| { | |
| "paper_id": "2020", | |
| "header": { | |
| "generated_with": "S2ORC 1.0.0", | |
| "date_generated": "2023-01-19T01:13:03.248073Z" | |
| }, | |
| "title": "Automated Extraction of Socio-political Events from News (AESPEN): Workshop and Shared Task Report", | |
| "authors": [ | |
| { | |
| "first": "Ali", | |
| "middle": [], | |
| "last": "H\u00fcrriyetoglu", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Vanni", | |
| "middle": [], | |
| "last": "Zavarella", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Hristo", | |
| "middle": [], | |
| "last": "Tanev", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Erdem", | |
| "middle": [], | |
| "last": "Y\u00f6r\u00fck", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Ali", | |
| "middle": [], | |
| "last": "Safaya", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Osman", | |
| "middle": [], | |
| "last": "Mutlu", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Fredrik", | |
| "middle": [], | |
| "last": "Olsson", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Magnus", | |
| "middle": [], | |
| "last": "Sahlgren", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Ben", | |
| "middle": [], | |
| "last": "Fehmi", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Ariel", | |
| "middle": [], | |
| "last": "Abdesslem", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Kristine", | |
| "middle": [], | |
| "last": "Ekgren", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "", | |
| "middle": [], | |
| "last": "Eck", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Matina", | |
| "middle": [], | |
| "last": "Halkia", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Stefano", | |
| "middle": [], | |
| "last": "Ferri", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Michail", | |
| "middle": [], | |
| "last": "Papazoglou", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Marie-Sophie", | |
| "middle": [], | |
| "last": "Van Damme", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Javier", | |
| "middle": [], | |
| "last": "Osorio", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Alejandro", | |
| "middle": [], | |
| "last": "Reyes", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Alejandro", | |
| "middle": [], | |
| "last": "Beltr\u00e1n", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Clionadh", | |
| "middle": [], | |
| "last": "Raleigh", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Philip", | |
| "middle": [ | |
| "A" | |
| ], | |
| "last": "Schrodt", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Berfu", | |
| "middle": [], | |
| "last": "B\u00fcy\u00fck\u00f6z", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| }, | |
| { | |
| "first": "Arzucan", | |
| "middle": [], | |
| "last": "\u00d6zg\u00fcr", | |
| "suffix": "", | |
| "affiliation": { | |
| "laboratory": "Analyzing ELMo and DistilBERT on Socio-political News Classification", | |
| "institution": "", | |
| "location": {} | |
| }, | |
| "email": "" | |
| } | |
| ], | |
| "year": "", | |
| "venue": null, | |
| "identifiers": {}, | |
| "abstract": "", | |
| "pdf_parse": { | |
| "paper_id": "2020", | |
| "_pdf_hash": "", | |
| "abstract": [], | |
| "body_text": [ | |
| { | |
| "text": "This year we've accepted for publication nine papers which look at event detection from different points of view.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "Nine papers were regular paper submissions and one was a shared task participation report. The shared task report and seven of the regular papers were accepted on the basis of reviews, which were five per paper, performed by the program committee members.", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The accepted regular papers can be grouped as i) evaluation of state-of-the-art machine learning approaches by Buyukoz et al., Olsson et al., and Piskorski, and Jacquet, ii) introduction of a new data set by Radford, iii) projects of event information collection by Osorio et al. and Papanikolaou and Papageorgiou, and iv) forecasting of political conflict by Halkia et al. Buyukoz et al. and Olsson et al. show that state-of-the-art deep learning models such as BERT and ELMo (Peters et al., 2018) yield consistently higher performance than traditional ML methods such as support vector machines (SVM) on conflict and protest event data respectively. Piskorski and Jacquet have found that TF-IDFweighted character n-gram based SVM model performs better than an SVM model that facilitates pre-trained em-beddings such as GLOVE (Pennington et al., 2014) , BERT, and FASTTEXT (Mikolov et al., 2018) in most of the experiments on conflict data. Radford introduces the dataset Headlines of War for cross-document coreference resolution for the news headlines.The dataset consists of positive samples from Militarized Interstate Disputes dataset and negative samples from NewYork Times. The description of this invaluable resource accompanied with a detailed discussion of its utility and caveats. Osorio et al. introduce Hadath that is a supervised protocol for event information collection from Arabic sources. The utility of Hadath was demonstrated in processing news reported between 2012 and 2012 in Afghanistan. In the scope of the other event information collection study, Papanikolaou and Papageorgiou processed two news sourcesi in Greek from Greece to create a database of protest events for the period between 1996 and 2014. Osorio et. al. and Pa-panikolaou and Papageorgiou utilized fully automatic toolsthat integrates supervised machine learning and rule based methodologies at various degrees. Finally, Halkia et al. presents a material conflict forecastingstudy that facilitates the available event databases GDELT and ICEWS. Their results demonstrate that it is possible to correctly predict social upheaval using the methodology they propose, which utilizes Long-Short Term Memory(LSTM) (Hochreiter and Schmidhuber, 1997).We have received expression of interest from 12 research teams, of which 6 teams signed the application form and received the data. Two of these teams sent their predictions on the test data. Finally, only Ors et al. submitted a paper about their work. This team reported their work as consisting of three steps. First, they use a transformer based model,which is ALBERT (Lan et al., 2020), to predict whether apair of sentences refer to the same event or not. Later, they use these predictions as the initial scores and recalculate their scores by considering the relation of sentences in a pair with respect to other sentences. As the last step, final scores between these sentences are used to construct the clusters,starting with the pairs with the highest scores.", | |
| "cite_spans": [ | |
| { | |
| "start": 111, | |
| "end": 173, | |
| "text": "Buyukoz et al., Olsson et al., and Piskorski, and Jacquet, ii)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 374, | |
| "end": 406, | |
| "text": "Buyukoz et al. and Olsson et al.", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 463, | |
| "end": 498, | |
| "text": "BERT and ELMo (Peters et al., 2018)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 821, | |
| "end": 852, | |
| "text": "GLOVE (Pennington et al., 2014)", | |
| "ref_id": null | |
| }, | |
| { | |
| "start": 855, | |
| "end": 896, | |
| "text": "BERT, and FASTTEXT (Mikolov et al., 2018)", | |
| "ref_id": null | |
| } | |
| ], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "Introduction", | |
| "sec_num": null | |
| }, | |
| { | |
| "text": "The variety of the submitted papers show that we could bring the ML, NLP, and social and political science communities together. Although the breadth of the topics were limited, the technical depth and timeliness of the contributions show that the workshop contribute to the discipline of automatic extraction of socio-political events. The papers about processing Arabic and Greek sources are significant contributions to the understanding of how should we handle languages other than English. Finally, the shared task ESCI demonstrated the prevalence of the event coreferences, some baselines for handling them, and a state-of-the-art system that is able to tackle this task. ", | |
| "cite_spans": [], | |
| "ref_spans": [], | |
| "eq_spans": [], | |
| "section": "The evaluation of", | |
| "sec_num": null | |
| } | |
| ], | |
| "back_matter": [], | |
| "bib_entries": {}, | |
| "ref_entries": { | |
| "TABREF1": { | |
| "num": null, | |
| "text": "", | |
| "type_str": "table", | |
| "html": null, | |
| "content": "<table/>" | |
| } | |
| } | |
| } | |
| } |