        "first": "Wajdi",
                        "middle": [],
                        "last": "Al-Homaid",
                        "suffix": ""
                    },
                    {
                        "first": "Tommaso",
                        "middle": [],
                        "last": "Zaghouani",
                        "suffix": ""
                    },
                    {
                        "first": "Gijs",
                        "middle": [],
                        "last": "Caselli",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Danoe",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Findings of EMNLP",
                "volume": "",
                "issue": "",
                "pages": "611--649",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2021.findings-emnlp.56"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Firoj Alam, Shaden Shaar, Fahim Dalvi, Hassan Saj- jad, Alex Nikolov, Hamdy Mubarak, Giovanni Da San Martino, Ahmed Abdelali, Nadir Durrani, Kareem Darwish, Abdulaziz Al-Homaid, Wajdi Za- ghouani, Tommaso Caselli, Gijs Danoe, Friso Stolk, Britt Bruntink, and Preslav Nakov. 2021b. Fighting the COVID-19 infodemic: Modeling the perspec- tive of journalists, fact-checkers, social media plat- forms, policy makers, and the society. In Findings of EMNLP, pages 611-649, Punta Cana, Dominican Republic.",
                "links": null
            },
            "BIBREF2": {
                "ref_id": "b2",
                "title": "Andreas Vlachos, Christos Christodoulopoulos, Oana Cocarascu, and Arpit Mittal. 2021. The fact extraction and VERification over unstructured and structured information (FEVEROUS) shared task",
                "authors": [
                    {
                        "first": "Rami",
                        "middle": [],
                        "last": "Aly",
                        "suffix": ""
                    },
                    {
                        "first": "Zhijiang",
                        "middle": [],
                        "last": "Guo",
                        "suffix": ""
                    },
                    {
                        "first": "Michael",
                        "middle": [
                            "Sejr"
                        ],
                        "last": "Schlichtkrull",
                        "suffix": ""
                    },
                    {
                        "first": "James",
                        "middle": [],
                        "last": "Thorne",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "Proceedings of the Fourth Workshop on Fact Extraction and VERification (FEVER)",
                "volume": "",
                "issue": "",
                "pages": "1--13",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2021.fever-1.1"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Rami Aly, Zhijiang Guo, Michael Sejr Schlichtkrull, James Thorne, Andreas Vlachos, Christos Christodoulopoulos, Oana Cocarascu, and Arpit Mittal. 2021. The fact extraction and VERification over unstructured and structured information (FEVEROUS) shared task. In Proceedings of the Fourth Workshop on Fact Extraction and VERification (FEVER), pages 1-13, Punta Cana, Dominican Republic.",
                "links": null
            },
            "BIBREF3": {
                "ref_id": "b3",
                "title": "BLOCK: Bilinear superdiagonal fusion for visual question answering and visual relationship detection",
                "authors": [
                    {
                        "first": "Hedi",
                        "middle": [],
                        "last": "Ben-Younes",
                        "suffix": ""
                    },
                    {
                        "first": "Remi",
                        "middle": [],
                        "last": "Cadene",
                        "suffix": ""
                    },
                    {
                        "first": "Nicolas",
                        "middle": [],
                        "last": "Thome",
                        "suffix": ""
                    },
                    {
                        "first": "Matthieu",
                        "middle": [],
                        "last": "Cord",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the Thirty-Third AAAI Conference on Artificial Intelligence, AAAI '19",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.1609/aaai.v33i01.33018102"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Hedi Ben-younes, Remi Cadene, Nicolas Thome, and Matthieu Cord. 2019. BLOCK: Bilinear superdiag- onal fusion for visual question answering and vi- sual relationship detection. In Proceedings of the Thirty-Third AAAI Conference on Artificial Intelli- gence, AAAI '19, Honolulu, Hawaii, USA.",
                "links": null
            },
            "BIBREF4": {
                "ref_id": "b4",
                "title": "Interannotator agreement in sentiment analysis: Machine learning perspective",
                "authors": [
                    {
                        "first": "Victoria",
                        "middle": [],
                        "last": "Bobicev",
                        "suffix": ""
                    },
                    {
                        "first": "Marina",
                        "middle": [],
                        "last": "Sokolova",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "Proceedings of the International Conference Recent Advances in Natural Language Processing, RANLP '17",
                "volume": "",
                "issue": "",
                "pages": "97--102",
                "other_ids": {
                    "DOI": [
                        "10.26615/978-954-452-049-6_015"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Victoria Bobicev and Marina Sokolova. 2017. Inter- annotator agreement in sentiment analysis: Machine learning perspective. In Proceedings of the Interna- tional Conference Recent Advances in Natural Lan- guage Processing, RANLP '17, pages 97-102, Varna, Bulgaria.",
                "links": null
            },
            "BIBREF5": {
                "ref_id": "b5",
                "title": "Once-for-All: Train One Network and Specialize it for Efficient Deployment",
                "authors": [
                    {
                        "first": "Han",
                        "middle": [],
                        "last": "Cai",
                        "suffix": ""
                    },
                    {
                        "first": "Chuang",
                        "middle": [],
                        "last": "Gan",
                        "suffix": ""
                    },
                    {
                        "first": "Tianzhe",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    },
                    {
                        "first": "Zhekai",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Song",
                        "middle": [],
                        "last": "Han",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:1908.09791"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Han Cai, Chuang Gan, Tianzhe Wang, Zhekai Zhang, and Song Han. 2020. Once-for-All: Train One Network and Specialize it for Efficient Deployment. arXiv:1908.09791 [cs, stat].",
                "links": null
            },
            "BIBREF6": {
                "ref_id": "b6",
                "title": "Findings of the shared task on offensive language identification in Tamil, Malayalam, and Kannada",
                "authors": [
                    {
                        "first": "Ruba",
                        "middle": [],
                        "last": "Bharathi Raja Chakravarthi",
                        "suffix": ""
                    },
                    {
                        "first": "Navya",
                        "middle": [],
                        "last": "Priyadharshini",
                        "suffix": ""
                    },
                    {
                        "first": "Anand",
                        "middle": [],
                        "last": "Jose",
                        "suffix": ""
                    },
                    {
                        "first": "M",
                        "middle": [],
                        "last": "Kumar",
                        "suffix": ""
                    },
                    {
                        "first": "Thomas",
                        "middle": [],
                        "last": "Mandl",
                        "suffix": ""
                    },
                    {
                        "first": "Prasanna",
                        "middle": [],
                        "last": "Kumar Kumaresan",
                        "suffix": ""
                    },
                    {
                        "first": "Rahul",
                        "middle": [],
                        "last": "Ponnusamy",
                        "suffix": ""
                    },
                    {
                        "first": "R L",
                        "middle": [],
                        "last": "Hariharan",
                        "suffix": ""
                    },
                    {
                        "first": "John",
                        "middle": [
                            "P"
                        ],
                        "last": "Mccrae",
                        "suffix": ""
                    },
                    {
                        "first": "Elizabeth",
                        "middle": [],
                        "last": "Sherly",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the First Workshop on Speech and Language Technologies for Dravidian Languages",
                "volume": "",
                "issue": "",
                "pages": "133--145",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Bharathi Raja Chakravarthi, Ruba Priyadharshini, Navya Jose, Anand Kumar M, Thomas Mandl, Prasanna Kumar Kumaresan, Rahul Ponnusamy, Har- iharan R L, John P. McCrae, and Elizabeth Sherly. 2021. Findings of the shared task on offensive lan- guage identification in Tamil, Malayalam, and Kan- nada. In Proceedings of the First Workshop on Speech and Language Technologies for Dravidian Languages, pages 133-145, Kyiv.",
                "links": null
            },
            "BIBREF7": {
                "ref_id": "b7",
                "title": "UNITER: UNiversal Image-TExt Representation Learning",
                "authors": [
                    {
                        "first": "Yen-Chun",
                        "middle": [],
                        "last": "Chen",
                        "suffix": ""
                    },
                    {
                        "first": "Linjie",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Licheng",
                        "middle": [],
                        "last": "Yu",
                        "suffix": ""
                    },
                    {
                        "first": "Ahmed",
                        "middle": [
                            "El"
                        ],
                        "last": "Kholy",
                        "suffix": ""
                    },
                    {
                        "first": "Faisal",
                        "middle": [],
                        "last": "Ahmed",
                        "suffix": ""
                    },
                    {
                        "first": "Zhe",
                        "middle": [],
                        "last": "Gan",
                        "suffix": ""
                    },
                    {
                        "first": "Yu",
                        "middle": [],
                        "last": "Cheng",
                        "suffix": ""
                    },
                    {
                        "first": "Jingjing",
                        "middle": [],
                        "last": "Liu",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:1909.11740"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Yen-Chun Chen, Linjie Li, Licheng Yu, Ahmed El Kholy, Faisal Ahmed, Zhe Gan, Yu Cheng, and Jingjing Liu. 2020. UNITER: UNiversal Image-TExt Representation Learning. arXiv:1909.11740 [cs].",
                "links": null
            },
            "BIBREF8": {
                "ref_id": "b8",
                "title": "SemEval-2020 task 11: Detection of propaganda techniques in news articles",
                "authors": [
                    {
                        "first": "Giovanni",
                        "middle": [],
                        "last": "Da San",
                        "suffix": ""
                    },
                    {
                        "first": "Alberto",
                        "middle": [],
                        "last": "Martino",
                        "suffix": ""
                    },
                    {
                        "first": "Henning",
                        "middle": [],
                        "last": "Barr\u00f3n-Cede\u00f1o",
                        "suffix": ""
                    },
                    {
                        "first": "Rostislav",
                        "middle": [],
                        "last": "Wachsmuth",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Petrov",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of the Fourteenth Workshop on Semantic Evaluation",
                "volume": "",
                "issue": "",
                "pages": "1377--1414",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2020.semeval-1.186"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Giovanni Da San Martino, Alberto Barr\u00f3n-Cede\u00f1o, Henning Wachsmuth, Rostislav Petrov, and Preslav Nakov. 2020a. SemEval-2020 task 11: Detection of propaganda techniques in news articles. In Pro- ceedings of the Fourteenth Workshop on Semantic Evaluation, pages 1377-1414, Barcelona (online).",
                "links": null
            },
            "BIBREF9": {
                "ref_id": "b9",
                "title": "A survey on computational propaganda detection",
                "authors": [
                    {
                        "first": "Giovanni",
                        "middle": [],
                        "last": "Da San",
                        "suffix": ""
                    },
                    {
                        "first": "Stefano",
                        "middle": [],
                        "last": "Martino",
                        "suffix": ""
                    },
                    {
                        "first": "Alberto",
                        "middle": [],
                        "last": "Cresci",
                        "suffix": ""
                    },
                    {
                        "first": "Seunghak",
                        "middle": [],
                        "last": "Barr\u00f3n-Cede\u00f1o",
                        "suffix": ""
                    },
                    {
                        "first": "Roberto",
                        "middle": [
                            "Di"
                        ],
                        "last": "Yu",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Pietro",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of the 29th International Joint Conference on Artificial Intelligence and the 17th Pacific Rim International Conference on Artificial Intelligence, IJCAI-PRICAI '20",
                "volume": "",
                "issue": "",
                "pages": "4826--4832",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Giovanni Da San Martino, Stefano Cresci, Alberto Barr\u00f3n-Cede\u00f1o, Seunghak Yu, Roberto Di Pietro, and Preslav Nakov. 2020b. A survey on computa- tional propaganda detection. In Proceedings of the 29th International Joint Conference on Artificial In- telligence and the 17th Pacific Rim International Con- ference on Artificial Intelligence, IJCAI-PRICAI '20, pages 4826-4832, Yokohama, Japan.",
                "links": null
            },
            "BIBREF10": {
                "ref_id": "b10",
                "title": "BERT: Pre-training of deep bidirectional transformers for language understanding",
                "authors": [
                    {
                        "first": "Jacob",
                        "middle": [],
                        "last": "Devlin",
                        "suffix": ""
                    },
                    {
                        "first": "Ming-Wei",
                        "middle": [],
                        "last": "Chang",
                        "suffix": ""
                    },
                    {
                        "first": "Kenton",
                        "middle": [],
                        "last": "Lee",
                        "suffix": ""
                    },
                    {
                        "first": "Kristina",
                        "middle": [],
                        "last": "Toutanova",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 2019 Conf. of the North American Chapter of the Association for Comp. Linguistics: Human Language Technologies",
                "volume": "",
                "issue": "",
                "pages": "4171--4186",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/N19-1423"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2019. BERT: Pre-training of deep bidirectional transformers for language under- standing. In Proceedings of the 2019 Conf. of the North American Chapter of the Association for Comp. Linguistics: Human Language Technologies, pages 4171-4186, Minneapolis, Minnesota.",
                "links": null
            },
            "BIBREF11": {
                "ref_id": "b11",
                "title": "Detecting propaganda techniques in memes",
                "authors": [
                    {
                        "first": "Dimitar",
                        "middle": [],
                        "last": "Dimitrov",
                        "suffix": ""
                    },
                    {
                        "first": "Shaden",
                        "middle": [],
                        "last": "Bishr Bin Ali",
                        "suffix": ""
                    },
                    {
                        "first": "Firoj",
                        "middle": [],
                        "last": "Shaar",
                        "suffix": ""
                    },
                    {
                        "first": "Fabrizio",
                        "middle": [],
                        "last": "Alam",
                        "suffix": ""
                    },
                    {
                        "first": "Hamed",
                        "middle": [],
                        "last": "Silvestri",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Firooz",
                        "suffix": ""
                    },
                    {
                        "first": "Giovanni Da San",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Martino",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the Joint Conference of the 59th Annual Meeting of the Association for Computational Linguistics and the 11th International Joint Conference on Natural Language Processing, ACL-IJCNLP '21",
                "volume": "",
                "issue": "",
                "pages": "6603--6617",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Dimitar Dimitrov, Bishr Bin Ali, Shaden Shaar, Firoj Alam, Fabrizio Silvestri, Hamed Firooz, Preslav Nakov, and Giovanni Da San Martino. 2021a. De- tecting propaganda techniques in memes. In Pro- ceedings of the Joint Conference of the 59th Annual Meeting of the Association for Computational Lin- guistics and the 11th International Joint Conference on Natural Language Processing, ACL-IJCNLP '21, pages 6603-6617.",
                "links": null
            },
            "BIBREF12": {
                "ref_id": "b12",
                "title": "SemEval-2021 task 6: Detection of persuasion techniques in texts and images",
                "authors": [
                    {
                        "first": "Dimitar",
                        "middle": [],
                        "last": "Dimitrov",
                        "suffix": ""
                    },
                    {
                        "first": "Shaden",
                        "middle": [],
                        "last": "Bishr Bin Ali",
                        "suffix": ""
                    },
                    {
                        "first": "Firoj",
                        "middle": [],
                        "last": "Shaar",
                        "suffix": ""
                    },
                    {
                        "first": "Fabrizio",
                        "middle": [],
                        "last": "Alam",
                        "suffix": ""
                    },
                    {
                        "first": "Hamed",
                        "middle": [],
                        "last": "Silvestri",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Firooz",
                        "suffix": ""
                    },
                    {
                        "first": "Giovanni Da San",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Martino",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the 15th International Workshop on Semantic Evaluation (SemEval-2021)",
                "volume": "",
                "issue": "",
                "pages": "70--98",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2021.semeval-1.7"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Dimitar Dimitrov, Bishr Bin Ali, Shaden Shaar, Firoj Alam, Fabrizio Silvestri, Hamed Firooz, Preslav Nakov, and Giovanni Da San Martino. 2021b. SemEval-2021 task 6: Detection of persuasion tech- niques in texts and images. In Proceedings of the 15th International Workshop on Semantic Evaluation (SemEval-2021), pages 70-98, Online.",
                "links": null
            },
            "BIBREF13": {
                "ref_id": "b13",
                "title": "Are you a hero or a villain? a semantic role labelling approach for detecting harmful memes",
                "authors": [],
                "year": null,
                "venue": "Proceedings of the Workshop on Combating Online Hostile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Shaik Fharook. 2022. Are you a hero or a villain? a se- mantic role labelling approach for detecting harmful memes. In Proceedings of the Workshop on Com- bating Online Hostile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22, Dublin, Ireland.",
                "links": null
            },
            "BIBREF14": {
                "ref_id": "b14",
                "title": "Ramit Sawhney, and Rajiv Ratn Shah. 2020. #MeTooMA: Multi-aspect annotations of tweets related to the MeToo movement. Proceedings of the International AAAI Conference on Web and Social Media",
                "authors": [
                    {
                        "first": "Akash",
                        "middle": [],
                        "last": "Gautam",
                        "suffix": ""
                    },
                    {
                        "first": "Puneet",
                        "middle": [],
                        "last": "Mathur",
                        "suffix": ""
                    },
                    {
                        "first": "Rakesh",
                        "middle": [],
                        "last": "Gosangi",
                        "suffix": ""
                    },
                    {
                        "first": "Debanjan",
                        "middle": [],
                        "last": "Mahata",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "",
                "volume": "14",
                "issue": "",
                "pages": "209--216",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Akash Gautam, Puneet Mathur, Rakesh Gosangi, De- banjan Mahata, Ramit Sawhney, and Rajiv Ratn Shah. 2020. #MeTooMA: Multi-aspect annotations of tweets related to the MeToo movement. Proceed- ings of the International AAAI Conference on Web and Social Media, 14(1):209-216.",
                "links": null
            },
            "BIBREF15": {
                "ref_id": "b15",
                "title": "Exploring hate speech detection in multimodal publications",
                "authors": [
                    {
                        "first": "Raul",
                        "middle": [],
                        "last": "Gomez",
                        "suffix": ""
                    },
                    {
                        "first": "Jaume",
                        "middle": [],
                        "last": "Gibert",
                        "suffix": ""
                    },
                    {
                        "first": "Lluis",
                        "middle": [],
                        "last": "Gomez",
                        "suffix": ""
                    },
                    {
                        "first": "Dimosthenis",
                        "middle": [],
                        "last": "Karatzas",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of the 2020 IEEE Winter Conference on Applications of Computer Vision, WACV '20",
                "volume": "",
                "issue": "",
                "pages": "1459--1467",
                "other_ids": {
                    "DOI": [
                        "10.1109/WACV45572.2020.9093414"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Raul Gomez, Jaume Gibert, Lluis Gomez, and Dimos- thenis Karatzas. 2020. Exploring hate speech detec- tion in multimodal publications. In Proceedings of the 2020 IEEE Winter Conference on Applications of Computer Vision, WACV '20, pages 1459-1467.",
                "links": null
            },
            "BIBREF16": {
                "ref_id": "b16",
                "title": "Who is the hero, the villain, and the victim? Detection of roles in news articles using natural language techniques",
                "authors": [
                    {
                        "first": "Diego",
                        "middle": [],
                        "last": "Gomez-Zara",
                        "suffix": ""
                    },
                    {
                        "first": "Miriam",
                        "middle": [],
                        "last": "Boon",
                        "suffix": ""
                    },
                    {
                        "first": "Larry",
                        "middle": [],
                        "last": "Birnbaum",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "23rd International Conference on Intelligent User Interfaces, IUI '18",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.1145/3172944.3172993"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Diego Gomez-Zara, Miriam Boon, and Larry Birnbaum. 2018. Who is the hero, the villain, and the victim? Detection of roles in news articles using natural lan- guage techniques. In 23rd International Conference on Intelligent User Interfaces, IUI '18, page 311315, Tokyo, Japan.",
                "links": null
            },
            "BIBREF17": {
                "ref_id": "b17",
                "title": "Preslav Nakov, and Isabelle Augenstein. 2022. A survey on stance detection for mis-and disinformation identification",
                "authors": [
                    {
                        "first": "Momchil",
                        "middle": [],
                        "last": "Hardalov",
                        "suffix": ""
                    },
                    {
                        "first": "Arnav",
                        "middle": [],
                        "last": "Arora",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "Findings of NAACL",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Momchil Hardalov, Arnav Arora, Preslav Nakov, and Isabelle Augenstein. 2022. A survey on stance de- tection for mis-and disinformation identification. In Findings of NAACL, Seattle, Washington, USA.",
                "links": null
            },
            "BIBREF18": {
                "ref_id": "b18",
                "title": "DeBERTa: Decodingenhanced BERT with Disentangled Attention",
                "authors": [
                    {
                        "first": "Pengcheng",
                        "middle": [],
                        "last": "He",
                        "suffix": ""
                    },
                    {
                        "first": "Xiaodong",
                        "middle": [],
                        "last": "Liu",
                        "suffix": ""
                    },
                    {
                        "first": "Jianfeng",
                        "middle": [],
                        "last": "Gao",
                        "suffix": ""
                    },
                    {
                        "first": "Weizhu",
                        "middle": [],
                        "last": "Chen",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:2006.03654"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Pengcheng He, Xiaodong Liu, Jianfeng Gao, and Weizhu Chen. 2021. DeBERTa: Decoding- enhanced BERT with Disentangled Attention. arXiv:2006.03654 [cs].",
                "links": null
            },
            "BIBREF19": {
                "ref_id": "b19",
                "title": "Fairface: Face attribute dataset for balanced race, gender, and age for bias measurement and mitigation",
                "authors": [
                    {
                        "first": "Kimmo",
                        "middle": [],
                        "last": "Karkkainen",
                        "suffix": ""
                    },
                    {
                        "first": "Jungseock",
                        "middle": [],
                        "last": "Joo",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, WACV '21",
                "volume": "",
                "issue": "",
                "pages": "1548--1558",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Kimmo Karkkainen and Jungseock Joo. 2021. Fair- face: Face attribute dataset for balanced race, gender, and age for bias measurement and mitigation. In Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, WACV '21, pages 1548-1558.",
                "links": null
            },
            "BIBREF20": {
                "ref_id": "b20",
                "title": "Pratik Ringshia, and Davide Testuggine. 2020. The hateful memes challenge: Detecting hate speech in multimodal memes",
                "authors": [
                    {
                        "first": "Douwe",
                        "middle": [],
                        "last": "Kiela",
                        "suffix": ""
                    },
                    {
                        "first": "Hamed",
                        "middle": [],
                        "last": "Firooz",
                        "suffix": ""
                    },
                    {
                        "first": "Aravind",
                        "middle": [],
                        "last": "Mohan",
                        "suffix": ""
                    },
                    {
                        "first": "Vedanuj",
                        "middle": [],
                        "last": "Goswami",
                        "suffix": ""
                    },
                    {
                        "first": "Amanpreet",
                        "middle": [],
                        "last": "Singh",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "Proceedings of the 34th International Conference on Neural Information Processing Systems",
                "volume": "33",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Douwe Kiela, Hamed Firooz, Aravind Mohan, Vedanuj Goswami, Amanpreet Singh, Pratik Ringshia, and Davide Testuggine. 2020. The hateful memes chal- lenge: Detecting hate speech in multimodal memes. In Proceedings of the 34th International Confer- ence on Neural Information Processing Systems, vol- ume 33 of NeurIPS '20.",
                "links": null
            },
            "BIBREF21": {
                "ref_id": "b21",
                "title": "ViLT: Vision-and-language transformer without convolution or region supervision",
                "authors": [
                    {
                        "first": "Wonjae",
                        "middle": [],
                        "last": "Kim",
                        "suffix": ""
                    },
                    {
                        "first": "Bokyung",
                        "middle": [],
                        "last": "Son",
                        "suffix": ""
                    },
                    {
                        "first": "Ildoo",
                        "middle": [],
                        "last": "Kim",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the 38th International Conference on Machine Learning",
                "volume": "139",
                "issue": "",
                "pages": "5583--5594",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Wonjae Kim, Bokyung Son, and Ildoo Kim. 2021. ViLT: Vision-and-language transformer without convolu- tion or region supervision. In Proceedings of the 38th International Conference on Machine Learning, volume 139 of ICML '21, pages 5583-5594.",
                "links": null
            },
            "BIBREF22": {
                "ref_id": "b22",
                "title": "Sarc-M: Sarcasm detection in typo-graphic memes",
                "authors": [
                    {
                        "first": "Akshi",
                        "middle": [],
                        "last": "Kumar",
                        "suffix": ""
                    },
                    {
                        "first": "Geetanjali",
                        "middle": [],
                        "last": "Garg",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the International Conference on Advances in Engineering Science Management & Technology, ICAESMT '19",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Akshi Kumar and Geetanjali Garg. 2019. Sarc-M: Sar- casm detection in typo-graphic memes. In Proceed- ings of the International Conference on Advances in Engineering Science Management & Technology, ICAESMT '19, Dehradun, India.",
                "links": null
            },
            "BIBREF23": {
                "ref_id": "b23",
                "title": "Logically at the CONSTRAINT 2022: Multimodal role labelling",
                "authors": [
                    {
                        "first": "Ludovic",
                        "middle": [],
                        "last": "Kun",
                        "suffix": ""
                    },
                    {
                        "first": "Jayesh",
                        "middle": [],
                        "last": "Bankoti",
                        "suffix": ""
                    },
                    {
                        "first": "David",
                        "middle": [],
                        "last": "Kiskovski",
                        "suffix": ""
                    }
                ],
                "year": 2022,
                "venue": "Proceedings of the Workshop on Combating Online Hostile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Ludovic Kun, Jayesh Bankoti, and David Kiskovski. 2022. Logically at the CONSTRAINT 2022: Multi- modal role labelling. In Proceedings of the Work- shop on Combating Online Hostile Posts in Re- gional Languages during Emergency Situations, CONSTRAINT '22, Dublin, Ireland.",
                "links": null
            },
            "BIBREF24": {
                "ref_id": "b24",
                "title": "The science of fake news",
                "authors": [
                    {
                        "first": "M",
                        "middle": [
                            "J"
                        ],
                        "last": "David",
                        "suffix": ""
                    },
                    {
                        "first": "Matthew",
                        "middle": [
                            "A"
                        ],
                        "last": "Lazer",
                        "suffix": ""
                    },
                    {
                        "first": "Yochai",
                        "middle": [],
                        "last": "Baum",
                        "suffix": ""
                    },
                    {
                        "first": "Adam",
                        "middle": [
                            "J"
                        ],
                        "last": "Benkler",
                        "suffix": ""
                    },
                    {
                        "first": "Kelly",
                        "middle": [
                            "M"
                        ],
                        "last": "Berinsky",
                        "suffix": ""
                    },
                    {
                        "first": "Filippo",
                        "middle": [],
                        "last": "Greenhill",
                        "suffix": ""
                    },
                    {
                        "first": "Miriam",
                        "middle": [
                            "J"
                        ],
                        "last": "Menczer",
                        "suffix": ""
                    },
                    {
                        "first": "Brendan",
                        "middle": [],
                        "last": "Metzger",
                        "suffix": ""
                    },
                    {
                        "first": ";",
                        "middle": [],
                        "last": "Nyhan",
                        "suffix": ""
                    },
                    {
                        "first": "A",
                        "middle": [],
                        "last": "Steven",
                        "suffix": ""
                    },
                    {
                        "first": "Cass",
                        "middle": [
                            "R"
                        ],
                        "last": "Sloman",
                        "suffix": ""
                    },
                    {
                        "first": "Emily",
                        "middle": [
                            "A"
                        ],
                        "last": "Sunstein",
                        "suffix": ""
                    },
                    {
                        "first": "Duncan",
                        "middle": [
                            "J"
                        ],
                        "last": "Thorson",
                        "suffix": ""
                    },
                    {
                        "first": "Jonathan",
                        "middle": [
                            "L"
                        ],
                        "last": "Watts",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Zittrain",
                "volume": "359",
                "issue": "6380",
                "pages": "1094--1096",
                "other_ids": {
                    "DOI": [
                        "10.1126/science.aao2998"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "David M.J. Lazer, Matthew A. Baum, Yochai Ben- kler, Adam J. Berinsky, Kelly M. Greenhill, Filippo Menczer, Miriam J. Metzger, Brendan Nyhan, Gor- don Pennycook, David Rothschild, Michael Schud- son, Steven A. Sloman, Cass R. Sunstein, Emily A. Thorson, Duncan J. Watts, and Jonathan L. Zit- train. 2018. The science of fake news. Science, 359(6380):1094-1096.",
                "links": null
            },
            "BIBREF25": {
                "ref_id": "b25",
                "title": "VisualBERT: A Simple and Performant Baseline for Vision and Language",
                "authors": [
                    {
                        "first": "Liunian Harold",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Mark",
                        "middle": [],
                        "last": "Yatskar",
                        "suffix": ""
                    },
                    {
                        "first": "Da",
                        "middle": [],
                        "last": "Yin",
                        "suffix": ""
                    },
                    {
                        "first": "Cho-Jui",
                        "middle": [],
                        "last": "Hsieh",
                        "suffix": ""
                    },
                    {
                        "first": "Kai-Wei",
                        "middle": [],
                        "last": "Chang",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:1908.03557"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Liunian Harold Li, Mark Yatskar, Da Yin, Cho-Jui Hsieh, and Kai-Wei Chang. 2019. VisualBERT: A Simple and Performant Baseline for Vision and Lan- guage. arXiv:1908.03557 [cs].",
                "links": null
            },
            "BIBREF26": {
                "ref_id": "b26",
                "title": "Oscar: Object-semantics aligned pre-training for vision-language tasks",
                "authors": [
                    {
                        "first": "Xiujun",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Xi",
                        "middle": [],
                        "last": "Yin",
                        "suffix": ""
                    },
                    {
                        "first": "Chunyuan",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Pengchuan",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Xiaowei",
                        "middle": [],
                        "last": "Hu",
                        "suffix": ""
                    },
                    {
                        "first": "Lei",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Lijuan",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    },
                    {
                        "first": "Houdong",
                        "middle": [],
                        "last": "Hu",
                        "suffix": ""
                    },
                    {
                        "first": "Li",
                        "middle": [],
                        "last": "Dong",
                        "suffix": ""
                    },
                    {
                        "first": "Furu",
                        "middle": [],
                        "last": "Wei",
                        "suffix": ""
                    },
                    {
                        "first": "Yejin",
                        "middle": [],
                        "last": "Choi",
                        "suffix": ""
                    },
                    {
                        "first": "Jianfeng",
                        "middle": [],
                        "last": "Gao",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of the 16th European Conference in Computer Vision",
                "volume": "12375",
                "issue": "",
                "pages": "121--137",
                "other_ids": {
                    "DOI": [
                        "10.1007/978-3-030-58577-8_8"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Xiujun Li, Xi Yin, Chunyuan Li, Pengchuan Zhang, Xiaowei Hu, Lei Zhang, Lijuan Wang, Houdong Hu, Li Dong, Furu Wei, Yejin Choi, and Jianfeng Gao. 2020. Oscar: Object-semantics aligned pre-training for vision-language tasks. In Proceedings of the 16th European Conference in Computer Vision, volume 12375 of ECCV '20, pages 121-137, Glasgow, UK.",
                "links": null
            },
            "BIBREF27": {
                "ref_id": "b27",
                "title": "RoBERTa: A Robustly Optimized BERT Pretraining Approach",
                "authors": [
                    {
                        "first": "Yinhan",
                        "middle": [],
                        "last": "Liu",
                        "suffix": ""
                    },
                    {
                        "first": "Myle",
                        "middle": [],
                        "last": "Ott",
                        "suffix": ""
                    },
                    {
                        "first": "Naman",
                        "middle": [],
                        "last": "Goyal",
                        "suffix": ""
                    },
                    {
                        "first": "Jingfei",
                        "middle": [],
                        "last": "Du",
                        "suffix": ""
                    },
                    {
                        "first": "Mandar",
                        "middle": [],
                        "last": "Joshi",
                        "suffix": ""
                    },
                    {
                        "first": "Danqi",
                        "middle": [],
                        "last": "Chen",
                        "suffix": ""
                    },
                    {
                        "first": "Omer",
                        "middle": [],
                        "last": "Levy",
                        "suffix": ""
                    },
                    {
                        "first": "Mike",
                        "middle": [],
                        "last": "Lewis",
                        "suffix": ""
                    },
                    {
                        "first": "Luke",
                        "middle": [],
                        "last": "Zettlemoyer",
                        "suffix": ""
                    },
                    {
                        "first": "Veselin",
                        "middle": [],
                        "last": "Stoyanov",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:1907.11692"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Yinhan Liu, Myle Ott, Naman Goyal, Jingfei Du, Man- dar Joshi, Danqi Chen, Omer Levy, Mike Lewis, Luke Zettlemoyer, and Veselin Stoyanov. 2019. RoBERTa: A Robustly Optimized BERT Pretrain- ing Approach. arXiv:1907.11692 [cs].",
                "links": null
            },
            "BIBREF28": {
                "ref_id": "b28",
                "title": "Joint learning for targeted sentiment analysis",
                "authors": [
                    {
                        "first": "Dehong",
                        "middle": [],
                        "last": "Ma",
                        "suffix": ""
                    },
                    {
                        "first": "Sujian",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Houfeng",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the 2018 Conference on Empirical Methods in Natural Language Processing",
                "volume": "",
                "issue": "",
                "pages": "4737--4742",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/D18-1504"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Dehong Ma, Sujian Li, and Houfeng Wang. 2018. Joint learning for targeted sentiment analysis. In Proceed- ings of the 2018 Conference on Empirical Methods in Natural Language Processing, pages 4737-4742, Brussels, Belgium.",
                "links": null
            },
            "BIBREF29": {
                "ref_id": "b29",
                "title": "Hate speech detection: Challenges and solutions",
                "authors": [
                    {
                        "first": "Sean",
                        "middle": [],
                        "last": "Macavaney",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Hao-Ren",
                        "suffix": ""
                    },
                    {
                        "first": "Eugene",
                        "middle": [],
                        "last": "Yao",
                        "suffix": ""
                    },
                    {
                        "first": "Katina",
                        "middle": [],
                        "last": "Yang",
                        "suffix": ""
                    },
                    {
                        "first": "Nazli",
                        "middle": [],
                        "last": "Russell",
                        "suffix": ""
                    },
                    {
                        "first": "Ophir",
                        "middle": [],
                        "last": "Goharian",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Frieder",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "PLOS ONE",
                "volume": "14",
                "issue": "8",
                "pages": "1--16",
                "other_ids": {
                    "DOI": [
                        "10.1371/journal.pone.0221152"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Sean MacAvaney, Hao-Ren Yao, Eugene Yang, Katina Russell, Nazli Goharian, and Ophir Frieder. 2019. Hate speech detection: Challenges and solutions. PLOS ONE, 14(8):1-16.",
                "links": null
            },
            "BIBREF30": {
                "ref_id": "b30",
                "title": "Hatexplain: A benchmark dataset for explainable hate speech detection",
                "authors": [
                    {
                        "first": "Binny",
                        "middle": [],
                        "last": "Mathew",
                        "suffix": ""
                    },
                    {
                        "first": "Punyajoy",
                        "middle": [],
                        "last": "Saha",
                        "suffix": ""
                    },
                    {
                        "first": "Chris",
                        "middle": [],
                        "last": "Seid Muhie Yimam",
                        "suffix": ""
                    },
                    {
                        "first": "Pawan",
                        "middle": [],
                        "last": "Biemann",
                        "suffix": ""
                    },
                    {
                        "first": "Animesh",
                        "middle": [],
                        "last": "Goyal",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Mukherjee",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the AAAI Conference on Artificial Intelligence",
                "volume": "35",
                "issue": "",
                "pages": "14867--14875",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Binny Mathew, Punyajoy Saha, Seid Muhie Yimam, Chris Biemann, Pawan Goyal, and Animesh Mukher- jee. 2021. Hatexplain: A benchmark dataset for explainable hate speech detection. Proceedings of the AAAI Conference on Artificial Intelligence, 35(17):14867-14875.",
                "links": null
            },
            "BIBREF31": {
                "ref_id": "b31",
                "title": "Batman, Pandaman and the Blind Man: A case study in social change memes and internet censorship in China",
                "authors": [
                    {
                        "first": "An",
                        "middle": [
                            "Xiao"
                        ],
                        "last": "Mina",
                        "suffix": ""
                    }
                ],
                "year": 2014,
                "venue": "Journal of Visual Culture",
                "volume": "13",
                "issue": "3",
                "pages": "359--375",
                "other_ids": {
                    "DOI": [
                        "10.1177/1470412914546576"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "An Xiao Mina. 2014. Batman, Pandaman and the Blind Man: A case study in social change memes and inter- net censorship in China. Journal of Visual Culture, 13(3):359-375.",
                "links": null
            },
            "BIBREF32": {
                "ref_id": "b32",
                "title": "Open domain targeted sentiment",
                "authors": [
                    {
                        "first": "Margaret",
                        "middle": [],
                        "last": "Mitchell",
                        "suffix": ""
                    },
                    {
                        "first": "Jacqui",
                        "middle": [],
                        "last": "Aguilar",
                        "suffix": ""
                    },
                    {
                        "first": "Theresa",
                        "middle": [],
                        "last": "Wilson",
                        "suffix": ""
                    },
                    {
                        "first": "Benjamin",
                        "middle": [],
                        "last": "Van Durme",
                        "suffix": ""
                    }
                ],
                "year": 2013,
                "venue": "Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing",
                "volume": "",
                "issue": "",
                "pages": "1643--1654",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Margaret Mitchell, Jacqui Aguilar, Theresa Wilson, and Benjamin Van Durme. 2013. Open domain targeted sentiment. In Proceedings of the 2013 Conference on Empirical Methods in Natural Language Processing, pages 1643-1654, Seattle, Washington, USA.",
                "links": null
            },
            "BIBREF33": {
                "ref_id": "b33",
                "title": "Overview of the HASOC subtrack at FIRE 2021: Hate speech and offensive content identification in english and indoaryan languages and conversational hate speech",
                "authors": [
                    {
                        "first": "Sandip",
                        "middle": [],
                        "last": "Modha",
                        "suffix": ""
                    },
                    {
                        "first": "Thomas",
                        "middle": [],
                        "last": "Mandl",
                        "suffix": ""
                    },
                    {
                        "first": "Hiren",
                        "middle": [],
                        "last": "Gautam Kishore Shahi",
                        "suffix": ""
                    },
                    {
                        "first": "Shrey",
                        "middle": [],
                        "last": "Madhu",
                        "suffix": ""
                    },
                    {
                        "first": "Tharindu",
                        "middle": [],
                        "last": "Satapara",
                        "suffix": ""
                    },
                    {
                        "first": "Marcos",
                        "middle": [],
                        "last": "Ranasinghe",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Zampieri",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the Forum for Information Retrieval Evaluation, FIRE '21",
                "volume": "",
                "issue": "",
                "pages": "1--3",
                "other_ids": {
                    "DOI": [
                        "10.1145/3503162.3503176"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Sandip Modha, Thomas Mandl, Gautam Kishore Shahi, Hiren Madhu, Shrey Satapara, Tharindu Ranasinghe, and Marcos Zampieri. 2021. Overview of the HASOC subtrack at FIRE 2021: Hate speech and offensive content identification in english and indo- aryan languages and conversational hate speech. In Proceedings of the Forum for Information Retrieval Evaluation, FIRE '21, pages 1-3.",
                "links": null
            },
            "BIBREF34": {
                "ref_id": "b34",
                "title": "Overview of the 2019 ALTA shared task: Sarcasm target identification",
                "authors": [
                    {
                        "first": "Diego",
                        "middle": [],
                        "last": "Molla",
                        "suffix": ""
                    },
                    {
                        "first": "Aditya",
                        "middle": [],
                        "last": "Joshi",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 17th Annual Workshop of the Australasian Language Technology Association, ALTA '19",
                "volume": "",
                "issue": "",
                "pages": "192--196",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Diego Molla and Aditya Joshi. 2019. Overview of the 2019 ALTA shared task: Sarcasm target identifica- tion. In Proceedings of the 17th Annual Workshop of the Australasian Language Technology Association, ALTA '19, pages 192-196, Sydney, Australia.",
                "links": null
            },
            "BIBREF35": {
                "ref_id": "b35",
                "title": "2022. Fine-tuning and sampling strategies for multimodal role labeling of entities under class imbalance",
                "authors": [
                    {
                        "first": "Syrielle",
                        "middle": [],
                        "last": "Montariol",
                        "suffix": ""
                    },
                    {
                        "first": "\u00c9tienne",
                        "middle": [],
                        "last": "Simon",
                        "suffix": ""
                    },
                    {
                        "first": "Arij",
                        "middle": [],
                        "last": "Riabi",
                        "suffix": ""
                    },
                    {
                        "first": "Djam\u00e9",
                        "middle": [],
                        "last": "Seddah",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "Proceedings of the Workshop on Combating Online Hostile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Syrielle Montariol, \u00c9tienne Simon, Arij Riabi, and Djam\u00e9 Seddah. 2022. Fine-tuning and sampling strategies for multimodal role labeling of entities under class imbalance. In Proceedings of the Workshop on Combating Online Hostile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22, Dublin, Ireland.",
                "links": null
            },
            "BIBREF36": {
                "ref_id": "b36",
                "title": "Overview of OSACT4 Arabic offensive language detection shared task",
                "authors": [
                    {
                        "first": "Hamdy",
                        "middle": [],
                        "last": "Mubarak",
                        "suffix": ""
                    },
                    {
                        "first": "Kareem",
                        "middle": [],
                        "last": "Darwish",
                        "suffix": ""
                    },
                    {
                        "first": "Walid",
                        "middle": [],
                        "last": "Magdy",
                        "suffix": ""
                    },
                    {
                        "first": "Tamer",
                        "middle": [],
                        "last": "Elsayed",
                        "suffix": ""
                    },
                    {
                        "first": "Hend",
                        "middle": [],
                        "last": "Al-Khalifa",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of the 4th Workshop on Open-Source Arabic Corpora and Processing Tools, with a Shared Task on Offensive Language Detection",
                "volume": "",
                "issue": "",
                "pages": "48--52",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Hamdy Mubarak, Kareem Darwish, Walid Magdy, Tamer Elsayed, and Hend Al-Khalifa. 2020. Overview of OSACT4 Arabic offensive language de- tection shared task. In Proceedings of the 4th Work- shop on Open-Source Arabic Corpora and Process- ing Tools, with a Shared Task on Offensive Language Detection, pages 48-52, Marseille, France.",
                "links": null
            },
            "BIBREF37": {
                "ref_id": "b37",
                "title": "COVID-Twitter-BERT: A Natural Language Processing Model to Analyse COVID-19 Content on Twitter",
                "authors": [
                    {
                        "first": "Martin",
                        "middle": [],
                        "last": "M\u00fcller",
                        "suffix": ""
                    },
                    {
                        "first": "Marcel",
                        "middle": [],
                        "last": "Salath\u00e9",
                        "suffix": ""
                    },
                    {
                        "first": "Per",
                        "middle": [
                            "E"
                        ],
                        "last": "Kummervold",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:2005.07503"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Martin M\u00fcller, Marcel Salath\u00e9, and Per E. Kummervold. 2020. COVID-Twitter-BERT: A Natural Language Processing Model to Analyse COVID-19 Content on Twitter. arXiv:2005.07503 [cs].",
                "links": null
            },
            "BIBREF38": {
                "ref_id": "b38",
                "title": "Yavuz Selim Kartal, and Javier Beltr\u00e1n. 2022. The CLEF-2022 CheckThat! lab on fighting the COVID-19 infodemic and fake news detection",
                "authors": [
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "Alberto",
                        "middle": [],
                        "last": "Barr\u00f3n-Cede\u00f1o",
                        "suffix": ""
                    },
                    {
                        "first": "Giovanni",
                        "middle": [],
                        "last": "Da San",
                        "suffix": ""
                    },
                    {
                        "first": "Firoj",
                        "middle": [],
                        "last": "Martino",
                        "suffix": ""
                    },
                    {
                        "first": "Julia",
                        "middle": [
                            "Maria"
                        ],
                        "last": "Alam",
                        "suffix": ""
                    },
                    {
                        "first": "Thomas",
                        "middle": [],
                        "last": "Stru\u00df",
                        "suffix": ""
                    },
                    {
                        "first": "Rub\u00e9n",
                        "middle": [],
                        "last": "Mandl",
                        "suffix": ""
                    },
                    {
                        "first": "Tommaso",
                        "middle": [],
                        "last": "M\u00edguez",
                        "suffix": ""
                    },
                    {
                        "first": "Mucahid",
                        "middle": [],
                        "last": "Caselli",
                        "suffix": ""
                    },
                    {
                        "first": "Wajdi",
                        "middle": [],
                        "last": "Kutlu",
                        "suffix": ""
                    },
                    {
                        "first": "Chengkai",
                        "middle": [],
                        "last": "Zaghouani",
                        "suffix": ""
                    },
                    {
                        "first": "Shaden",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Shaar",
                        "suffix": ""
                    },
                    {
                        "first": "Hamdy",
                        "middle": [],
                        "last": "Gautam Kishore Shahi",
                        "suffix": ""
                    },
                    {
                        "first": "Alex",
                        "middle": [],
                        "last": "Mubarak",
                        "suffix": ""
                    },
                    {
                        "first": "Nikolay",
                        "middle": [],
                        "last": "Nikolov",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Babulkov",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "Proceedings of the 44th European Conference on IR Research: Advances in Information Retrieval, ECIR '22",
                "volume": "",
                "issue": "",
                "pages": "416--428",
                "other_ids": {
                    "DOI": [
                        "10.1007/978-3-030-99739-7_52"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Preslav Nakov, Alberto Barr\u00f3n-Cede\u00f1o, Giovanni Da San Martino, Firoj Alam, Julia Maria Stru\u00df, Thomas Mandl, Rub\u00e9n M\u00edguez, Tom- maso Caselli, Mucahid Kutlu, Wajdi Zaghouani, Chengkai Li, Shaden Shaar, Gautam Kishore Shahi, Hamdy Mubarak, Alex Nikolov, Nikolay Babulkov, Yavuz Selim Kartal, and Javier Beltr\u00e1n. 2022. The CLEF-2022 CheckThat! lab on fighting the COVID- 19 infodemic and fake news detection. In Proceed- ings of the 44th European Conference on IR Re- search: Advances in Information Retrieval, ECIR '22, pages 416-428, Berlin, Heidelberg.",
                "links": null
            },
            "BIBREF39": {
                "ref_id": "b39",
                "title": "Guillaume Bouchard, and Isabelle Augenstein. 2021. Detecting abusive language on online platforms: A critical analysis",
                "authors": [
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "Vibha",
                        "middle": [],
                        "last": "Nayak",
                        "suffix": ""
                    },
                    {
                        "first": "Kyle",
                        "middle": [],
                        "last": "Dent",
                        "suffix": ""
                    },
                    {
                        "first": "Ameya",
                        "middle": [],
                        "last": "Bhatawdekar",
                        "suffix": ""
                    },
                    {
                        "first": "Muhammad",
                        "middle": [],
                        "last": "Sheikh",
                        "suffix": ""
                    },
                    {
                        "first": "Momchil",
                        "middle": [],
                        "last": "Sarwar",
                        "suffix": ""
                    },
                    {
                        "first": "Yoan",
                        "middle": [],
                        "last": "Hardalov",
                        "suffix": ""
                    },
                    {
                        "first": "Dimitrina",
                        "middle": [],
                        "last": "Dinkov",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Zlatkova",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Preslav Nakov, Vibha Nayak, Kyle Dent, Ameya Bhatawdekar, Sheikh Muhammad Sarwar, Momchil Hardalov, Yoan Dinkov, Dimitrina Zlatkova, Guil- laume Bouchard, and Isabelle Augenstein. 2021. De- tecting abusive language on online platforms: A criti- cal analysis. arXiv/2103.00153.",
                "links": null
            },
            "BIBREF40": {
                "ref_id": "b40",
                "title": "Detecting the role of an entity in harmful memes: Techniques and their limitations",
                "authors": [
                    {
                        "first": "Firoj",
                        "middle": [],
                        "last": "Rabindra Nath Nandi",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Alam",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    }
                ],
                "year": 2022,
                "venue": "Proceedings of the Workshop on Combating Online Hostile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Rabindra Nath Nandi, Firoj Alam, and Preslav Nakov. 2022. Detecting the role of an entity in harmful memes: Techniques and their limitations. In Pro- ceedings of the Workshop on Combating Online Hos- tile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22, Dublin, Ireland.",
                "links": null
            },
            "BIBREF41": {
                "ref_id": "b41",
                "title": "BERTweet: A pre-trained language model for English tweets",
                "authors": [
                    {
                        "first": "Thanh",
                        "middle": [],
                        "last": "Dat Quoc Nguyen",
                        "suffix": ""
                    },
                    {
                        "first": "Anh",
                        "middle": [
                            "Tuan"
                        ],
                        "last": "Vu",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Nguyen",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing: System Demonstrations, EMNLP '20",
                "volume": "",
                "issue": "",
                "pages": "9--14",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2020.emnlp-demos.2"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Dat Quoc Nguyen, Thanh Vu, and Anh Tuan Nguyen. 2020. BERTweet: A pre-trained language model for English tweets. In Proceedings of the 2020 Con- ference on Empirical Methods in Natural Language Processing: System Demonstrations, EMNLP '20, pages 9-14, Online.",
                "links": null
            },
            "BIBREF42": {
                "ref_id": "b42",
                "title": "Multilingual and multi-aspect hate speech analysis",
                "authors": [
                    {
                        "first": "Nedjma",
                        "middle": [],
                        "last": "Ousidhoum",
                        "suffix": ""
                    },
                    {
                        "first": "Zizheng",
                        "middle": [],
                        "last": "Lin",
                        "suffix": ""
                    },
                    {
                        "first": "Hongming",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Yangqiu",
                        "middle": [],
                        "last": "Song",
                        "suffix": ""
                    },
                    {
                        "first": "Dit-Yan",
                        "middle": [],
                        "last": "Yeung",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP)",
                "volume": "",
                "issue": "",
                "pages": "4675--4684",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/D19-1474"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Nedjma Ousidhoum, Zizheng Lin, Hongming Zhang, Yangqiu Song, and Dit-Yan Yeung. 2019. Multi- lingual and multi-aspect hate speech analysis. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP), pages 4675-4684, Hong Kong, China.",
                "links": null
            },
            "BIBREF43": {
                "ref_id": "b43",
                "title": "Detecting harmful memes and their targets",
                "authors": [
                    {
                        "first": "Shraman",
                        "middle": [],
                        "last": "Pramanick",
                        "suffix": ""
                    },
                    {
                        "first": "Dimitar",
                        "middle": [],
                        "last": "Dimitrov",
                        "suffix": ""
                    },
                    {
                        "first": "Rituparna",
                        "middle": [],
                        "last": "Mukherjee",
                        "suffix": ""
                    },
                    {
                        "first": "Shivam",
                        "middle": [],
                        "last": "Sharma",
                        "suffix": ""
                    },
                    {
                        "first": "Md",
                        "middle": [
                            "Shad"
                        ],
                        "last": "Akhtar",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "Tanmoy",
                        "middle": [],
                        "last": "Chakraborty",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Findings of ACL, ACL-IJCNLP '21",
                "volume": "",
                "issue": "",
                "pages": "2783--2796",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2021.findings-acl.246"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Shraman Pramanick, Dimitar Dimitrov, Rituparna Mukherjee, Shivam Sharma, Md. Shad Akhtar, Preslav Nakov, and Tanmoy Chakraborty. 2021a. De- tecting harmful memes and their targets. In Findings of ACL, ACL-IJCNLP '21, pages 2783-2796.",
                "links": null
            },
            "BIBREF44": {
                "ref_id": "b44",
                "title": "MOMENTA: A multimodal framework for detecting harmful memes and their targets",
                "authors": [
                    {
                        "first": "Shraman",
                        "middle": [],
                        "last": "Pramanick",
                        "suffix": ""
                    },
                    {
                        "first": "Shivam",
                        "middle": [],
                        "last": "Sharma",
                        "suffix": ""
                    },
                    {
                        "first": "Dimitar",
                        "middle": [],
                        "last": "Dimitrov",
                        "suffix": ""
                    },
                    {
                        "first": "Md",
                        "middle": [
                            "Shad"
                        ],
                        "last": "Akhtar",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "Tanmoy",
                        "middle": [],
                        "last": "Chakraborty",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Findings of EMNLP 2021",
                "volume": "",
                "issue": "",
                "pages": "4439--4455",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Shraman Pramanick, Shivam Sharma, Dimitar Dim- itrov, Md. Shad Akhtar, Preslav Nakov, and Tanmoy Chakraborty. 2021b. MOMENTA: A multimodal framework for detecting harmful memes and their tar- gets. In Findings of EMNLP 2021, pages 4439-4455, Punta Cana, Dominican Republic.",
                "links": null
            },
            "BIBREF45": {
                "ref_id": "b45",
                "title": "Learning transferable visual models from natural language supervision",
                "authors": [
                    {
                        "first": "Alec",
                        "middle": [],
                        "last": "Radford",
                        "suffix": ""
                    },
                    {
                        "first": "Jong",
                        "middle": [
                            "Wook"
                        ],
                        "last": "Kim",
                        "suffix": ""
                    },
                    {
                        "first": "Chris",
                        "middle": [],
                        "last": "Hallacy",
                        "suffix": ""
                    },
                    {
                        "first": "Aditya",
                        "middle": [],
                        "last": "Ramesh",
                        "suffix": ""
                    },
                    {
                        "first": "Gabriel",
                        "middle": [],
                        "last": "Goh",
                        "suffix": ""
                    },
                    {
                        "first": "Sandhini",
                        "middle": [],
                        "last": "Agarwal",
                        "suffix": ""
                    },
                    {
                        "first": "Girish",
                        "middle": [],
                        "last": "Sastry",
                        "suffix": ""
                    },
                    {
                        "first": "Amanda",
                        "middle": [],
                        "last": "Askell",
                        "suffix": ""
                    },
                    {
                        "first": "Pamela",
                        "middle": [],
                        "last": "Mishkin",
                        "suffix": ""
                    },
                    {
                        "first": "Jack",
                        "middle": [],
                        "last": "Clark",
                        "suffix": ""
                    },
                    {
                        "first": "Gretchen",
                        "middle": [],
                        "last": "Krueger",
                        "suffix": ""
                    },
                    {
                        "first": "Ilya",
                        "middle": [],
                        "last": "Sutskever",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the 38th International Conference on Machine Learning",
                "volume": "139",
                "issue": "",
                "pages": "8748--8763",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sas- try, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, and Ilya Sutskever. 2021. Learn- ing transferable visual models from natural language supervision. In Proceedings of the 38th International Conference on Machine Learning, volume 139 of Proceedings of Machine Learning Research, pages 8748-8763.",
                "links": null
            },
            "BIBREF46": {
                "ref_id": "b46",
                "title": "SemEval-2017 task 4: Sentiment analysis in Twitter",
                "authors": [
                    {
                        "first": "Sara",
                        "middle": [],
                        "last": "Rosenthal",
                        "suffix": ""
                    },
                    {
                        "first": "Noura",
                        "middle": [],
                        "last": "Farra",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "Proceedings of the 11th International Workshop on Semantic Evaluation, SemEval '17",
                "volume": "",
                "issue": "",
                "pages": "502--518",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/S17-2088"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Sara Rosenthal, Noura Farra, and Preslav Nakov. 2017. SemEval-2017 task 4: Sentiment analysis in Twitter. In Proceedings of the 11th International Workshop on Semantic Evaluation, SemEval '17, pages 502-518, Vancouver, Canada.",
                "links": null
            },
            "BIBREF47": {
                "ref_id": "b47",
                "title": "Hate speech in pixels: Detection of offensive memes towards automatic moderation",
                "authors": [
                    {
                        "first": "",
                        "middle": [],
                        "last": "Benet Oriol",
                        "suffix": ""
                    },
                    {
                        "first": "Cristian Canton",
                        "middle": [],
                        "last": "Sabat",
                        "suffix": ""
                    },
                    {
                        "first": "Xavier Giro-I",
                        "middle": [],
                        "last": "Ferrer",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Nieto",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:1910.02334"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Benet Oriol Sabat, Cristian Canton Ferrer, and Xavier Giro-i Nieto. 2019. Hate speech in pixels: Detection of offensive memes towards automatic moderation. arXiv:1910.02334.",
                "links": null
            },
            "BIBREF48": {
                "ref_id": "b48",
                "title": "Social bias frames: Reasoning about social and power implications of language",
                "authors": [
                    {
                        "first": "Maarten",
                        "middle": [],
                        "last": "Sap",
                        "suffix": ""
                    },
                    {
                        "first": "Saadia",
                        "middle": [],
                        "last": "Gabriel",
                        "suffix": ""
                    },
                    {
                        "first": "Lianhui",
                        "middle": [],
                        "last": "Qin",
                        "suffix": ""
                    },
                    {
                        "first": "Dan",
                        "middle": [],
                        "last": "Jurafsky",
                        "suffix": ""
                    },
                    {
                        "first": "Noah",
                        "middle": [
                            "A"
                        ],
                        "last": "Smith",
                        "suffix": ""
                    },
                    {
                        "first": "Yejin",
                        "middle": [],
                        "last": "Choi",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of the 58th",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2020.acl-main.486"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Maarten Sap, Saadia Gabriel, Lianhui Qin, Dan Juraf- sky, Noah A. Smith, and Yejin Choi. 2020. Social bias frames: Reasoning about social and power im- plications of language. In Proceedings of the 58th",
                "links": null
            },
            "BIBREF49": {
                "ref_id": "b49",
                "title": "Annual Meeting of the Association for Computational Linguistics, ACL '20",
                "authors": [],
                "year": null,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "5477--5490",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Annual Meeting of the Association for Computational Linguistics, ACL '20, pages 5477-5490, Online.",
                "links": null
            },
            "BIBREF50": {
                "ref_id": "b50",
                "title": "Findings of the NLP4IF-2021 shared tasks on fighting the COVID-19 infodemic and censorship detection",
                "authors": [
                    {
                        "first": "Shaden",
                        "middle": [],
                        "last": "Shaar",
                        "suffix": ""
                    },
                    {
                        "first": "Firoj",
                        "middle": [],
                        "last": "Alam",
                        "suffix": ""
                    },
                    {
                        "first": "Giovanni",
                        "middle": [],
                        "last": "Da San",
                        "suffix": ""
                    },
                    {
                        "first": "Alex",
                        "middle": [],
                        "last": "Martino",
                        "suffix": ""
                    },
                    {
                        "first": "Wajdi",
                        "middle": [],
                        "last": "Nikolov",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Zaghouani",
                        "suffix": ""
                    },
                    {
                        "first": "Anna",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Feldman",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the Fourth Workshop on NLP for Internet Freedom: Censorship, Disinformation, and Propaganda",
                "volume": "",
                "issue": "",
                "pages": "82--92",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2021.nlp4if-1.12"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Shaden Shaar, Firoj Alam, Giovanni Da San Martino, Alex Nikolov, Wajdi Zaghouani, Preslav Nakov, and Anna Feldman. 2021. Findings of the NLP4IF-2021 shared tasks on fighting the COVID-19 infodemic and censorship detection. In Proceedings of the Fourth Workshop on NLP for Internet Freedom: Cen- sorship, Disinformation, and Propaganda, pages 82-92, Online.",
                "links": null
            },
            "BIBREF51": {
                "ref_id": "b51",
                "title": "KnowMeme: A knowledgeenriched graph neural network solution to offensive meme detection",
                "authors": [
                    {
                        "first": "Lanyu",
                        "middle": [],
                        "last": "Shang",
                        "suffix": ""
                    },
                    {
                        "first": "Christina",
                        "middle": [],
                        "last": "Youn",
                        "suffix": ""
                    },
                    {
                        "first": "Yuheng",
                        "middle": [],
                        "last": "Zha",
                        "suffix": ""
                    },
                    {
                        "first": "Yang",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Dong",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the 2021 IEEE 17th International Conference on eScience, eScience '21",
                "volume": "",
                "issue": "",
                "pages": "186--195",
                "other_ids": {
                    "DOI": [
                        "10.1109/eScience51609.2021.00029"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Lanyu Shang, Christina Youn, Yuheng Zha, Yang Zhang, and Dong Wang. 2021. KnowMeme: A knowledge- enriched graph neural network solution to offen- sive meme detection. In Proceedings of the 2021 IEEE 17th International Conference on eScience, eScience '21, pages 186-195.",
                "links": null
            },
            "BIBREF52": {
                "ref_id": "b52",
                "title": "Viswanath Pulabaigari, and Bj\u00f6rn Gamb\u00e4ck. 2020. SemEval-2020 task 8: Memotion analysis-the visuolingual metaphor! In Proceedings of the Fourteenth Workshop on Semantic Evaluation, SemEval '20",
                "authors": [
                    {
                        "first": "Chhavi",
                        "middle": [],
                        "last": "Sharma",
                        "suffix": ""
                    },
                    {
                        "first": "Deepesh",
                        "middle": [],
                        "last": "Bhageria",
                        "suffix": ""
                    },
                    {
                        "first": "William",
                        "middle": [],
                        "last": "Scott",
                        "suffix": ""
                    },
                    {
                        "first": "Pykl",
                        "middle": [],
                        "last": "Srinivas",
                        "suffix": ""
                    },
                    {
                        "first": "Amitava",
                        "middle": [],
                        "last": "Das",
                        "suffix": ""
                    },
                    {
                        "first": "Tanmoy",
                        "middle": [],
                        "last": "Chakraborty",
                        "suffix": ""
                    }
                ],
                "year": null,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "759--773",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Chhavi Sharma, Deepesh Bhageria, William Scott, Srinivas PYKL, Amitava Das, Tanmoy Chakraborty, Viswanath Pulabaigari, and Bj\u00f6rn Gamb\u00e4ck. 2020. SemEval-2020 task 8: Memotion analysis-the visuo- lingual metaphor! In Proceedings of the Fourteenth Workshop on Semantic Evaluation, SemEval '20, pages 759-773.",
                "links": null
            },
            "BIBREF53": {
                "ref_id": "b53",
                "title": "DISARM: Detecting the victims targeted by harmful memes",
                "authors": [
                    {
                        "first": "Shivam",
                        "middle": [],
                        "last": "Sharma",
                        "suffix": ""
                    },
                    {
                        "first": "Shad",
                        "middle": [],
                        "last": "Md",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Akhtar",
                        "suffix": ""
                    },
                    {
                        "first": "Tanmoy",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Chakraborty",
                        "suffix": ""
                    }
                ],
                "year": 2022,
                "venue": "Findings of NAACL",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Shivam Sharma, Md Shad Akhtar, Preslav Nakov, and Tanmoy Chakraborty. 2022a. DISARM: Detecting the victims targeted by harmful memes. In Findings of NAACL, Seattle, Washington, USA.",
                "links": null
            },
            "BIBREF54": {
                "ref_id": "b54",
                "title": "Detecting and understanding harmful memes: A survey",
                "authors": [
                    {
                        "first": "Shivam",
                        "middle": [],
                        "last": "Sharma",
                        "suffix": ""
                    },
                    {
                        "first": "Firoj",
                        "middle": [],
                        "last": "Alam",
                        "suffix": ""
                    },
                    {
                        "first": "Md",
                        "middle": [
                            "Shad"
                        ],
                        "last": "Akhtar",
                        "suffix": ""
                    },
                    {
                        "first": "Dimitar",
                        "middle": [],
                        "last": "Dimitrov",
                        "suffix": ""
                    },
                    {
                        "first": "Giovanni",
                        "middle": [],
                        "last": "Da San",
                        "suffix": ""
                    },
                    {
                        "first": "Hamed",
                        "middle": [],
                        "last": "Martino",
                        "suffix": ""
                    },
                    {
                        "first": "Alon",
                        "middle": [],
                        "last": "Firooz",
                        "suffix": ""
                    },
                    {
                        "first": "Fabrizio",
                        "middle": [],
                        "last": "Halevy",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Silvestri",
                        "suffix": ""
                    },
                    {
                        "first": "Tanmoy",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Chakraborty",
                        "suffix": ""
                    }
                ],
                "year": 2022,
                "venue": "Proceedings of the 31st International Joint Conference on Artificial Intelligence, IJCAI-ECAI '22",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Shivam Sharma, Firoj Alam, Md. Shad Akhtar, Dimitar Dimitrov, Giovanni Da San Martino, Hamed Firooz, Alon Halevy, Fabrizio Silvestri, Preslav Nakov, and Tanmoy Chakraborty. 2022b. Detecting and under- standing harmful memes: A survey. In Proceedings of the 31st International Joint Conference on Artifi- cial Intelligence, IJCAI-ECAI '22, Vienna, Austria.",
                "links": null
            },
            "BIBREF55": {
                "ref_id": "b55",
                "title": "Targets and aspects in social media hate speech",
                "authors": [
                    {
                        "first": "Alexander",
                        "middle": [],
                        "last": "Shvets",
                        "suffix": ""
                    },
                    {
                        "first": "Paula",
                        "middle": [],
                        "last": "Fortuna",
                        "suffix": ""
                    },
                    {
                        "first": "Juan",
                        "middle": [],
                        "last": "Soler",
                        "suffix": ""
                    },
                    {
                        "first": "Leo",
                        "middle": [],
                        "last": "Wanner",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the 5th Workshop on Online Abuse and Harms (WOAH 2021)",
                "volume": "",
                "issue": "",
                "pages": "179--190",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2021.woah-1.19"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Alexander Shvets, Paula Fortuna, Juan Soler, and Leo Wanner. 2021. Targets and aspects in social media hate speech. In Proceedings of the 5th Workshop on Online Abuse and Harms (WOAH 2021), pages 179-190, Online.",
                "links": null
            },
            "BIBREF56": {
                "ref_id": "b56",
                "title": "Combining language models and linguistic information to label entities in memes",
                "authors": [
                    {
                        "first": "Pranaydeep",
                        "middle": [],
                        "last": "Singh",
                        "suffix": ""
                    },
                    {
                        "first": "Aaron",
                        "middle": [],
                        "last": "Maladry",
                        "suffix": ""
                    },
                    {
                        "first": "Els",
                        "middle": [],
                        "last": "Lefever",
                        "suffix": ""
                    }
                ],
                "year": 2022,
                "venue": "Proceedings of the Workshop on Combating Online Hostile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Pranaydeep Singh, Aaron Maladry, and Els Lefever. 2022. Combining language models and linguistic in- formation to label entities in memes. In Proceedings of the Workshop on Combating Online Hostile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22, Dublin, Ireland.",
                "links": null
            },
            "BIBREF57": {
                "ref_id": "b57",
                "title": "Overview of GermEval task 2",
                "authors": [
                    {
                        "first": "Julia",
                        "middle": [
                            "Maria"
                        ],
                        "last": "Strus",
                        "suffix": ""
                    },
                    {
                        "first": "Melanie",
                        "middle": [],
                        "last": "Siegel",
                        "suffix": ""
                    },
                    {
                        "first": "Josef",
                        "middle": [],
                        "last": "Ruppenhofer",
                        "suffix": ""
                    },
                    {
                        "first": "Michael",
                        "middle": [],
                        "last": "Wiegand",
                        "suffix": ""
                    },
                    {
                        "first": "Manfred",
                        "middle": [],
                        "last": "Klenner",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "shared task on the identification of offensive language. Proceedings of the 15th Conference on Natural Language Processing",
                "volume": "",
                "issue": "",
                "pages": "352--363",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Julia Maria Strus, Melanie Siegel, Josef Ruppen- hofer, Michael Wiegand, and Manfred Klenner. 2019. Overview of GermEval task 2, 2019 shared task on the identification of offensive language. Proceed- ings of the 15th Conference on Natural Language Processing, pages 352 -363, M\u00fcnchen [u.a.].",
                "links": null
            },
            "BIBREF58": {
                "ref_id": "b58",
                "title": "Multimodal meme dataset (MultiOFF) for identifying offensive content in image and text",
                "authors": [
                    {
                        "first": "Shardul",
                        "middle": [],
                        "last": "Suryawanshi",
                        "suffix": ""
                    },
                    {
                        "first": "Mihael",
                        "middle": [],
                        "last": "Bharathi Raja Chakravarthi",
                        "suffix": ""
                    },
                    {
                        "first": "Paul",
                        "middle": [],
                        "last": "Arcan",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Buitelaar",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of the Second Workshop on Trolling, Aggression and Cyberbullying",
                "volume": "",
                "issue": "",
                "pages": "32--41",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Shardul Suryawanshi, Bharathi Raja Chakravarthi, Mi- hael Arcan, and Paul Buitelaar. 2020. Multimodal meme dataset (MultiOFF) for identifying offensive content in image and text. In Proceedings of the Second Workshop on Trolling, Aggression and Cyber- bullying, pages 32-41, Marseille, France.",
                "links": null
            },
            "BIBREF59": {
                "ref_id": "b59",
                "title": "EfficientNet: Rethinking model scaling for convolutional neural networks",
                "authors": [
                    {
                        "first": "Mingxing",
                        "middle": [],
                        "last": "Tan",
                        "suffix": ""
                    },
                    {
                        "first": "V",
                        "middle": [],
                        "last": "Quoc",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Le",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 36th International Conference on Machine Learning",
                "volume": "97",
                "issue": "",
                "pages": "6105--6114",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Mingxing Tan and Quoc V. Le. 2019. EfficientNet: Rethinking model scaling for convolutional neural networks. In Proceedings of the 36th International Conference on Machine Learning, volume 97 of ICML '19, pages 6105-6114, Long Beach, Califor- nia, USA.",
                "links": null
            },
            "BIBREF60": {
                "ref_id": "b60",
                "title": "The spread of true and false news online",
                "authors": [
                    {
                        "first": "Soroush",
                        "middle": [],
                        "last": "Vosoughi",
                        "suffix": ""
                    },
                    {
                        "first": "Deb",
                        "middle": [],
                        "last": "Roy",
                        "suffix": ""
                    },
                    {
                        "first": "Sinan",
                        "middle": [],
                        "last": "Aral",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Science",
                "volume": "359",
                "issue": "6380",
                "pages": "1146--1151",
                "other_ids": {
                    "DOI": [
                        "10.1126/science.aap9559"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Soroush Vosoughi, Deb Roy, and Sinan Aral. 2018. The spread of true and false news online. Science, 359(6380):1146-1151.",
                "links": null
            },
            "BIBREF61": {
                "ref_id": "b61",
                "title": "Misinformation in social media: Definition, manipulation, and detection",
                "authors": [
                    {
                        "first": "Liang",
                        "middle": [],
                        "last": "Wu",
                        "suffix": ""
                    },
                    {
                        "first": "Fred",
                        "middle": [],
                        "last": "Morstatter",
                        "suffix": ""
                    },
                    {
                        "first": "Kathleen",
                        "middle": [
                            "M"
                        ],
                        "last": "Carley",
                        "suffix": ""
                    },
                    {
                        "first": "Huan",
                        "middle": [],
                        "last": "Liu",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "SIGKDD Explor. Newsl",
                "volume": "21",
                "issue": "2",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.1145/3373464.3373475"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Liang Wu, Fred Morstatter, Kathleen M. Carley, and Huan Liu. 2019. Misinformation in social media: Definition, manipulation, and detection. SIGKDD Explor. Newsl., 21(2):8090.",
                "links": null
            },
            "BIBREF62": {
                "ref_id": "b62",
                "title": "Overview of NLPCC shared task 4: Stance detection in Chinese microblogs",
                "authors": [
                    {
                        "first": "Ruifeng",
                        "middle": [],
                        "last": "Xu",
                        "suffix": ""
                    },
                    {
                        "first": "Yu",
                        "middle": [],
                        "last": "Zhou",
                        "suffix": ""
                    },
                    {
                        "first": "Dongyin",
                        "middle": [],
                        "last": "Wu",
                        "suffix": ""
                    },
                    {
                        "first": "Lin",
                        "middle": [],
                        "last": "Gui",
                        "suffix": ""
                    },
                    {
                        "first": "Jiachen",
                        "middle": [],
                        "last": "Du",
                        "suffix": ""
                    },
                    {
                        "first": "Yun",
                        "middle": [],
                        "last": "Xue",
                        "suffix": ""
                    }
                ],
                "year": 2016,
                "venue": "Natural Language Understanding and Intelligent Applications",
                "volume": "",
                "issue": "",
                "pages": "907--916",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Ruifeng Xu, Yu Zhou, Dongyin Wu, Lin Gui, Jiachen Du, and Yun Xue. 2016. Overview of NLPCC shared task 4: Stance detection in Chinese microblogs. In Natural Language Understanding and Intelligent Ap- plications, pages 907-916, Cham.",
                "links": null
            },
            "BIBREF63": {
                "ref_id": "b63",
                "title": "ERNIE-ViL: Knowledge enhanced vision-language representations through scene graphs",
                "authors": [
                    {
                        "first": "Fei",
                        "middle": [],
                        "last": "Yu",
                        "suffix": ""
                    },
                    {
                        "first": "Jiji",
                        "middle": [],
                        "last": "Tang",
                        "suffix": ""
                    },
                    {
                        "first": "Weichong",
                        "middle": [],
                        "last": "Yin",
                        "suffix": ""
                    },
                    {
                        "first": "Yu",
                        "middle": [],
                        "last": "Sun",
                        "suffix": ""
                    },
                    {
                        "first": "Hua",
                        "middle": [],
                        "last": "Hao Tian",
                        "suffix": ""
                    },
                    {
                        "first": "Haifeng",
                        "middle": [],
                        "last": "Wu",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the AAAI Conference on Artificial Intelligence",
                "volume": "35",
                "issue": "",
                "pages": "3208--3216",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Fei Yu, Jiji Tang, Weichong Yin, Yu Sun, Hao Tian, Hua Wu, and Haifeng Wang. 2021. ERNIE- ViL: Knowledge enhanced vision-language repre- sentations through scene graphs. Proceedings of the AAAI Conference on Artificial Intelligence, 35(4):3208-3216.",
                "links": null
            },
            "BIBREF64": {
                "ref_id": "b64",
                "title": "Twitter hate aspect extraction using association analysis and dictionary-based approach",
                "authors": [
                    {
                        "first": "Nurulhuda",
                        "middle": [],
                        "last": "Zainuddin",
                        "suffix": ""
                    },
                    {
                        "first": "Ali",
                        "middle": [],
                        "last": "Selamat",
                        "suffix": ""
                    },
                    {
                        "first": "Roliana",
                        "middle": [],
                        "last": "Ibrahim",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "New Trends in Intelligent Software Methodologies, Tools and Techniques -Proceedings of the 16th International Conference (SoMeT'17)",
                "volume": "297",
                "issue": "",
                "pages": "641--651",
                "other_ids": {
                    "DOI": [
                        "10.3233/978-1-61499-800-6-641"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Nurulhuda Zainuddin, Ali Selamat, and Roliana Ibrahim. 2017. Twitter hate aspect extraction using association analysis and dictionary-based approach. In New Trends in Intelligent Software Methodologies, Tools and Techniques -Proceedings of the 16th In- ternational Conference (SoMeT'17), volume 297 of Frontiers in Artificial Intelligence and Applications, pages 641-651, Kitakyushu City, Japan.",
                "links": null
            },
            "BIBREF65": {
                "ref_id": "b65",
                "title": "Evaluating aspect-based sentiment classification on Twitter hate speech using neural networks and word embedding features",
                "authors": [
                    {
                        "first": "Nurulhuda",
                        "middle": [],
                        "last": "Zainuddin",
                        "suffix": ""
                    },
                    {
                        "first": "Ali",
                        "middle": [],
                        "last": "Selamat",
                        "suffix": ""
                    },
                    {
                        "first": "Roliana",
                        "middle": [],
                        "last": "Ibrahim",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "New Trends in Intelligent Software Methodologies, Tools and Techniques",
                "volume": "",
                "issue": "",
                "pages": "723--734",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Nurulhuda Zainuddin, Ali Selamat, and Roliana Ibrahim. 2018. Evaluating aspect-based sentiment classification on Twitter hate speech using neural networks and word embedding features. In New Trends in Intelligent Software Methodologies, Tools and Techniques, pages 723-734.",
                "links": null
            },
            "BIBREF66": {
                "ref_id": "b66",
                "title": "Hate crime on Twitter: Aspectbased sentiment analysis approach",
                "authors": [
                    {
                        "first": "Nurulhuda",
                        "middle": [],
                        "last": "Zainuddin",
                        "suffix": ""
                    },
                    {
                        "first": "Ali",
                        "middle": [],
                        "last": "Selamat",
                        "suffix": ""
                    },
                    {
                        "first": "Roliana",
                        "middle": [],
                        "last": "Ibrahim",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Advancing Technology Industrialization Through Intelligent Software Methodologies, Tools and Techniques",
                "volume": "",
                "issue": "",
                "pages": "284--297",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Nurulhuda Zainuddin, Ali Selamat, and Roliana Ibrahim. 2019. Hate crime on Twitter: Aspect- based sentiment analysis approach. In Advancing Technology Industrialization Through Intelligent Soft- ware Methodologies, Tools and Techniques, pages 284-297.",
                "links": null
            },
            "BIBREF67": {
                "ref_id": "b67",
                "title": "Predicting the type and target of offensive posts in social media",
                "authors": [
                    {
                        "first": "Marcos",
                        "middle": [],
                        "last": "Zampieri",
                        "suffix": ""
                    },
                    {
                        "first": "Shervin",
                        "middle": [],
                        "last": "Malmasi",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "Sara",
                        "middle": [],
                        "last": "Rosenthal",
                        "suffix": ""
                    },
                    {
                        "first": "Noura",
                        "middle": [],
                        "last": "Farra",
                        "suffix": ""
                    },
                    {
                        "first": "Ritesh",
                        "middle": [],
                        "last": "Kumar",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 17th Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT '19",
                "volume": "",
                "issue": "",
                "pages": "1415--1420",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Marcos Zampieri, Shervin Malmasi, Preslav Nakov, Sara Rosenthal, Noura Farra, and Ritesh Kumar. 2019a. Predicting the type and target of offensive posts in social media. In Proceedings of the 17th Annual Conference of the North American Chap- ter of the Association for Computational Linguistics: Human Language Technologies, NAACL-HLT '19, pages 1415-1420, Minneapolis, MN, USA.",
                "links": null
            },
            "BIBREF68": {
                "ref_id": "b68",
                "title": "SemEval-2019 task 6: Identifying and categorizing offensive language in social media (OffensEval)",
                "authors": [
                    {
                        "first": "Marcos",
                        "middle": [],
                        "last": "Zampieri",
                        "suffix": ""
                    },
                    {
                        "first": "Shervin",
                        "middle": [],
                        "last": "Malmasi",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "Sara",
                        "middle": [],
                        "last": "Rosenthal",
                        "suffix": ""
                    },
                    {
                        "first": "Noura",
                        "middle": [],
                        "last": "Farra",
                        "suffix": ""
                    },
                    {
                        "first": "Ritesh",
                        "middle": [],
                        "last": "Kumar",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the International Workshop on Semantic Evaluation, SemEval '19",
                "volume": "",
                "issue": "",
                "pages": "75--86",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Marcos Zampieri, Shervin Malmasi, Preslav Nakov, Sara Rosenthal, Noura Farra, and Ritesh Kumar. 2019b. SemEval-2019 task 6: Identifying and catego- rizing offensive language in social media (OffensE- val). In Proceedings of the International Workshop on Semantic Evaluation, SemEval '19, pages 75-86, Minneapolis, MN, USA.",
                "links": null
            },
            "BIBREF69": {
                "ref_id": "b69",
                "title": "Zeses Pitenis, and \u00c7agr\u0131 \u00c7\u00f6ltekin. 2020. SemEval-2020 task 12: Multilingual offensive language identification in social media",
                "authors": [
                    {
                        "first": "Marcos",
                        "middle": [],
                        "last": "Zampieri",
                        "suffix": ""
                    },
                    {
                        "first": "Preslav",
                        "middle": [],
                        "last": "Nakov",
                        "suffix": ""
                    },
                    {
                        "first": "Sara",
                        "middle": [],
                        "last": "Rosenthal",
                        "suffix": ""
                    },
                    {
                        "first": "Pepa",
                        "middle": [],
                        "last": "Atanasova",
                        "suffix": ""
                    },
                    {
                        "first": "Georgi",
                        "middle": [],
                        "last": "Karadzhov",
                        "suffix": ""
                    },
                    {
                        "first": "Hamdy",
                        "middle": [],
                        "last": "Mubarak",
                        "suffix": ""
                    },
                    {
                        "first": "Leon",
                        "middle": [],
                        "last": "Derczynski",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "1425--1447",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2020.semeval-1.188"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Marcos Zampieri, Preslav Nakov, Sara Rosenthal, Pepa Atanasova, Georgi Karadzhov, Hamdy Mubarak, Leon Derczynski, Zeses Pitenis, and \u00c7agr\u0131 \u00c7\u00f6ltekin. 2020. SemEval-2020 task 12: Multilingual offensive language identification in social media (OffensEval 2020). In Proceedings of the Fourteenth Workshop on Semantic Evaluation, pages 1425-1447.",
                "links": null
            },
            "BIBREF70": {
                "ref_id": "b70",
                "title": "On the origins of memes by means of fringe web communities",
                "authors": [
                    {
                        "first": "Savvas",
                        "middle": [],
                        "last": "Zannettou",
                        "suffix": ""
                    },
                    {
                        "first": "Tristan",
                        "middle": [],
                        "last": "Caulfield",
                        "suffix": ""
                    },
                    {
                        "first": "Jeremy",
                        "middle": [],
                        "last": "Blackburn",
                        "suffix": ""
                    },
                    {
                        "first": "Emiliano",
                        "middle": [],
                        "last": "De Cristofaro",
                        "suffix": ""
                    },
                    {
                        "first": "Michael",
                        "middle": [],
                        "last": "Sirivianos",
                        "suffix": ""
                    },
                    {
                        "first": "Gianluca",
                        "middle": [],
                        "last": "Stringhini",
                        "suffix": ""
                    },
                    {
                        "first": "Guillermo",
                        "middle": [],
                        "last": "Suarez-Tangil",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the Internet Measurement Conference 2018, IMC '18",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "DOI": [
                        "10.1145/3278532.3278550"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Savvas Zannettou, Tristan Caulfield, Jeremy Blackburn, Emiliano De Cristofaro, Michael Sirivianos, Gian- luca Stringhini, and Guillermo Suarez-Tangil. 2018. On the origins of memes by means of fringe web communities. In Proceedings of the Internet Mea- surement Conference 2018, IMC '18, page 188202, New York, NY, USA.",
                "links": null
            },
            "BIBREF71": {
                "ref_id": "b71",
                "title": "Grunn2019 at SemEval-2019 task 5: Shared task on multilingual detection of hate",
                "authors": [
                    {
                        "first": "Mike",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Roy",
                        "middle": [],
                        "last": "David",
                        "suffix": ""
                    },
                    {
                        "first": "Leon",
                        "middle": [],
                        "last": "Graumans",
                        "suffix": ""
                    },
                    {
                        "first": "Gerben",
                        "middle": [],
                        "last": "Timmerman",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 13th International Workshop on Semantic Evaluation",
                "volume": "",
                "issue": "",
                "pages": "391--395",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/S19-2069"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Mike Zhang, Roy David, Leon Graumans, and Gerben Timmerman. 2019. Grunn2019 at SemEval-2019 task 5: Shared task on multilingual detection of hate. In Proceedings of the 13th International Workshop on Semantic Evaluation, pages 391-395, Minneapolis, Minnesota, USA.",
                "links": null
            },
            "BIBREF72": {
                "ref_id": "b72",
                "title": "Multimodal learning for hateful memes detection",
                "authors": [
                    {
                        "first": "Yi",
                        "middle": [],
                        "last": "Zhou",
                        "suffix": ""
                    },
                    {
                        "first": "Zhenhao",
                        "middle": [],
                        "last": "Chen",
                        "suffix": ""
                    },
                    {
                        "first": "Huiyuan",
                        "middle": [],
                        "last": "Yang",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Proceedings of the International Conference on Multimedia Expo Workshops, ICMEW '21",
                "volume": "",
                "issue": "",
                "pages": "1--6",
                "other_ids": {
                    "DOI": [
                        "10.1109/ICMEW53276.2021.9455994"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Yi Zhou, Zhenhao Chen, and Huiyuan Yang. 2021. Mul- timodal learning for hateful memes detection. In Proceedings of the International Conference on Mul- timedia Expo Workshops, ICMEW '21, pages 1-6.",
                "links": null
            },
            "BIBREF73": {
                "ref_id": "b73",
                "title": "DD-TIG at Constraint@ACL2022: Multimodal understanding and reasoning for role labeling of entities in hateful memes",
                "authors": [
                    {
                        "first": "Ziming",
                        "middle": [],
                        "last": "Zhou",
                        "suffix": ""
                    },
                    {
                        "first": "Han",
                        "middle": [],
                        "last": "Zhao",
                        "suffix": ""
                    },
                    {
                        "first": "Jingjing",
                        "middle": [],
                        "last": "Dong",
                        "suffix": ""
                    },
                    {
                        "first": "Jun",
                        "middle": [],
                        "last": "Gao",
                        "suffix": ""
                    },
                    {
                        "first": "Xiaolong",
                        "middle": [],
                        "last": "Liu",
                        "suffix": ""
                    }
                ],
                "year": 2022,
                "venue": "Proceedings of the Workshop on Combating Online Hostile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Ziming Zhou, Han Zhao, Jingjing Dong, Jun Gao, and Xiaolong Liu. 2022. DD-TIG at Con- straint@ACL2022: Multimodal understanding and reasoning for role labeling of entities in hateful memes. In Proceedings of the Workshop on Com- bating Online Hostile Posts in Regional Languages during Emergency Situations, CONSTRAINT '22, Dublin, Ireland.",
                "links": null
            },
            "BIBREF74": {
                "ref_id": "b74",
                "title": "Memes and the spread of misinformation: Establishing the importance of media literacy in the era of information disorder",
                "authors": [
                    {
                        "first": "Sulafa",
                        "middle": [],
                        "last": "Zidani",
                        "suffix": ""
                    },
                    {
                        "first": "Rachel",
                        "middle": [],
                        "last": "Moran",
                        "suffix": ""
                    }
                ],
                "year": 2021,
                "venue": "Teaching Media Quarterly",
                "volume": "9",
                "issue": "1",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Sulafa Zidani and Rachel Moran. 2021. Memes and the spread of misinformation: Establishing the im- portance of media literacy in the era of information disorder. Teaching Media Quarterly, 9(1).",
                "links": null
            }
        },
        "ref_entries": {
            "FIGREF0": {
                "uris": null,
                "text": "Examples of heroes, villains and victims, as portrayed within memes.",
                "type_str": "figure",
                "num": null
            },
            "FIGREF1": {
                "uris": null,
                "text": "Word clouds for (a) COVID-19 and (b) US Politics domains in HVVMemes.",
                "type_str": "figure",
                "num": null
            },
            "FIGREF2": {
                "uris": null,
                "text": "",
                "type_str": "figure",
                "num": null
            },
            "TABREF0": {
                "text": "modeled sarcastic targeting of specific entities. Rosenthal et al. (2017) focused on sentiment analysis in Twitter.",
                "html": null,
                "content": "<table><tr><td>Domain Splits # Memes</td><td># Referenced Entities Hero Villain Victim Other Total</td></tr></table>",
                "type_str": "table",
                "num": null
            },
            "TABREF1": {
                "text": "Statistics about our HVVMemes dataset.",
                "html": null,
                "content": "<table/>",
                "type_str": "table",
                "num": null
            },
            "TABREF2": {
                "text": "Key considerations in our annotation guidelines.",
                "html": null,
                "content": "<table><tr><td>Entity</td><td>Resolution Remark</td></tr><tr><td colspan=\"2\">Corona resolved to Corona Beer (whenever valid).</td></tr><tr><td>Govt.</td><td>resolved to Government.</td></tr><tr><td>Putin</td><td>resolved to Vladimir Putin.</td></tr><tr><td>CDC</td><td>standardised as Centre of Disease Control (CDC).</td></tr></table>",
                "type_str": "table",
                "num": null
            },
            "TABREF3": {
                "text": "Examples of resolution remarks that we provided to the annotators towards entity identification.",
                "html": null,
                "content": "<table><tr><td>(a) COVID-19</td><td>(b) US Politics</td></tr></table>",
                "type_str": "table",
                "num": null
            },
            "TABREF5": {
                "text": "",
                "html": null,
                "content": "<table><tr><td>: Inter-annotator agreement (IAA) summary for completed (Stage-2) and consolidated (Stage-3) stages of the annotation process. Note that the average IAA for the dry-run (Stage-1), for COVID-19 and US Politics combined, was 0.50 (hero), 0.35 (villain), 0.14 (victim), and 0.55 (other).</td></tr></table>",
                "type_str": "table",
                "num": null
            }
        }
    }
}