Base_JSON/prefixE/json/ecnlp/2021.ecnlp-1.6.json

{
    "paper_id": "2021",
    "header": {
        "generated_with": "S2ORC 1.0.0",
        "date_generated": "2023-01-19T10:33:52.427152Z"
    },
    "title": "Personalized Entity Resolution with Dynamic Heterogeneous Knowledge Graph Representations",
    "authors": [
        {
            "first": "Ying",
            "middle": [],
            "last": "Lin",
            "suffix": "",
            "affiliation": {},
            "email": "[email protected]"
        },
        {
            "first": "Han",
            "middle": [],
            "last": "Wang",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Jiangning",
            "middle": [],
            "last": "Chen",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Tong",
            "middle": [],
            "last": "Wang",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Yue",
            "middle": [],
            "last": "Liu",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Heng",
            "middle": [],
            "last": "Ji",
            "suffix": "",
            "affiliation": {},
            "email": ""
        },
        {
            "first": "Yang",
            "middle": [],
            "last": "Liu",
            "suffix": "",
            "affiliation": {},
            "email": "[email protected]"
        },
        {
            "first": "Premkumar",
            "middle": [],
            "last": "Natarajan",
            "suffix": "",
            "affiliation": {},
            "email": ""
        }
    ],
    "year": "",
    "venue": null,
    "identifiers": {},
    "abstract": "The growing popularity of Virtual Assistants poses new challenges for Entity Resolution, the task of linking mentions in text to their referent entities in a knowledge base. Specifically, in the shopping domain, customers tend to mention the entities implicitly (e.g., \"organic milk\") rather than use the entity names explicitly, leading to a large number of candidate products. Meanwhile, for the same query, different customers may expect different results. For example, with \"add milk to my cart\", a customer may refer to a certain product from his/her favorite brand, while some customers may want to reorder products they regularly purchase. Moreover, new customers may lack persistent shopping history, which requires us to enrich the connections between customers through products and their attributes. To address these issues, we propose a new framework that leverages personalized features to improve the accuracy of product ranking. We first build a cross-source heterogeneous knowledge graph from customer purchase history and product knowledge graph to jointly learn customer and product embeddings. After that, we incorporate product, customer, and history representations into a neural reranking model to predict which candidate is most likely to be purchased by a specific customer. Experiment results show that our model substantially improves the accuracy of the top ranked candidates by 24.6% compared to the state-of-theart product search model. * * This work was done when the first author was on an internship at Amazon Alexa AI.",
    "pdf_parse": {
        "paper_id": "2021",
        "_pdf_hash": "",
        "abstract": [
            {
                "text": "The growing popularity of Virtual Assistants poses new challenges for Entity Resolution, the task of linking mentions in text to their referent entities in a knowledge base. Specifically, in the shopping domain, customers tend to mention the entities implicitly (e.g., \"organic milk\") rather than use the entity names explicitly, leading to a large number of candidate products. Meanwhile, for the same query, different customers may expect different results. For example, with \"add milk to my cart\", a customer may refer to a certain product from his/her favorite brand, while some customers may want to reorder products they regularly purchase. Moreover, new customers may lack persistent shopping history, which requires us to enrich the connections between customers through products and their attributes. To address these issues, we propose a new framework that leverages personalized features to improve the accuracy of product ranking. We first build a cross-source heterogeneous knowledge graph from customer purchase history and product knowledge graph to jointly learn customer and product embeddings. After that, we incorporate product, customer, and history representations into a neural reranking model to predict which candidate is most likely to be purchased by a specific customer. Experiment results show that our model substantially improves the accuracy of the top ranked candidates by 24.6% compared to the state-of-theart product search model. * * This work was done when the first author was on an internship at Amazon Alexa AI.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Abstract",
                "sec_num": null
            }
        ],
        "body_text": [
            {
                "text": "Given an entity mention as a query, the goal of entity resolution (or entity linking) (Ji and Grishman, 2011) is to link the mention to its corresponding entry in a target knowledge base (KB). In an academic shared task setting, an entity mention is usually a name string, which can be a person, organization or geo-political entity in a news context, and the KB is usually a Wikipedia dump with rich structured properties and unstructured text descriptions. State-of-the-art entity resolution methods can achieve higher than 90% accuracy in such settings (Ji and Grishman, 2011; Agarwal and Bikel, 2020) , and they have been successfully applied in hundreds of languages (Pan et al., 2017) and various domains such as disaster management (Zhang et al., 2018a) and scientific discovery (Zheng et al., 2014; . Therefore, we tend to think entity resolution is a solved problem in academia. However, in industry, with the rise in popularity of Virtual Assistants (VAs) in recent years, an increasing number of customers now rely on VAs to perform daily tasks involving entities, including shopping, playing music or movies, calling a person, booking a flight, and managing schedules. The scale and complexity of industrial applications presents the following unique new challenges.",
                "cite_spans": [
                    {
                        "start": 86,
                        "end": 109,
                        "text": "(Ji and Grishman, 2011)",
                        "ref_id": "BIBREF12"
                    },
                    {
                        "start": 556,
                        "end": 579,
                        "text": "(Ji and Grishman, 2011;",
                        "ref_id": "BIBREF12"
                    },
                    {
                        "start": 580,
                        "end": 604,
                        "text": "Agarwal and Bikel, 2020)",
                        "ref_id": "BIBREF1"
                    },
                    {
                        "start": 672,
                        "end": 690,
                        "text": "(Pan et al., 2017)",
                        "ref_id": "BIBREF26"
                    },
                    {
                        "start": 739,
                        "end": 760,
                        "text": "(Zhang et al., 2018a)",
                        "ref_id": "BIBREF52"
                    },
                    {
                        "start": 786,
                        "end": 806,
                        "text": "(Zheng et al., 2014;",
                        "ref_id": "BIBREF56"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "Unpopular majority. There is a massive number of new entities emerging every day. The entity resolver may know very little about them since very few users interact with them. Handling these tail entities effectively requires the use of property linkages between entities and shared user interests. Similarly, there might be many new users with limited interaction history, and we need to infer their interests from other users who have interacted with similar entities.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "Large number of ambiguous variants. When interacting with VAs, users tend to use short and less informative utterances with the expectation that the VAs can intelligently infer their actual intentions. This raises the need for personalization when resolving the entities. In the shopping domain, this problem is even more challenging as customers typically use implicit entity mentions (c) Cross-source Heterogeneous Customer-product Graph (e.g., \"organic milk\") instead of explicit names (e.g., \"Horizon Organic Shelf-Stable 1% Lowfat Milk\") which usually leads to a large number of candidates due to the ambiguity. However, with VAs' voice user interface (VUI), the number of products that can be presented to the customers is very limited, if not only one. In this work, we focus on the problem of personalized entity resolution in the shopping domain. Given a query and a list of retrieved candidates, we aim to return the product that is most likely to be purchased by a customer.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "Beyond ambiguity. In the traditional news entity linking setting, each entity in the KB refers to a unique world object. In contrast, in e-commerce, the same product can have multiple variants. For example, a customer may like to stick to a toothpaste product of a certain brand and flavor, but choose different sizes (thus different entities) in each purchase. These entities in the target KB refer to the same product but have different properties (in this case, size). Therefore it is important to construct fine-grained knowledge graphs to profile products and capture the implicit connections between customers based on the properties of their purchased products.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "We make three assumptions: (H1) customers tend to purchase products they have purchased in the past; (H2) customers tend to purchase products that share some properties; (H3) customers who purchased products with similar properties share similar interests. Based on these assumptions, we propose to represent customers and products as low-dimensional vectors learned from a graph of customers and products. Unlike social media networks with rich interactions among users, the customers of most shopping services are isolated, which prevents us from learning customer embeddings effectively. To address this issue, we propose to build a cross-source heterogeneous knowledge graph as Figure 1 depicts to indirectly establish rich connections among customers from a) users' purchase history (customer-product graph) and b) a product knowledge graph, and further jointly learn the representations of nodes in this graph using a Graph Neural Network (GNN)-based method. In Figure 1 (c), for instance, we can build connections between Customer 1 and Customer 2 because their purchased products share the same ingredient attribute, and thus possibly recommend Product 2 to Customer 1 even though it does not appear in his/her purchase history. In addition to static customer embeddings, we further propose an attentive model to dynamically generate a history representation for each user based on the current query. Finally, the model predicts how likely a candidate will be purchased using entity, customer, and history representations.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 682,
                        "end": 690,
                        "text": "Figure 1",
                        "ref_id": "FIGREF1"
                    },
                    {
                        "start": 968,
                        "end": 976,
                        "text": "Figure 1",
                        "ref_id": "FIGREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "Experiments on real purchase records collected from an online shopping service show that our method significantly improves the purchase rate of the top ranked products.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Introduction",
                "sec_num": "1"
            },
            {
                "text": "Given a query q from a customer c, and a list of candidate products P = {p 1 , ..., p L }, where L is the number of candidates, our goal is to predict the product that the customer will purchase based on their purchase history and the product knowledge graph. Specifically, we use purchase records {r 1 , ..., r H } where H is the number of historical records. As Figure 2 illustrates, we jointly learn customer and product embeddings from a cross-source customer-product graph using GNN. To perform personalized ranking, we incorporate the learned customer embedding and history representation as additional features when calculating the confidence score of each candidate. We then rank all candidates by confidence score and return the top one. ",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 364,
                        "end": 372,
                        "text": "Figure 2",
                        "ref_id": "FIGREF2"
                    }
                ],
                "eq_spans": [],
                "section": "Methodology",
                "sec_num": "2"
            },
            {
                "text": "We first retrieve candidate products for each query using QUARTS (Nigam et al., 2019; Nguyen et al., 2020) , which is an end-to-end neural model for product search. QUARTS has three major components: (1) an LSTM-based (long short-term memory) classifier adapted from the entailment model in (Rockt\u00e4schel et al., 2016) to predict whether a product-query pair is matched; (2) a variational encoder-decoder (VED) query generator that generates difficult negative examples to tackle the class imbalance issue in the training data as a search engine typically returns much more matched queryitem examples than mismatched ones, and (3) a state combiner that switches between query representations computed by the classifier and generator. During training, the VED generator takes as input a matched product-query pair (I, Q) and generates a mismatched query Q gen which is lexically similar to Q. The state combiner then merges H, the representation computed by the classifier, and H gen , the representation computed by the generator, as sH gen + (1 \u2212 s)H, where s is a binary value that controls which query to use and whether the gradients are back-propagated to the classifier or generator.",
                "cite_spans": [
                    {
                        "start": 65,
                        "end": 85,
                        "text": "(Nigam et al., 2019;",
                        "ref_id": "BIBREF24"
                    },
                    {
                        "start": 86,
                        "end": 106,
                        "text": "Nguyen et al., 2020)",
                        "ref_id": "BIBREF22"
                    },
                    {
                        "start": 291,
                        "end": 317,
                        "text": "(Rockt\u00e4schel et al., 2016)",
                        "ref_id": "BIBREF33"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Retrieval",
                "sec_num": "2.1"
            },
            {
                "text": "The next step is to obtain the representations of customers and products. Customer embeddings are usually learned from user-generated texts (Preo\u0163iuc-Pietro et al., 2015; Yu et al., 2016; Ribeiro et al., 2018) or social relations (Perozzi et al., 2014a; Grover and Leskovec, 2016; Zhang et al., 2018b) , neither of which are available in the shopping dataset we use. Alternatively, we establish indirect connections among customers through their purchased products under hypothesis H3, and form a customer-product graph as shown in Figure 1 (a). This graph only contains a single type of relation (i.e., purchase) and ignores product attributes. As a result, it tends to be sparse and less effective for customer representation learning.",
                "cite_spans": [
                    {
                        "start": 140,
                        "end": 170,
                        "text": "(Preo\u0163iuc-Pietro et al., 2015;",
                        "ref_id": "BIBREF30"
                    },
                    {
                        "start": 171,
                        "end": 187,
                        "text": "Yu et al., 2016;",
                        "ref_id": "BIBREF51"
                    },
                    {
                        "start": 188,
                        "end": 209,
                        "text": "Ribeiro et al., 2018)",
                        "ref_id": "BIBREF32"
                    },
                    {
                        "start": 230,
                        "end": 253,
                        "text": "(Perozzi et al., 2014a;",
                        "ref_id": "BIBREF28"
                    },
                    {
                        "start": 254,
                        "end": 280,
                        "text": "Grover and Leskovec, 2016;",
                        "ref_id": "BIBREF7"
                    },
                    {
                        "start": 281,
                        "end": 301,
                        "text": "Zhang et al., 2018b)",
                        "ref_id": "BIBREF54"
                    }
                ],
                "ref_spans": [
                    {
                        "start": 532,
                        "end": 540,
                        "text": "Figure 1",
                        "ref_id": "FIGREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Joint Customer and Product Embedding",
                "sec_num": "2.2"
            },
            {
                "text": "In order to learn more informative embeddings, we propose to incorporate richer information from a product knowledge graph (Figure 1 (b)) where products are not only connected to different attribute nodes (e.g., brands, flavors), but they may also be associated with textual features (e.g., title) and boolean features (e.g., isOrganic, encoded as a boolean vector).",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 123,
                        "end": 132,
                        "text": "(Figure 1",
                        "ref_id": "FIGREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Joint Customer and Product Embedding",
                "sec_num": "2.2"
            },
            {
                "text": "By merging the product knowledge graph and the customer-product graph, we obtain a more comprehensive graph (Figure 1 (c)) of higher connectivity. For example, in the original customer-product graph, Customer 1 and Customer 2 are disconnected because they do not share any purchase. In the new graph, they have an indirect connection through Product 2 and Product 3, which share the same flavor and ingredient.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 108,
                        "end": 117,
                        "text": "(Figure 1",
                        "ref_id": "FIGREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Joint Customer and Product Embedding",
                "sec_num": "2.2"
            },
            {
                "text": "From this heterogeneous graph, we jointly learn customer and product representations using a two-layer Relational Graph Convolutional Network (Schlichtkrull et al., 2018) . The embedding of each node is updated as:",
                "cite_spans": [
                    {
                        "start": 142,
                        "end": 170,
                        "text": "(Schlichtkrull et al., 2018)",
                        "ref_id": "BIBREF34"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Joint Customer and Product Embedding",
                "sec_num": "2.2"
            },
            {
                "text": "h l+1 i = ReLU W l 0 h l i + r\u2208R j\u2208N r i 1 |N r i | W l r h l j ,",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Joint Customer and Product Embedding",
                "sec_num": "2.2"
            },
            {
                "text": "where h l i is the representation of node i at the l-th layer, N r i is the set of neighbor indices of node i under relation r \u2208 R, and W l 0 and W l r are learnable weight matrices.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Joint Customer and Product Embedding",
                "sec_num": "2.2"
            },
            {
                "text": "In order to capture textual features (i.e., product titles, descriptions, and bullet 1 ), we use a pretrained RoBERTa (Liu et al., 2019) encoder to generate a fix-sized representation for each product.",
                "cite_spans": [
                    {
                        "start": 118,
                        "end": 136,
                        "text": "(Liu et al., 2019)",
                        "ref_id": "BIBREF18"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Joint Customer and Product Embedding",
                "sec_num": "2.2"
            },
            {
                "text": "Specifically, we concatenate textual features using a special separator token [SEP] , obtain the RoBERTa representation for each token, and then use the averaged embedding to represent the whole sequence. To reduce the runtime, we calculate customer and product embeddings offline and cache the results.",
                "cite_spans": [
                    {
                        "start": 78,
                        "end": 83,
                        "text": "[SEP]",
                        "ref_id": null
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Joint Customer and Product Embedding",
                "sec_num": "2.2"
            },
            {
                "text": "In addition to the product embedding, we further incorporate the following features to enrich the representation of each candidate.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Representation",
                "sec_num": "2.3"
            },
            {
                "text": "Rank: the order of the candidate returned by the product retrieval system.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Representation",
                "sec_num": "2.3"
            },
            {
                "text": "Relative Price: how much a product's absolute price is higher or lower than the average price of all retrieved candidates as price is an important factor that affects purchasing decision.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Representation",
                "sec_num": "2.3"
            },
            {
                "text": "Previously Purchased: a binary flag indicating whether a candidate has been purchased by the customer or not.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Representation",
                "sec_num": "2.3"
            },
            {
                "text": "Textual Feature: we use RoBERTa to encode each candidate's textual features (i.e., title, bullet, description). This RoBERTa encoder is fine-tuned during training.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Representation",
                "sec_num": "2.3"
            },
            {
                "text": "We concatenate these features with the product embedding and project the vector into a lower dimensional space using a feed forward network.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Representation",
                "sec_num": "2.3"
            },
            {
                "text": "Although customer embeddings can encode purchase history information, they are static and may not effectively provide the most relevant information for each specific query. For example, if the query is \"bookshelf\", the furniture-related purchase records are more likely to help the model predict the product that the customer will purchase, while if the query is \"sulfate-free shampoo\", the purchase records of beauty products are more relevant. To tackle this issue, we propose to generate a queryaware history representation v based on the current query q from all purchase record representations {v 1 , ..., v H } of the customer.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "History Representation",
                "sec_num": "2.4"
            },
            {
                "text": "We first represent each purchase record as the concatenation of the product embedding, product price, and purchase timestamp. The queryaware history representation is then calculated as a weighted sum of the customer's purchase record representations using an attention mechanism as follows.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "History Representation",
                "sec_num": "2.4"
            },
            {
                "text": "e i = v tanh W q q + W v v i , a i = Softmax(e i ) = exp (e i ) M k exp (e k ) , v = H i a i v i ,",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "History Representation",
                "sec_num": "2.4"
            },
            {
                "text": "where v , W q , and W v are learnable weights.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "History Representation",
                "sec_num": "2.4"
            },
            {
                "text": "We adopt a feed forward neural network that takes in the candidate, customer, and history representations, and returns a confidence score\u0177 i which indicates how likely a candidate will be purchased. The confidence score is scaled to (0, 1) using a Sigmoid function. During training, we optimize the model by minimizing the following binary cross entropy loss function.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Ranking",
                "sec_num": "2.5"
            },
            {
                "text": "L = \u2212 1 N N i=1 y i log\u0177 i + (1 \u2212 y i ) log (1 \u2212\u0177 i ),",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Ranking",
                "sec_num": "2.5"
            },
            {
                "text": "where N denotes the total number of candidates, and y i \u2208 {0, 1} is the true label. In the inference phase, we calculate confidence scores for all candidates for each session and return the one with the highest score.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Ranking",
                "sec_num": "2.5"
            },
            {
                "text": "3 Experiment",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Candidate Ranking",
                "sec_num": "2.5"
            },
            {
                "text": "Product Knowledge Graph. In our experiment, we use a knowledge graph of products in five categories (i.e., grocery, beauty, luxury beauty, baby, and health care), which contains 24,287,337 unique product entities. As Figure 1(b) depicts, the products in this knowledge graph are connected through attribute nodes, including brands, scents, flavors, and ingredients. This knowledge graph also provides rich attributes for each product node. We use two types of attributes in this work, textual features (i.e., title, description, and bullet) and binary features (e.g., isOrganic, isNatural). Evalution Dataset. We randomly collect 1 million users' purchase sessions from November 2018 to October 2019 on an online shopping service. Each session contains a query, an obfuscated identifier, a timestamp, and a list of candidate products retrieved using QUARTS where only one product is purchased.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 217,
                        "end": 228,
                        "text": "Figure 1(b)",
                        "ref_id": "FIGREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Data",
                "sec_num": "3.1"
            },
            {
                "text": "We split the sessions before and after September 1, 2019 into two subsets. The first subset only serves as the purchase history and is used to construct the customer-product graph. From the second subset, we randomly sample 22,000 customers with at least one purchase record in the first subset and take their last purchase sessions for training or evaluation. Specifically, we use 20,000 sessions for training, 1,000 for validation, and 1,000 for test. If a customer has multiple purchase sessions in the second subset, other sessions before the last one are also considered as purchase history when we generate history representations, while they are excluded from the customer-product graph, which is constructed from the first subset.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Data",
                "sec_num": "3.1"
            },
            {
                "text": "We optimize our model with AdamW (Loshchilov and Hutter, 2018) for 10 epochs with a learning rate of 1e-5 for the RoBERTa encoder, a learning rate of 1e-4 for other parameters, weight decay of 1e-3, a warmup rate of 10%, and a batch size of 100.",
                "cite_spans": [
                    {
                        "start": 33,
                        "end": 62,
                        "text": "(Loshchilov and Hutter, 2018)",
                        "ref_id": "BIBREF21"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Experimental Setup",
                "sec_num": "3.2"
            },
            {
                "text": "To encode textual features, we use the RoBERTa base model 2 with an output dropout rate of 0.5. To represent query words, we use 100-dimensional GloVe embeddings (Pennington et al., 2014) pretrained on Wikipedia and Gigaword 3 . We set the size of pre-trained customer and product embeddings to 100 and freeze them during training.",
                "cite_spans": [
                    {
                        "start": 162,
                        "end": 187,
                        "text": "(Pennington et al., 2014)",
                        "ref_id": "BIBREF27"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Experimental Setup",
                "sec_num": "3.2"
            },
            {
                "text": "We use separate fully connected layers to project candidate and history representations into 100dimensional feature vectors before concatenating them for ranking. We use a two-layer feed forward neural network with a hidden layer size of 50 as the ranker and apply a dropout layer with a dropout rate of 0.5 to its input.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Experimental Setup",
                "sec_num": "3.2"
            },
            {
                "text": "We compare our model to the state-of-the-art product search model QUARTS as the baseline. Because our target usage scenarios are VAs where only one result will be returned to the user, we use accuracy as our evaluation metric. We implement the following baseline ranking methods. Purchased: We prioritize products previously purchased by the customer. If multiple candidates are previously purchased, we return the one ranked highest by QUARTS. ComplEx: Customer and product embeddings are learned using ComplEx (Trouillon et al., 2016) , a widely used knowledge embedding model that represents nodes in a knowledge graph as complex vectors and is able to capture antisymmetric relations using efficient dot product.",
                "cite_spans": [
                    {
                        "start": 512,
                        "end": 536,
                        "text": "(Trouillon et al., 2016)",
                        "ref_id": "BIBREF41"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Quantitative Analysis",
                "sec_num": "3.3"
            },
            {
                "text": "In Table 1 , we show the relative gains compared to the baseline model QUARTS. With personalized features, our method effectively improves accuracy on both development and test sets.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 3,
                        "end": 10,
                        "text": "Table 1",
                        "ref_id": "TABREF1"
                    }
                ],
                "eq_spans": [],
                "section": "Quantitative Analysis",
                "sec_num": "3.3"
            },
            {
                "text": "We also conduct ablation studies by removing the following features and show results in Table 2 . Ranking: In this setting, our model ignores the original retrieval ranking returned by QUARTS. Personalized Features: We remove personalized features (e.g., customer embedding, whether a product is previously purchased) in this setting. Product Embedding: We remove pre-trained product embedding but still use textual features and binary features to represent products. Joint Embedding: Customer and product embeddings are not jointly learned from the merged graph. Alternatively, customer embeddings are learned from the customer-product graph, and product embeddings are learned from the product knowledge graph.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 88,
                        "end": 95,
                        "text": "Table 2",
                        "ref_id": "TABREF2"
                    }
                ],
                "eq_spans": [],
                "section": "Quantitative Analysis",
                "sec_num": "3.3"
            },
            {
                "text": "In Table 2 , from the results of Methods 6 and 7, we can see that removing either product or customer embedding degrades the performance of the model. The result of Method 8 shows that embeddings jointly learned from the merged crosssource graph achieve better performance on our downstream task. We also observe that the ranking returned by the product search system is still an important feature as Method 6 shows.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 3,
                        "end": 10,
                        "text": "Table 2",
                        "ref_id": "TABREF2"
                    }
                ],
                "eq_spans": [],
                "section": "Quantitative Analysis",
                "sec_num": "3.3"
            },
            {
                "text": "Dev Accuracy Test Accuracy 1 QUARTS 0.0 0.0 2 Purchased +10.5 +8.5 3 ComplEx +25.7 +16.1 4 Our Model +32.9 +24.6 Our model promotes an organic product as the customer probably prefers organic products based on the shopping records. Table 3 : Positive examples in the data set. Candidates are listed in the order returned by our method. The number before each candidate is the original ranking returned by QUARTS. In the candidate column, we highlight the purchased products . In the history column, we highlight related records .",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 232,
                        "end": 239,
                        "text": "Table 3",
                        "ref_id": null
                    }
                ],
                "eq_spans": [],
                "section": "Method",
                "sec_num": null
            },
            {
                "text": "In Table 3 and Table 4 , we show some positive and negative examples in the data set. From Table 3 we can see that multiple sources of evidence in the constructed heterogeneous knowledge graphs are complimentary and the combination of them successfully promotes various entities which match customers' interests. Table 4 shows examples where our model fails to return the correct item. In many cases, such as Example #4, the purchased product and the top ranked one only differ in packaging size. We also observe that sometimes customers may not repurchase a product even if it is in the candidate list.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 3,
                        "end": 22,
                        "text": "Table 3 and Table 4",
                        "ref_id": "TABREF4"
                    },
                    {
                        "start": 91,
                        "end": 98,
                        "text": "Table 3",
                        "ref_id": null
                    },
                    {
                        "start": 313,
                        "end": 320,
                        "text": "Table 4",
                        "ref_id": "TABREF4"
                    }
                ],
                "eq_spans": [],
                "section": "Qualitative Analysis",
                "sec_num": "3.4"
            },
            {
                "text": "To better understand the remaining errors, we randomly sample 100 examples where our model fails to predict the purchased items. As Figure 3 illustrates, we analyze these examples and classify the possible reasons into the following categories. Different size. The predicted product and ground truth are the same product but differ in size. For example, while our model predicts \"Lipton Herbal Tea Bags, Peach Mango, 20 ct\", the customer purchases another item \"Lipton Tea Herbal Peach Mango (pack of 2)\", which is actually the same product in 2 pack. Purchased. The customer purchased the predicted product before but decides not to repurchase it. This usually happens in categories (e.g., toothpaste) where customers are more willing to try new products. Additionally, customers may be less likely to repurchase a product in some categories such as books and electronics. Uninformative title. The purchased product has 2lbfor smoothie, coffee and drink mixes Our model promote \"Anthony's Organic Cocoa Powder\" as it has been purchased twice by the customer. an uninformative title and is therefore not promoted. For example, when the customer searches for \"masaman curry paste maesri\", our model promotes \"Maesri Thai Masaman Curry -4 oz (pack of 4)\", while the customer purchases \"6 Can (4oz. Each) of Thai Green Red Yellow Curry Pastes Set\", which is also a Maesri product, but this key information is missing from its title. Similar title. The title of the predicted product is similar to the titles of some purchased products in the customer's history in a less important aspect. For example, the model promotes a \"moisturizing\" shave gel because the customer has purchased a \"moisturizing\" body wash, whereas the customer decides to purchase a product for \"sensitive skin\". Brand. The customer has purchased one or more products of the same brand. Attribute. The customer has purchased one or more products with the same attribute (e.g., organic, keto, kosher). Other. The model may fail to predict the purchased item in other uncategorized cases. For example, when a customer searches for \"nail clippers\" but has purchased only food in the past, the model is unlikely to utilize the history records to improve the ranking.",
                "cite_spans": [],
                "ref_spans": [
                    {
                        "start": 132,
                        "end": 140,
                        "text": "Figure 3",
                        "ref_id": "FIGREF3"
                    }
                ],
                "eq_spans": [],
                "section": "Qualitative Analysis",
                "sec_num": "3.4"
            },
            {
                "text": "Although our framework can improve the accuracy of predicting products that will be purchased, there are still some remaining challenges. We pro-pose the following potential solutions for further improvement. Incorporating more informative features. Some important features that affect purchase decisions are still missing in our framework, such as the average rating, customer review comments, and number of ratings. For example, we may promote the highest rated product for a customer who usually buys products with high ratings.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Remaining Challenges",
                "sec_num": "3.5"
            },
            {
                "text": "Building a more comprehensive cross-source customer-product graph. In this work, we merge the customer-product graph and product knowledge graph into a single graph, which has been proved to produce better embeddings for our target task. A natural extension is to include records from more sources, such as music or video playing history, and multimedia features such as product pictures. Modeling the interactions among purchase behaviors. Our current attention-based method that generates history representations is \"flat\" and ignores the relationship among purchase behaviors. For example, for a customer who previously purchases a pod coffee maker, we should promote coffee capsules in the candidates over coffee beans or grounds. Incorporating cohort features. When dealing with customers with limited or even no previous shopping records, a step forward is to cluster customers and produce cohort-based representations. In this way, customers can be better represented collectively through other similar customers, espe-cially when we combine their interaction history from other domains and build a more comprehensive graph as suggested above.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Remaining Challenges",
                "sec_num": "3.5"
            },
            {
                "text": "To sum up, shopping is a complex behavior, the importance weights of various features may vary across types of products. For instance, customers may like to stick to the same brand for beauty products while changing the size, depending on their needs. In contrast, for clothing customers may care more about visual features rather than text descriptions, and for books customers rarely purchase the same book more than once. To tackle these remaining challenges, we aim to extend our framework to incorporate more multimedia features, extract knowledge from review comments, and present recommendation results in a more hierarchical way by clustering variants of the same product and presenting their different attributes.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Remaining Challenges",
                "sec_num": "3.5"
            },
            {
                "text": "A variety of neural models (Gupta et al., 2017; Kolitsas et al., 2018; Cao et al., 2018; Sil et al., 2018; Gillick et al., 2019; Logeswaran et al., 2019; Wu et al., 2019; Agarwal and Bikel, 2020) have been applied to entity linking in recent years. Compared to traditional entity linking, our task is different in three aspects: (1) Our mentions are typically vague and occur in uninformative contexts, such as \"add toothpaste to my cart\" ; (2) A mention may be reasonably linked to multiple entities, while only one of them is considered \"correct\" (purchased by the customer); (3) The ground truth for the same mention can be different for different customers.",
                "cite_spans": [
                    {
                        "start": 27,
                        "end": 47,
                        "text": "(Gupta et al., 2017;",
                        "ref_id": "BIBREF8"
                    },
                    {
                        "start": 48,
                        "end": 70,
                        "text": "Kolitsas et al., 2018;",
                        "ref_id": "BIBREF15"
                    },
                    {
                        "start": 71,
                        "end": 88,
                        "text": "Cao et al., 2018;",
                        "ref_id": "BIBREF4"
                    },
                    {
                        "start": 89,
                        "end": 106,
                        "text": "Sil et al., 2018;",
                        "ref_id": "BIBREF36"
                    },
                    {
                        "start": 107,
                        "end": 128,
                        "text": "Gillick et al., 2019;",
                        "ref_id": "BIBREF6"
                    },
                    {
                        "start": 129,
                        "end": 153,
                        "text": "Logeswaran et al., 2019;",
                        "ref_id": "BIBREF20"
                    },
                    {
                        "start": 154,
                        "end": 170,
                        "text": "Wu et al., 2019;",
                        "ref_id": "BIBREF47"
                    },
                    {
                        "start": 171,
                        "end": 195,
                        "text": "Agarwal and Bikel, 2020)",
                        "ref_id": "BIBREF1"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Neural Entity Linking",
                "sec_num": "4.1"
            },
            {
                "text": "A recommender system is an information filtering system that aims to suggest a list of items in which a user may be interested. Content-based filtering (Billsus and Pazzani, 2000; Aciar et al., 2007; and collaborative filtering (Shardanand and Maes, 1995; Konstan et al., 1997; Linden et al., 2003; Zhao and Shang, 2010) are two common approaches used in recommender systems. In recent years, researchers have also applied neural methods to improve the quality of recommendations (Xue et al., 2017; He et al., 2017; Wang et al., 2019a,b) . Recommender systems usually rank items based on the user's past behaviors (e.g., purchasing, browsing, rating) and current context (Linden et al., 2003; Smith and Linden, 2017) , whereas the results are not constrained by queries. Instead, our task requires a specific query and only returns the product that is most likely to be purchased from a list of relevant candidates.",
                "cite_spans": [
                    {
                        "start": 152,
                        "end": 179,
                        "text": "(Billsus and Pazzani, 2000;",
                        "ref_id": "BIBREF2"
                    },
                    {
                        "start": 180,
                        "end": 199,
                        "text": "Aciar et al., 2007;",
                        "ref_id": "BIBREF0"
                    },
                    {
                        "start": 228,
                        "end": 255,
                        "text": "(Shardanand and Maes, 1995;",
                        "ref_id": "BIBREF35"
                    },
                    {
                        "start": 256,
                        "end": 277,
                        "text": "Konstan et al., 1997;",
                        "ref_id": "BIBREF16"
                    },
                    {
                        "start": 278,
                        "end": 298,
                        "text": "Linden et al., 2003;",
                        "ref_id": "BIBREF17"
                    },
                    {
                        "start": 299,
                        "end": 320,
                        "text": "Zhao and Shang, 2010)",
                        "ref_id": "BIBREF55"
                    },
                    {
                        "start": 480,
                        "end": 498,
                        "text": "(Xue et al., 2017;",
                        "ref_id": "BIBREF48"
                    },
                    {
                        "start": 499,
                        "end": 515,
                        "text": "He et al., 2017;",
                        "ref_id": "BIBREF9"
                    },
                    {
                        "start": 516,
                        "end": 537,
                        "text": "Wang et al., 2019a,b)",
                        "ref_id": null
                    },
                    {
                        "start": 671,
                        "end": 692,
                        "text": "(Linden et al., 2003;",
                        "ref_id": "BIBREF17"
                    },
                    {
                        "start": 693,
                        "end": 716,
                        "text": "Smith and Linden, 2017)",
                        "ref_id": "BIBREF37"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Personalized Recommendation",
                "sec_num": "4.2"
            },
            {
                "text": "Various methods have been proposed to learn lowdimensional vectors for nodes in knowledge graphs. Knowledge graph embedding methods, such as TransE (Bordes et al., 2013) , DistMult (Yang et al., 2014) , ComplEx (Trouillon et al., 2016) , and Ro-tatE (Sun et al., 2018) , typically represent the head entity, relation, and tail entity in each triplet in the knowledge graph as vectors and aim to rank true triplets higher than corresponding corrupted triplets. Matrix Factorization-based methods (He and Niyogi, 2004; Nickel et al., 2011; Qiu et al., 2018) represent the graph as a matrix and obtain node vectors by factorizing this matrix. Another category of frameworks (Perozzi et al., 2014b; Yang et al., 2015; Grover and Leskovec, 2016) use random walk to sample paths from the input graph and learn node embeddings from the sampled paths using neural models such as SkipGram and LSTM.",
                "cite_spans": [
                    {
                        "start": 148,
                        "end": 169,
                        "text": "(Bordes et al., 2013)",
                        "ref_id": "BIBREF3"
                    },
                    {
                        "start": 181,
                        "end": 200,
                        "text": "(Yang et al., 2014)",
                        "ref_id": "BIBREF49"
                    },
                    {
                        "start": 211,
                        "end": 235,
                        "text": "(Trouillon et al., 2016)",
                        "ref_id": "BIBREF41"
                    },
                    {
                        "start": 250,
                        "end": 268,
                        "text": "(Sun et al., 2018)",
                        "ref_id": "BIBREF40"
                    },
                    {
                        "start": 495,
                        "end": 516,
                        "text": "(He and Niyogi, 2004;",
                        "ref_id": "BIBREF10"
                    },
                    {
                        "start": 517,
                        "end": 537,
                        "text": "Nickel et al., 2011;",
                        "ref_id": "BIBREF23"
                    },
                    {
                        "start": 538,
                        "end": 555,
                        "text": "Qiu et al., 2018)",
                        "ref_id": "BIBREF31"
                    },
                    {
                        "start": 671,
                        "end": 694,
                        "text": "(Perozzi et al., 2014b;",
                        "ref_id": "BIBREF29"
                    },
                    {
                        "start": 695,
                        "end": 713,
                        "text": "Yang et al., 2015;",
                        "ref_id": "BIBREF50"
                    },
                    {
                        "start": 714,
                        "end": 740,
                        "text": "Grover and Leskovec, 2016)",
                        "ref_id": "BIBREF7"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Graph Embedding",
                "sec_num": "4.3"
            },
            {
                "text": "The earliest study of mining heterogeneous network dates back to (Sun et al., 2009) , which coins the concept of Heterogeneous Information Network. After that, heterogeneous network has been applied to a range of tasks, including ranking , similarity search (Sun et al., 2011) , link prediction (Dong et al., 2015) , academic paper recommendation (Pan et al., 2015) , and malicious account detection .",
                "cite_spans": [
                    {
                        "start": 65,
                        "end": 83,
                        "text": "(Sun et al., 2009)",
                        "ref_id": "BIBREF39"
                    },
                    {
                        "start": 258,
                        "end": 276,
                        "text": "(Sun et al., 2011)",
                        "ref_id": "BIBREF38"
                    },
                    {
                        "start": 295,
                        "end": 314,
                        "text": "(Dong et al., 2015)",
                        "ref_id": "BIBREF5"
                    },
                    {
                        "start": 347,
                        "end": 365,
                        "text": "(Pan et al., 2015)",
                        "ref_id": "BIBREF25"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Heterogeneous Network",
                "sec_num": "4.4"
            },
            {
                "text": "Recently, with the advent of graph neural network, many methods based on this new paradigm have been proposed to learn graph representations on heterogeneous graphs, such as Heterogeneous Graph Neural Network (HetGNN) (Zhang et al., 2019) , Heterogeneous Graph Attention Network (HAN) (Wang et al., 2019c) , and Heterogeneous Graph Transformer (HGT) (Hu et al., 2020) .",
                "cite_spans": [
                    {
                        "start": 218,
                        "end": 238,
                        "text": "(Zhang et al., 2019)",
                        "ref_id": "BIBREF53"
                    },
                    {
                        "start": 285,
                        "end": 305,
                        "text": "(Wang et al., 2019c)",
                        "ref_id": "BIBREF46"
                    },
                    {
                        "start": 350,
                        "end": 367,
                        "text": "(Hu et al., 2020)",
                        "ref_id": "BIBREF11"
                    }
                ],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Heterogeneous Network",
                "sec_num": "4.4"
            },
            {
                "text": "We propose a novel framework to jointly learn customer and product representations based on a crosssource heterogeneous graph constructed from customers' purchase history and the product knowledge graph to improve personalized entity resolution. Experiments show that our framework can effectively increase the purchase rate of the top ranked products. In the future, we plan to investigate better approaches to integrating personalized features and extend the framework to cross-lingual cross-media settings and generate conversations for more proactive and explainable entity recommendation and summarization.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "Conclusion and Future Work",
                "sec_num": "5"
            },
            {
                "text": "Bullet points that outline the main features of a product.",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            },
            {
                "text": "https://huggingface.co/transformers/ pretrained_models.html 3 https://nlp.stanford.edu/projects/glove/",
                "cite_spans": [],
                "ref_spans": [],
                "eq_spans": [],
                "section": "",
                "sec_num": null
            }
        ],
        "back_matter": [],
        "bib_entries": {
            "BIBREF0": {
                "ref_id": "b0",
                "title": "Informed recommender: Basing recommendations on consumer product reviews",
                "authors": [
                    {
                        "first": "Silvana",
                        "middle": [],
                        "last": "Aciar",
                        "suffix": ""
                    },
                    {
                        "first": "Debbie",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Simeon",
                        "middle": [],
                        "last": "Simoff",
                        "suffix": ""
                    },
                    {
                        "first": "John",
                        "middle": [],
                        "last": "Debenham",
                        "suffix": ""
                    }
                ],
                "year": 2007,
                "venue": "IEEE Intelligent systems",
                "volume": "22",
                "issue": "3",
                "pages": "39--47",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Silvana Aciar, Debbie Zhang, Simeon Simoff, and John Debenham. 2007. Informed recommender: Bas- ing recommendations on consumer product reviews. IEEE Intelligent systems, 22(3):39-47.",
                "links": null
            },
            "BIBREF1": {
                "ref_id": "b1",
                "title": "Entity linking via dual and cross-attention encoders",
                "authors": [
                    {
                        "first": "Oshin",
                        "middle": [],
                        "last": "Agarwal",
                        "suffix": ""
                    },
                    {
                        "first": "M",
                        "middle": [],
                        "last": "Daniel",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Bikel",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:2004.03555"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Oshin Agarwal and Daniel M Bikel. 2020. Entity link- ing via dual and cross-attention encoders. arXiv preprint arXiv:2004.03555.",
                "links": null
            },
            "BIBREF2": {
                "ref_id": "b2",
                "title": "User modeling for adaptive news access. User modeling and user-adapted interaction",
                "authors": [
                    {
                        "first": "Daniel",
                        "middle": [],
                        "last": "Billsus",
                        "suffix": ""
                    },
                    {
                        "first": "J",
                        "middle": [],
                        "last": "Michael",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Pazzani",
                        "suffix": ""
                    }
                ],
                "year": 2000,
                "venue": "",
                "volume": "10",
                "issue": "",
                "pages": "147--180",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Daniel Billsus and Michael J Pazzani. 2000. User mod- eling for adaptive news access. User modeling and user-adapted interaction, 10(2-3):147-180.",
                "links": null
            },
            "BIBREF3": {
                "ref_id": "b3",
                "title": "Translating embeddings for modeling multirelational data",
                "authors": [
                    {
                        "first": "Antoine",
                        "middle": [],
                        "last": "Bordes",
                        "suffix": ""
                    },
                    {
                        "first": "Nicolas",
                        "middle": [],
                        "last": "Usunier",
                        "suffix": ""
                    },
                    {
                        "first": "Alberto",
                        "middle": [],
                        "last": "Garcia-Duran",
                        "suffix": ""
                    },
                    {
                        "first": "Jason",
                        "middle": [],
                        "last": "Weston",
                        "suffix": ""
                    },
                    {
                        "first": "Oksana",
                        "middle": [],
                        "last": "Yakhnenko",
                        "suffix": ""
                    }
                ],
                "year": 2013,
                "venue": "Advances in neural information processing systems",
                "volume": "26",
                "issue": "",
                "pages": "2787--2795",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Antoine Bordes, Nicolas Usunier, Alberto Garcia- Duran, Jason Weston, and Oksana Yakhnenko. 2013. Translating embeddings for modeling multi- relational data. Advances in neural information pro- cessing systems, 26:2787-2795.",
                "links": null
            },
            "BIBREF4": {
                "ref_id": "b4",
                "title": "Neural collective entity linking",
                "authors": [
                    {
                        "first": "Yixin",
                        "middle": [],
                        "last": "Cao",
                        "suffix": ""
                    },
                    {
                        "first": "Lei",
                        "middle": [],
                        "last": "Hou",
                        "suffix": ""
                    },
                    {
                        "first": "Juanzi",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Zhiyuan",
                        "middle": [],
                        "last": "Liu",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the 27th International Conference on Computational Linguistics",
                "volume": "",
                "issue": "",
                "pages": "675--686",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Yixin Cao, Lei Hou, Juanzi Li, and Zhiyuan Liu. 2018. Neural collective entity linking. In Proceedings of the 27th International Conference on Computational Linguistics, pages 675-686.",
                "links": null
            },
            "BIBREF5": {
                "ref_id": "b5",
                "title": "Coupledlp: Link prediction in coupled networks",
                "authors": [
                    {
                        "first": "Yuxiao",
                        "middle": [],
                        "last": "Dong",
                        "suffix": ""
                    },
                    {
                        "first": "Jing",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Jie",
                        "middle": [],
                        "last": "Tang",
                        "suffix": ""
                    },
                    {
                        "first": "V",
                        "middle": [],
                        "last": "Nitesh",
                        "suffix": ""
                    },
                    {
                        "first": "Bai",
                        "middle": [],
                        "last": "Chawla",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    }
                ],
                "year": 2015,
                "venue": "Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining",
                "volume": "",
                "issue": "",
                "pages": "199--208",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Yuxiao Dong, Jing Zhang, Jie Tang, Nitesh V Chawla, and Bai Wang. 2015. Coupledlp: Link prediction in coupled networks. In Proceedings of the 21th ACM SIGKDD International Conference on Knowl- edge Discovery and Data Mining, pages 199-208.",
                "links": null
            },
            "BIBREF6": {
                "ref_id": "b6",
                "title": "Learning dense representations for entity retrieval",
                "authors": [
                    {
                        "first": "Dan",
                        "middle": [],
                        "last": "Gillick",
                        "suffix": ""
                    },
                    {
                        "first": "Sayali",
                        "middle": [],
                        "last": "Kulkarni",
                        "suffix": ""
                    },
                    {
                        "first": "Larry",
                        "middle": [],
                        "last": "Lansing",
                        "suffix": ""
                    },
                    {
                        "first": "Alessandro",
                        "middle": [],
                        "last": "Presta",
                        "suffix": ""
                    },
                    {
                        "first": "Jason",
                        "middle": [],
                        "last": "Baldridge",
                        "suffix": ""
                    },
                    {
                        "first": "Eugene",
                        "middle": [],
                        "last": "Ie",
                        "suffix": ""
                    },
                    {
                        "first": "Diego",
                        "middle": [],
                        "last": "Garcia-Olano",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 23rd Conference on Computational Natural Language Learning (CoNLL)",
                "volume": "",
                "issue": "",
                "pages": "528--537",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Dan Gillick, Sayali Kulkarni, Larry Lansing, Alessan- dro Presta, Jason Baldridge, Eugene Ie, and Diego Garcia-Olano. 2019. Learning dense representa- tions for entity retrieval. In Proceedings of the 23rd Conference on Computational Natural Lan- guage Learning (CoNLL), pages 528-537.",
                "links": null
            },
            "BIBREF7": {
                "ref_id": "b7",
                "title": "node2vec: Scalable feature learning for networks",
                "authors": [
                    {
                        "first": "Aditya",
                        "middle": [],
                        "last": "Grover",
                        "suffix": ""
                    },
                    {
                        "first": "Jure",
                        "middle": [],
                        "last": "Leskovec",
                        "suffix": ""
                    }
                ],
                "year": 2016,
                "venue": "Proceedings of the 22nd ACM SIGKDD international conference on Knowledge discovery and data mining",
                "volume": "",
                "issue": "",
                "pages": "855--864",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Aditya Grover and Jure Leskovec. 2016. node2vec: Scalable feature learning for networks. In Proceed- ings of the 22nd ACM SIGKDD international con- ference on Knowledge discovery and data mining, pages 855-864.",
                "links": null
            },
            "BIBREF8": {
                "ref_id": "b8",
                "title": "Entity linking via joint encoding of types, descriptions, and context",
                "authors": [
                    {
                        "first": "Nitish",
                        "middle": [],
                        "last": "Gupta",
                        "suffix": ""
                    },
                    {
                        "first": "Sameer",
                        "middle": [],
                        "last": "Singh",
                        "suffix": ""
                    },
                    {
                        "first": "Dan",
                        "middle": [],
                        "last": "Roth",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "Proceedings of the 2017 Conference on Empirical Methods in Natural Language Processing",
                "volume": "",
                "issue": "",
                "pages": "2681--2690",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Nitish Gupta, Sameer Singh, and Dan Roth. 2017. En- tity linking via joint encoding of types, descriptions, and context. In Proceedings of the 2017 Conference on Empirical Methods in Natural Language Process- ing, pages 2681-2690.",
                "links": null
            },
            "BIBREF9": {
                "ref_id": "b9",
                "title": "Neural collaborative filtering",
                "authors": [
                    {
                        "first": "Xiangnan",
                        "middle": [],
                        "last": "He",
                        "suffix": ""
                    },
                    {
                        "first": "Lizi",
                        "middle": [],
                        "last": "Liao",
                        "suffix": ""
                    },
                    {
                        "first": "Hanwang",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Liqiang",
                        "middle": [],
                        "last": "Nie",
                        "suffix": ""
                    },
                    {
                        "first": "Xia",
                        "middle": [],
                        "last": "Hu",
                        "suffix": ""
                    },
                    {
                        "first": "Tat-Seng",
                        "middle": [],
                        "last": "Chua",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "Proceedings of the 26th international conference on world wide web",
                "volume": "",
                "issue": "",
                "pages": "173--182",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Xiangnan He, Lizi Liao, Hanwang Zhang, Liqiang Nie, Xia Hu, and Tat-Seng Chua. 2017. Neural collab- orative filtering. In Proceedings of the 26th inter- national conference on world wide web, pages 173- 182.",
                "links": null
            },
            "BIBREF10": {
                "ref_id": "b10",
                "title": "Locality preserving projections",
                "authors": [
                    {
                        "first": "Xiaofei",
                        "middle": [],
                        "last": "He",
                        "suffix": ""
                    },
                    {
                        "first": "Partha",
                        "middle": [],
                        "last": "Niyogi",
                        "suffix": ""
                    }
                ],
                "year": 2004,
                "venue": "Advances in neural information processing systems",
                "volume": "",
                "issue": "",
                "pages": "153--160",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Xiaofei He and Partha Niyogi. 2004. Locality preserv- ing projections. In Advances in neural information processing systems, pages 153-160.",
                "links": null
            },
            "BIBREF11": {
                "ref_id": "b11",
                "title": "Heterogeneous graph transformer",
                "authors": [
                    {
                        "first": "Ziniu",
                        "middle": [],
                        "last": "Hu",
                        "suffix": ""
                    },
                    {
                        "first": "Yuxiao",
                        "middle": [],
                        "last": "Dong",
                        "suffix": ""
                    },
                    {
                        "first": "Kuansan",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    },
                    {
                        "first": "Yizhou",
                        "middle": [],
                        "last": "Sun",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of The Web Conference 2020",
                "volume": "",
                "issue": "",
                "pages": "2704--2710",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Ziniu Hu, Yuxiao Dong, Kuansan Wang, and Yizhou Sun. 2020. Heterogeneous graph transformer. In Proceedings of The Web Conference 2020, pages 2704-2710.",
                "links": null
            },
            "BIBREF12": {
                "ref_id": "b12",
                "title": "Knowledge base population: Successful approaches and challenges",
                "authors": [
                    {
                        "first": "Heng",
                        "middle": [],
                        "last": "Ji",
                        "suffix": ""
                    },
                    {
                        "first": "Ralph",
                        "middle": [],
                        "last": "Grishman",
                        "suffix": ""
                    }
                ],
                "year": 2011,
                "venue": "Proc. ACL2011",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Heng Ji and Ralph Grishman. 2011. Knowledge base population: Successful approaches and challenges. In Proc. ACL2011.",
                "links": null
            },
            "BIBREF13": {
                "ref_id": "b13",
                "title": "Overview of tac-kbp2015 tri-lingual entity discovery and linking",
                "authors": [
                    {
                        "first": "Heng",
                        "middle": [],
                        "last": "Ji",
                        "suffix": ""
                    },
                    {
                        "first": "Joel",
                        "middle": [],
                        "last": "Nothman",
                        "suffix": ""
                    },
                    {
                        "first": "Ben",
                        "middle": [],
                        "last": "Hachey",
                        "suffix": ""
                    },
                    {
                        "first": "Radu",
                        "middle": [],
                        "last": "Florian",
                        "suffix": ""
                    }
                ],
                "year": 2015,
                "venue": "Proc. Text Analysis Conference (TAC2015)",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Heng Ji, Joel Nothman, Ben Hachey, and Radu Florian. 2015. Overview of tac-kbp2015 tri-lingual entity discovery and linking. In Proc. Text Analysis Con- ference (TAC2015).",
                "links": null
            },
            "BIBREF14": {
                "ref_id": "b14",
                "title": "Ranking-based classification of heterogeneous information networks",
                "authors": [
                    {
                        "first": "Ming",
                        "middle": [],
                        "last": "Ji",
                        "suffix": ""
                    },
                    {
                        "first": "Jiawei",
                        "middle": [],
                        "last": "Han",
                        "suffix": ""
                    },
                    {
                        "first": "Marina",
                        "middle": [],
                        "last": "Danilevsky",
                        "suffix": ""
                    }
                ],
                "year": 2011,
                "venue": "Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining",
                "volume": "",
                "issue": "",
                "pages": "1298--1306",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Ming Ji, Jiawei Han, and Marina Danilevsky. 2011. Ranking-based classification of heterogeneous in- formation networks. In Proceedings of the 17th ACM SIGKDD international conference on Knowl- edge discovery and data mining, pages 1298-1306.",
                "links": null
            },
            "BIBREF15": {
                "ref_id": "b15",
                "title": "End-to-end neural entity linking",
                "authors": [
                    {
                        "first": "Nikolaos",
                        "middle": [],
                        "last": "Kolitsas",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Octavian-Eugen",
                        "suffix": ""
                    },
                    {
                        "first": "Thomas",
                        "middle": [],
                        "last": "Ganea",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Hofmann",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the 22nd Conference on Computational Natural Language Learning",
                "volume": "",
                "issue": "",
                "pages": "519--529",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Nikolaos Kolitsas, Octavian-Eugen Ganea, and Thomas Hofmann. 2018. End-to-end neural entity linking. In Proceedings of the 22nd Conference on Computational Natural Language Learning, pages 519-529.",
                "links": null
            },
            "BIBREF16": {
                "ref_id": "b16",
                "title": "Grouplens: applying collaborative filtering to usenet news",
                "authors": [
                    {
                        "first": "",
                        "middle": [],
                        "last": "Joseph A Konstan",
                        "suffix": ""
                    },
                    {
                        "first": "N",
                        "middle": [],
                        "last": "Bradley",
                        "suffix": ""
                    },
                    {
                        "first": "David",
                        "middle": [],
                        "last": "Miller",
                        "suffix": ""
                    },
                    {
                        "first": "Jonathan",
                        "middle": [
                            "L"
                        ],
                        "last": "Maltz",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Herlocker",
                        "suffix": ""
                    },
                    {
                        "first": "John",
                        "middle": [],
                        "last": "Lee R Gordon",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Riedl",
                        "suffix": ""
                    }
                ],
                "year": 1997,
                "venue": "Communications of the ACM",
                "volume": "40",
                "issue": "3",
                "pages": "77--87",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Joseph A Konstan, Bradley N Miller, David Maltz, Jonathan L Herlocker, Lee R Gordon, and John Riedl. 1997. Grouplens: applying collaborative fil- tering to usenet news. Communications of the ACM, 40(3):77-87.",
                "links": null
            },
            "BIBREF17": {
                "ref_id": "b17",
                "title": "Amazon. com recommendations: Item-to-item collaborative filtering",
                "authors": [
                    {
                        "first": "Greg",
                        "middle": [],
                        "last": "Linden",
                        "suffix": ""
                    },
                    {
                        "first": "Brent",
                        "middle": [],
                        "last": "Smith",
                        "suffix": ""
                    },
                    {
                        "first": "Jeremy",
                        "middle": [],
                        "last": "York",
                        "suffix": ""
                    }
                ],
                "year": 2003,
                "venue": "IEEE Internet computing",
                "volume": "7",
                "issue": "1",
                "pages": "76--80",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Greg Linden, Brent Smith, and Jeremy York. 2003. Amazon. com recommendations: Item-to-item col- laborative filtering. IEEE Internet computing, 7(1):76-80.",
                "links": null
            },
            "BIBREF18": {
                "ref_id": "b18",
                "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 pretraining ap- proach. arXiv preprint arXiv:1907.11692.",
                "links": null
            },
            "BIBREF19": {
                "ref_id": "b19",
                "title": "Heterogeneous graph neural networks for malicious account detection",
                "authors": [
                    {
                        "first": "Ziqi",
                        "middle": [],
                        "last": "Liu",
                        "suffix": ""
                    },
                    {
                        "first": "Chaochao",
                        "middle": [],
                        "last": "Chen",
                        "suffix": ""
                    },
                    {
                        "first": "Xinxing",
                        "middle": [],
                        "last": "Yang",
                        "suffix": ""
                    },
                    {
                        "first": "Jun",
                        "middle": [],
                        "last": "Zhou",
                        "suffix": ""
                    },
                    {
                        "first": "Xiaolong",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Le",
                        "middle": [],
                        "last": "Song",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the 27th ACM International Conference on Information and Knowledge Management",
                "volume": "",
                "issue": "",
                "pages": "2077--2085",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Ziqi Liu, Chaochao Chen, Xinxing Yang, Jun Zhou, Xi- aolong Li, and Le Song. 2018. Heterogeneous graph neural networks for malicious account detection. In Proceedings of the 27th ACM International Confer- ence on Information and Knowledge Management, pages 2077-2085.",
                "links": null
            },
            "BIBREF20": {
                "ref_id": "b20",
                "title": "Zero-shot entity linking by reading entity descriptions",
                "authors": [
                    {
                        "first": "Lajanugen",
                        "middle": [],
                        "last": "Logeswaran",
                        "suffix": ""
                    },
                    {
                        "first": "Ming-Wei",
                        "middle": [],
                        "last": "Chang",
                        "suffix": ""
                    },
                    {
                        "first": "Kenton",
                        "middle": [],
                        "last": "Lee",
                        "suffix": ""
                    },
                    {
                        "first": "Kristina",
                        "middle": [],
                        "last": "Toutanova",
                        "suffix": ""
                    },
                    {
                        "first": "Jacob",
                        "middle": [],
                        "last": "Devlin",
                        "suffix": ""
                    },
                    {
                        "first": "Honglak",
                        "middle": [],
                        "last": "Lee",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 57th Annual Meeting of the Association for Computational Linguistics",
                "volume": "",
                "issue": "",
                "pages": "3449--3460",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Lajanugen Logeswaran, Ming-Wei Chang, Kenton Lee, Kristina Toutanova, Jacob Devlin, and Honglak Lee. 2019. Zero-shot entity linking by reading entity de- scriptions. In Proceedings of the 57th Annual Meet- ing of the Association for Computational Linguistics, pages 3449-3460.",
                "links": null
            },
            "BIBREF21": {
                "ref_id": "b21",
                "title": "Decoupled weight decay regularization",
                "authors": [
                    {
                        "first": "Ilya",
                        "middle": [],
                        "last": "Loshchilov",
                        "suffix": ""
                    },
                    {
                        "first": "Frank",
                        "middle": [],
                        "last": "Hutter",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "International Conference on Learning Representations",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Ilya Loshchilov and Frank Hutter. 2018. Decoupled weight decay regularization. In International Con- ference on Learning Representations.",
                "links": null
            },
            "BIBREF22": {
                "ref_id": "b22",
                "title": "Learning robust models for e-commerce product search",
                "authors": [
                    {
                        "first": "Thanh",
                        "middle": [],
                        "last": "Nguyen",
                        "suffix": ""
                    },
                    {
                        "first": "Nikhil",
                        "middle": [],
                        "last": "Rao",
                        "suffix": ""
                    },
                    {
                        "first": "Karthik",
                        "middle": [],
                        "last": "Subbian",
                        "suffix": ""
                    }
                ],
                "year": 2020,
                "venue": "Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics",
                "volume": "",
                "issue": "",
                "pages": "6861--6869",
                "other_ids": {
                    "DOI": [
                        "10.18653/v1/2020.acl-main.614"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Thanh Nguyen, Nikhil Rao, and Karthik Subbian. 2020. Learning robust models for e-commerce product search. In Proceedings of the 58th Annual Meet- ing of the Association for Computational Linguistics, pages 6861-6869, Online. Association for Computa- tional Linguistics.",
                "links": null
            },
            "BIBREF23": {
                "ref_id": "b23",
                "title": "A three-way model for collective learning on multi-relational data",
                "authors": [
                    {
                        "first": "Maximilian",
                        "middle": [],
                        "last": "Nickel",
                        "suffix": ""
                    },
                    {
                        "first": "Hans-Peter",
                        "middle": [],
                        "last": "Volker Tresp",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Kriegel",
                        "suffix": ""
                    }
                ],
                "year": 2011,
                "venue": "In Icml",
                "volume": "",
                "issue": "11",
                "pages": "809--816",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Maximilian Nickel, Volker Tresp, and Hans-Peter Kriegel. 2011. A three-way model for collective learning on multi-relational data. In Icml, vol- ume 11, pages 809-816.",
                "links": null
            },
            "BIBREF24": {
                "ref_id": "b24",
                "title": "Semantic product search",
                "authors": [
                    {
                        "first": "Priyanka",
                        "middle": [],
                        "last": "Nigam",
                        "suffix": ""
                    },
                    {
                        "first": "Yiwei",
                        "middle": [],
                        "last": "Song",
                        "suffix": ""
                    },
                    {
                        "first": "Vijai",
                        "middle": [],
                        "last": "Mohan",
                        "suffix": ""
                    },
                    {
                        "first": "Vihan",
                        "middle": [],
                        "last": "Lakshman",
                        "suffix": ""
                    },
                    {
                        "first": "Ankit",
                        "middle": [],
                        "last": "Weitian (allen) Ding",
                        "suffix": ""
                    },
                    {
                        "first": "Choon",
                        "middle": [
                            "Hui"
                        ],
                        "last": "Shingavi",
                        "suffix": ""
                    },
                    {
                        "first": "Hao",
                        "middle": [],
                        "last": "Teo",
                        "suffix": ""
                    },
                    {
                        "first": "Bing",
                        "middle": [],
                        "last": "Gu",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Yin",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining",
                "volume": "",
                "issue": "",
                "pages": "2876--2885",
                "other_ids": {
                    "DOI": [
                        "10.1145/3292500.3330759"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Priyanka Nigam, Yiwei Song, Vijai Mohan, Vihan Lakshman, Weitian (Allen) Ding, Ankit Shingavi, Choon Hui Teo, Hao Gu, and Bing Yin. 2019. Se- mantic product search. In Proceedings of the 25th ACM SIGKDD International Conference on Knowl- edge Discovery & Data Mining, page 2876-2885, New York, NY, USA. Association for Computing Machinery.",
                "links": null
            },
            "BIBREF25": {
                "ref_id": "b25",
                "title": "Academic paper recommendation based on heterogeneous graph",
                "authors": [
                    {
                        "first": "Linlin",
                        "middle": [],
                        "last": "Pan",
                        "suffix": ""
                    },
                    {
                        "first": "Xinyu",
                        "middle": [],
                        "last": "Dai",
                        "suffix": ""
                    },
                    {
                        "first": "Shujian",
                        "middle": [],
                        "last": "Huang",
                        "suffix": ""
                    },
                    {
                        "first": "Jiajun",
                        "middle": [],
                        "last": "Chen",
                        "suffix": ""
                    }
                ],
                "year": 2015,
                "venue": "Chinese computational linguistics and natural language processing based on naturally annotated big data",
                "volume": "",
                "issue": "",
                "pages": "381--392",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Linlin Pan, Xinyu Dai, Shujian Huang, and Jiajun Chen. 2015. Academic paper recommendation based on heterogeneous graph. In Chinese computational linguistics and natural language processing based on naturally annotated big data, pages 381-392. Springer.",
                "links": null
            },
            "BIBREF26": {
                "ref_id": "b26",
                "title": "Crosslingual name tagging and linking for 282 languages",
                "authors": [
                    {
                        "first": "Xiaoman",
                        "middle": [],
                        "last": "Pan",
                        "suffix": ""
                    },
                    {
                        "first": "Boliang",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Jonathan",
                        "middle": [],
                        "last": "May",
                        "suffix": ""
                    },
                    {
                        "first": "Joel",
                        "middle": [],
                        "last": "Nothman",
                        "suffix": ""
                    },
                    {
                        "first": "Kevin",
                        "middle": [],
                        "last": "Knight",
                        "suffix": ""
                    },
                    {
                        "first": "Heng",
                        "middle": [],
                        "last": "Ji",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "Proc. the 55th Annual Meeting of the Association for Computational Linguistics (ACL2017)",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Xiaoman Pan, Boliang Zhang, Jonathan May, Joel Nothman, Kevin Knight, and Heng Ji. 2017. Cross- lingual name tagging and linking for 282 languages. In Proc. the 55th Annual Meeting of the Association for Computational Linguistics (ACL2017).",
                "links": null
            },
            "BIBREF27": {
                "ref_id": "b27",
                "title": "Glove: Global vectors for word representation",
                "authors": [
                    {
                        "first": "Jeffrey",
                        "middle": [],
                        "last": "Pennington",
                        "suffix": ""
                    },
                    {
                        "first": "Richard",
                        "middle": [],
                        "last": "Socher",
                        "suffix": ""
                    },
                    {
                        "first": "Christopher",
                        "middle": [
                            "D"
                        ],
                        "last": "Manning",
                        "suffix": ""
                    }
                ],
                "year": 2014,
                "venue": "Empirical Methods in Natural Language Processing (EMNLP)",
                "volume": "",
                "issue": "",
                "pages": "1532--1543",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Jeffrey Pennington, Richard Socher, and Christopher D. Manning. 2014. Glove: Global vectors for word rep- resentation. In Empirical Methods in Natural Lan- guage Processing (EMNLP), pages 1532-1543.",
                "links": null
            },
            "BIBREF28": {
                "ref_id": "b28",
                "title": "Deepwalk: online learning of social representations",
                "authors": [
                    {
                        "first": "Bryan",
                        "middle": [],
                        "last": "Perozzi",
                        "suffix": ""
                    },
                    {
                        "first": "Rami",
                        "middle": [],
                        "last": "Al-Rfou",
                        "suffix": ""
                    },
                    {
                        "first": "S",
                        "middle": [],
                        "last": "Skiena",
                        "suffix": ""
                    }
                ],
                "year": 2014,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Bryan Perozzi, Rami Al-Rfou, and S. Skiena. 2014a. Deepwalk: online learning of social representations. In KDD '14.",
                "links": null
            },
            "BIBREF29": {
                "ref_id": "b29",
                "title": "Deepwalk: Online learning of social representations",
                "authors": [
                    {
                        "first": "Bryan",
                        "middle": [],
                        "last": "Perozzi",
                        "suffix": ""
                    },
                    {
                        "first": "Rami",
                        "middle": [],
                        "last": "Al-Rfou",
                        "suffix": ""
                    },
                    {
                        "first": "Steven",
                        "middle": [],
                        "last": "Skiena",
                        "suffix": ""
                    }
                ],
                "year": 2014,
                "venue": "Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining",
                "volume": "",
                "issue": "",
                "pages": "701--710",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Bryan Perozzi, Rami Al-Rfou, and Steven Skiena. 2014b. Deepwalk: Online learning of social rep- resentations. In Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining, pages 701-710.",
                "links": null
            },
            "BIBREF30": {
                "ref_id": "b30",
                "title": "An analysis of the user occupational class through twitter content",
                "authors": [
                    {
                        "first": "Daniel",
                        "middle": [],
                        "last": "Preo\u0163iuc-Pietro",
                        "suffix": ""
                    },
                    {
                        "first": "Vasileios",
                        "middle": [],
                        "last": "Lampos",
                        "suffix": ""
                    },
                    {
                        "first": "Nikolaos",
                        "middle": [],
                        "last": "Aletras",
                        "suffix": ""
                    }
                ],
                "year": 2015,
                "venue": "Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing",
                "volume": "1",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Daniel Preo\u0163iuc-Pietro, Vasileios Lampos, and Niko- laos Aletras. 2015. An analysis of the user occupa- tional class through twitter content. In Proceedings of the 53rd Annual Meeting of the Association for Computational Linguistics and the 7th International Joint Conference on Natural Language Processing (Volume 1: Long Papers).",
                "links": null
            },
            "BIBREF31": {
                "ref_id": "b31",
                "title": "Network embedding as matrix factorization: Unifying deepwalk, line, pte, and node2vec",
                "authors": [
                    {
                        "first": "Jiezhong",
                        "middle": [],
                        "last": "Qiu",
                        "suffix": ""
                    },
                    {
                        "first": "Yuxiao",
                        "middle": [],
                        "last": "Dong",
                        "suffix": ""
                    },
                    {
                        "first": "Hao",
                        "middle": [],
                        "last": "Ma",
                        "suffix": ""
                    },
                    {
                        "first": "Jian",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Kuansan",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    },
                    {
                        "first": "Jie",
                        "middle": [],
                        "last": "Tang",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining",
                "volume": "",
                "issue": "",
                "pages": "459--467",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Jiezhong Qiu, Yuxiao Dong, Hao Ma, Jian Li, Kuansan Wang, and Jie Tang. 2018. Network embedding as matrix factorization: Unifying deepwalk, line, pte, and node2vec. In Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining, pages 459-467.",
                "links": null
            },
            "BIBREF32": {
                "ref_id": "b32",
                "title": "Characterizing and detecting hateful users on Twitter",
                "authors": [
                    {
                        "first": "M",
                        "middle": [
                            "H"
                        ],
                        "last": "Ribeiro",
                        "suffix": ""
                    },
                    {
                        "first": "Pedro",
                        "middle": [
                            "H"
                        ],
                        "last": "Calais",
                        "suffix": ""
                    },
                    {
                        "first": "Yuri",
                        "middle": [
                            "A"
                        ],
                        "last": "Santos",
                        "suffix": ""
                    },
                    {
                        "first": "V",
                        "middle": [],
                        "last": "Almeida",
                        "suffix": ""
                    },
                    {
                        "first": "W",
                        "middle": [],
                        "last": "Meira",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the Twelfth International AAAI Conference on Web and Social Media",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "M. H. Ribeiro, Pedro H. Calais, Yuri A. Santos, V. Almeida, and W. Meira. 2018. Characterizing and detecting hateful users on Twitter. In Proceedings of the Twelfth International AAAI Conference on Web and Social Media (ICWSM 2018).",
                "links": null
            },
            "BIBREF33": {
                "ref_id": "b33",
                "title": "Reasoning about entailment with neural attention",
                "authors": [
                    {
                        "first": "Tim",
                        "middle": [],
                        "last": "Rockt\u00e4schel",
                        "suffix": ""
                    },
                    {
                        "first": "Edward",
                        "middle": [],
                        "last": "Grefenstette",
                        "suffix": ""
                    },
                    {
                        "first": "K",
                        "middle": [],
                        "last": "Hermann",
                        "suffix": ""
                    },
                    {
                        "first": "Tom\u00e1s",
                        "middle": [],
                        "last": "Kocisk\u00fd",
                        "suffix": ""
                    },
                    {
                        "first": "P",
                        "middle": [],
                        "last": "Blunsom",
                        "suffix": ""
                    }
                ],
                "year": 2016,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Tim Rockt\u00e4schel, Edward Grefenstette, K. Hermann, Tom\u00e1s Kocisk\u00fd, and P. Blunsom. 2016. Reason- ing about entailment with neural attention. CoRR, abs/1509.06664.",
                "links": null
            },
            "BIBREF34": {
                "ref_id": "b34",
                "title": "Modeling relational data with graph convolutional networks",
                "authors": [
                    {
                        "first": "Michael",
                        "middle": [],
                        "last": "Schlichtkrull",
                        "suffix": ""
                    },
                    {
                        "first": "N",
                        "middle": [],
                        "last": "Thomas",
                        "suffix": ""
                    },
                    {
                        "first": "Peter",
                        "middle": [],
                        "last": "Kipf",
                        "suffix": ""
                    },
                    {
                        "first": "Rianne",
                        "middle": [],
                        "last": "Bloem",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Van Den",
                        "suffix": ""
                    },
                    {
                        "first": "Ivan",
                        "middle": [],
                        "last": "Berg",
                        "suffix": ""
                    },
                    {
                        "first": "Max",
                        "middle": [],
                        "last": "Titov",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Welling",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "European Semantic Web Conference",
                "volume": "",
                "issue": "",
                "pages": "593--607",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Michael Schlichtkrull, Thomas N Kipf, Peter Bloem, Rianne Van Den Berg, Ivan Titov, and Max Welling. 2018. Modeling relational data with graph convolu- tional networks. In European Semantic Web Confer- ence, pages 593-607. Springer.",
                "links": null
            },
            "BIBREF35": {
                "ref_id": "b35",
                "title": "Social information filtering: algorithms for automating \"word of mouth",
                "authors": [
                    {
                        "first": "Upendra",
                        "middle": [],
                        "last": "Shardanand",
                        "suffix": ""
                    },
                    {
                        "first": "Pattie",
                        "middle": [],
                        "last": "Maes",
                        "suffix": ""
                    }
                ],
                "year": 1995,
                "venue": "Proceedings of the SIGCHI conference on Human factors in computing systems",
                "volume": "",
                "issue": "",
                "pages": "210--217",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Upendra Shardanand and Pattie Maes. 1995. So- cial information filtering: algorithms for automating \"word of mouth\". In Proceedings of the SIGCHI conference on Human factors in computing systems, pages 210-217.",
                "links": null
            },
            "BIBREF36": {
                "ref_id": "b36",
                "title": "Neural cross-lingual entity linking",
                "authors": [
                    {
                        "first": "Avirup",
                        "middle": [],
                        "last": "Sil",
                        "suffix": ""
                    },
                    {
                        "first": "Gourab",
                        "middle": [],
                        "last": "Kundu",
                        "suffix": ""
                    },
                    {
                        "first": "Radu",
                        "middle": [],
                        "last": "Florian",
                        "suffix": ""
                    },
                    {
                        "first": "Wael",
                        "middle": [],
                        "last": "Hamza",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proceedings of the AAAI Conference on Artificial Intelligence",
                "volume": "32",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Avirup Sil, Gourab Kundu, Radu Florian, and Wael Hamza. 2018. Neural cross-lingual entity linking. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 32.",
                "links": null
            },
            "BIBREF37": {
                "ref_id": "b37",
                "title": "Two decades of recommender systems at amazon.com",
                "authors": [
                    {
                        "first": "Brent",
                        "middle": [],
                        "last": "Smith",
                        "suffix": ""
                    },
                    {
                        "first": "Greg",
                        "middle": [],
                        "last": "Linden",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "Ieee internet computing",
                "volume": "21",
                "issue": "3",
                "pages": "12--18",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Brent Smith and Greg Linden. 2017. Two decades of recommender systems at amazon.com. Ieee internet computing, 21(3):12-18.",
                "links": null
            },
            "BIBREF38": {
                "ref_id": "b38",
                "title": "Pathsim: Meta path-based top-k similarity search in heterogeneous information networks",
                "authors": [
                    {
                        "first": "Yizhou",
                        "middle": [],
                        "last": "Sun",
                        "suffix": ""
                    },
                    {
                        "first": "Jiawei",
                        "middle": [],
                        "last": "Han",
                        "suffix": ""
                    },
                    {
                        "first": "Xifeng",
                        "middle": [],
                        "last": "Yan",
                        "suffix": ""
                    },
                    {
                        "first": "S",
                        "middle": [],
                        "last": "Philip",
                        "suffix": ""
                    },
                    {
                        "first": "Tianyi",
                        "middle": [],
                        "last": "Yu",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Wu",
                        "suffix": ""
                    }
                ],
                "year": 2011,
                "venue": "Proceedings of the VLDB Endowment",
                "volume": "4",
                "issue": "",
                "pages": "992--1003",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Yizhou Sun, Jiawei Han, Xifeng Yan, Philip S Yu, and Tianyi Wu. 2011. Pathsim: Meta path-based top-k similarity search in heterogeneous information networks. Proceedings of the VLDB Endowment, 4(11):992-1003.",
                "links": null
            },
            "BIBREF39": {
                "ref_id": "b39",
                "title": "Rankclus: integrating clustering with ranking for heterogeneous information network analysis",
                "authors": [
                    {
                        "first": "Yizhou",
                        "middle": [],
                        "last": "Sun",
                        "suffix": ""
                    },
                    {
                        "first": "Jiawei",
                        "middle": [],
                        "last": "Han",
                        "suffix": ""
                    },
                    {
                        "first": "Peixiang",
                        "middle": [],
                        "last": "Zhao",
                        "suffix": ""
                    },
                    {
                        "first": "Zhijun",
                        "middle": [],
                        "last": "Yin",
                        "suffix": ""
                    },
                    {
                        "first": "Hong",
                        "middle": [],
                        "last": "Cheng",
                        "suffix": ""
                    },
                    {
                        "first": "Tianyi",
                        "middle": [],
                        "last": "Wu",
                        "suffix": ""
                    }
                ],
                "year": 2009,
                "venue": "Proceedings of the 12th International Conference on Extending Database Technology: Advances in Database Technology",
                "volume": "",
                "issue": "",
                "pages": "565--576",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Yizhou Sun, Jiawei Han, Peixiang Zhao, Zhijun Yin, Hong Cheng, and Tianyi Wu. 2009. Rankclus: in- tegrating clustering with ranking for heterogeneous information network analysis. In Proceedings of the 12th International Conference on Extending Database Technology: Advances in Database Tech- nology, pages 565-576.",
                "links": null
            },
            "BIBREF40": {
                "ref_id": "b40",
                "title": "RotatE: Knowledge graph embedding by relational rotation in complex space",
                "authors": [
                    {
                        "first": "Zhiqing",
                        "middle": [],
                        "last": "Sun",
                        "suffix": ""
                    },
                    {
                        "first": "Zhi-Hong",
                        "middle": [],
                        "last": "Deng",
                        "suffix": ""
                    },
                    {
                        "first": "Jian-Yun",
                        "middle": [],
                        "last": "Nie",
                        "suffix": ""
                    },
                    {
                        "first": "Jian",
                        "middle": [],
                        "last": "Tang",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "International Conference on Learning Representations",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Zhiqing Sun, Zhi-Hong Deng, Jian-Yun Nie, and Jian Tang. 2018. RotatE: Knowledge graph embedding by relational rotation in complex space. In Interna- tional Conference on Learning Representations.",
                "links": null
            },
            "BIBREF41": {
                "ref_id": "b41",
                "title": "Complex embeddings for simple link prediction",
                "authors": [
                    {
                        "first": "Th\u00e9o",
                        "middle": [],
                        "last": "Trouillon",
                        "suffix": ""
                    },
                    {
                        "first": "Johannes",
                        "middle": [],
                        "last": "Welbl",
                        "suffix": ""
                    },
                    {
                        "first": "Sebastian",
                        "middle": [],
                        "last": "Riedel",
                        "suffix": ""
                    },
                    {
                        "first": "\u00c9ric",
                        "middle": [],
                        "last": "Gaussier",
                        "suffix": ""
                    },
                    {
                        "first": "Guillaume",
                        "middle": [],
                        "last": "Bouchard",
                        "suffix": ""
                    }
                ],
                "year": 2016,
                "venue": "International Conference on Machine Learning (ICML)",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Th\u00e9o Trouillon, Johannes Welbl, Sebastian Riedel, \u00c9ric Gaussier, and Guillaume Bouchard. 2016. Complex embeddings for simple link prediction. International Conference on Machine Learning (ICML).",
                "links": null
            },
            "BIBREF42": {
                "ref_id": "b42",
                "title": "A content-based recommender system for computer science publications. Knowledge-Based Systems",
                "authors": [
                    {
                        "first": "Donghui",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    },
                    {
                        "first": "Yanchun",
                        "middle": [],
                        "last": "Liang",
                        "suffix": ""
                    },
                    {
                        "first": "Dong",
                        "middle": [],
                        "last": "Xu",
                        "suffix": ""
                    },
                    {
                        "first": "Xiaoyue",
                        "middle": [],
                        "last": "Feng",
                        "suffix": ""
                    },
                    {
                        "first": "Renchu",
                        "middle": [],
                        "last": "Guan",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "",
                "volume": "157",
                "issue": "",
                "pages": "1--9",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Donghui Wang, Yanchun Liang, Dong Xu, Xiaoyue Feng, and Renchu Guan. 2018. A content-based recommender system for computer science publica- tions. Knowledge-Based Systems, 157:1-9.",
                "links": null
            },
            "BIBREF43": {
                "ref_id": "b43",
                "title": "Language and domain independent entity linking with quantified collective validation",
                "authors": [
                    {
                        "first": "Han",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    },
                    {
                        "first": "Jin",
                        "middle": [
                            "Guang"
                        ],
                        "last": "Zheng",
                        "suffix": ""
                    },
                    {
                        "first": "Xiaogang",
                        "middle": [],
                        "last": "Ma",
                        "suffix": ""
                    },
                    {
                        "first": "Peter",
                        "middle": [],
                        "last": "Fox",
                        "suffix": ""
                    },
                    {
                        "first": "Heng",
                        "middle": [],
                        "last": "Ji",
                        "suffix": ""
                    }
                ],
                "year": 2015,
                "venue": "Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing",
                "volume": "",
                "issue": "",
                "pages": "695--704",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Han Wang, Jin Guang Zheng, Xiaogang Ma, Peter Fox, and Heng Ji. 2015. Language and domain indepen- dent entity linking with quantified collective valida- tion. In Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pages 695-704.",
                "links": null
            },
            "BIBREF44": {
                "ref_id": "b44",
                "title": "Knowledge-aware graph neural networks with label smoothness regularization for recommender systems",
                "authors": [
                    {
                        "first": "Hongwei",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    },
                    {
                        "first": "Fuzheng",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Mengdi",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Jure",
                        "middle": [],
                        "last": "Leskovec",
                        "suffix": ""
                    },
                    {
                        "first": "Miao",
                        "middle": [],
                        "last": "Zhao",
                        "suffix": ""
                    },
                    {
                        "first": "Wenjie",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Zhongyuan",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, KDD '19",
                "volume": "",
                "issue": "",
                "pages": "968--977",
                "other_ids": {
                    "DOI": [
                        "10.1145/3292500.3330836"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Hongwei Wang, Fuzheng Zhang, Mengdi Zhang, Jure Leskovec, Miao Zhao, Wenjie Li, and Zhongyuan Wang. 2019a. Knowledge-aware graph neural net- works with label smoothness regularization for rec- ommender systems. In Proceedings of the 25th ACM SIGKDD International Conference on Knowl- edge Discovery & Data Mining, KDD '19, page 968-977, New York, NY, USA. Association for Computing Machinery.",
                "links": null
            },
            "BIBREF45": {
                "ref_id": "b45",
                "title": "Multi-task feature learning for knowledge graph enhanced recommendation",
                "authors": [
                    {
                        "first": "Hongwei",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    },
                    {
                        "first": "Fuzheng",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Miao",
                        "middle": [],
                        "last": "Zhao",
                        "suffix": ""
                    },
                    {
                        "first": "Wenjie",
                        "middle": [],
                        "last": "Li",
                        "suffix": ""
                    },
                    {
                        "first": "Xing",
                        "middle": [],
                        "last": "Xie",
                        "suffix": ""
                    },
                    {
                        "first": "Minyi",
                        "middle": [],
                        "last": "Guo",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "The World Wide Web Conference",
                "volume": "",
                "issue": "",
                "pages": "2000--2010",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Hongwei Wang, Fuzheng Zhang, Miao Zhao, Wenjie Li, Xing Xie, and Minyi Guo. 2019b. Multi-task feature learning for knowledge graph enhanced rec- ommendation. In The World Wide Web Conference, pages 2000-2010.",
                "links": null
            },
            "BIBREF46": {
                "ref_id": "b46",
                "title": "Heterogeneous graph attention network",
                "authors": [
                    {
                        "first": "Xiao",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    },
                    {
                        "first": "Houye",
                        "middle": [],
                        "last": "Ji",
                        "suffix": ""
                    },
                    {
                        "first": "Chuan",
                        "middle": [],
                        "last": "Shi",
                        "suffix": ""
                    },
                    {
                        "first": "Bai",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    },
                    {
                        "first": "Yanfang",
                        "middle": [],
                        "last": "Ye",
                        "suffix": ""
                    },
                    {
                        "first": "Peng",
                        "middle": [],
                        "last": "Cui",
                        "suffix": ""
                    },
                    {
                        "first": "Philip S",
                        "middle": [],
                        "last": "Yu",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "The World Wide Web Conference",
                "volume": "",
                "issue": "",
                "pages": "2022--2032",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Xiao Wang, Houye Ji, Chuan Shi, Bai Wang, Yanfang Ye, Peng Cui, and Philip S Yu. 2019c. Heteroge- neous graph attention network. In The World Wide Web Conference, pages 2022-2032.",
                "links": null
            },
            "BIBREF47": {
                "ref_id": "b47",
                "title": "Scalable zeroshot entity linking with dense entity retrieval",
                "authors": [
                    {
                        "first": "Ledell",
                        "middle": [],
                        "last": "Wu",
                        "suffix": ""
                    },
                    {
                        "first": "Fabio",
                        "middle": [],
                        "last": "Petroni",
                        "suffix": ""
                    },
                    {
                        "first": "Martin",
                        "middle": [],
                        "last": "Josifoski",
                        "suffix": ""
                    },
                    {
                        "first": "Sebastian",
                        "middle": [],
                        "last": "Riedel",
                        "suffix": ""
                    },
                    {
                        "first": "Luke",
                        "middle": [],
                        "last": "Zettlemoyer",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:1911.03814"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Ledell Wu, Fabio Petroni, Martin Josifoski, Sebastian Riedel, and Luke Zettlemoyer. 2019. Scalable zero- shot entity linking with dense entity retrieval. arXiv preprint arXiv:1911.03814.",
                "links": null
            },
            "BIBREF48": {
                "ref_id": "b48",
                "title": "Deep matrix factorization models for recommender systems",
                "authors": [
                    {
                        "first": "Xinyu",
                        "middle": [],
                        "last": "Hong-Jian Xue",
                        "suffix": ""
                    },
                    {
                        "first": "Jianbing",
                        "middle": [],
                        "last": "Dai",
                        "suffix": ""
                    },
                    {
                        "first": "Shujian",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Jiajun",
                        "middle": [],
                        "last": "Huang",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Chen",
                        "suffix": ""
                    }
                ],
                "year": 2017,
                "venue": "IJCAI",
                "volume": "17",
                "issue": "",
                "pages": "3203--3209",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Hong-Jian Xue, Xinyu Dai, Jianbing Zhang, Shujian Huang, and Jiajun Chen. 2017. Deep matrix factor- ization models for recommender systems. In IJCAI, volume 17, pages 3203-3209. Melbourne, Australia.",
                "links": null
            },
            "BIBREF49": {
                "ref_id": "b49",
                "title": "Embedding entities and relations for learning and inference in knowledge bases",
                "authors": [
                    {
                        "first": "Bishan",
                        "middle": [],
                        "last": "Yang",
                        "suffix": ""
                    },
                    {
                        "first": "Wen-Tau",
                        "middle": [],
                        "last": "Yih",
                        "suffix": ""
                    },
                    {
                        "first": "Xiaodong",
                        "middle": [],
                        "last": "He",
                        "suffix": ""
                    },
                    {
                        "first": "Jianfeng",
                        "middle": [],
                        "last": "Gao",
                        "suffix": ""
                    },
                    {
                        "first": "Li",
                        "middle": [],
                        "last": "Deng",
                        "suffix": ""
                    }
                ],
                "year": 2014,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:1412.6575"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Bishan Yang, Wen-tau Yih, Xiaodong He, Jianfeng Gao, and Li Deng. 2014. Embedding entities and relations for learning and inference in knowledge bases. arXiv preprint arXiv:1412.6575.",
                "links": null
            },
            "BIBREF50": {
                "ref_id": "b50",
                "title": "Multi-modal bayesian embeddings for learning social knowledge graphs",
                "authors": [
                    {
                        "first": "Zhilin",
                        "middle": [],
                        "last": "Yang",
                        "suffix": ""
                    },
                    {
                        "first": "Jie",
                        "middle": [],
                        "last": "Tang",
                        "suffix": ""
                    },
                    {
                        "first": "William",
                        "middle": [],
                        "last": "Cohen",
                        "suffix": ""
                    }
                ],
                "year": 2015,
                "venue": "",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {
                    "arXiv": [
                        "arXiv:1508.00715"
                    ]
                },
                "num": null,
                "urls": [],
                "raw_text": "Zhilin Yang, Jie Tang, and William Cohen. 2015. Multi-modal bayesian embeddings for learn- ing social knowledge graphs. arXiv preprint arXiv:1508.00715.",
                "links": null
            },
            "BIBREF51": {
                "ref_id": "b51",
                "title": "User embedding for scholarly microblog recommendation",
                "authors": [
                    {
                        "first": "Yang",
                        "middle": [],
                        "last": "Yu",
                        "suffix": ""
                    },
                    {
                        "first": "Xiaojun",
                        "middle": [],
                        "last": "Wan",
                        "suffix": ""
                    },
                    {
                        "first": "Xinjie",
                        "middle": [],
                        "last": "Zhou",
                        "suffix": ""
                    }
                ],
                "year": 2016,
                "venue": "Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Yang Yu, Xiaojun Wan, and Xinjie Zhou. 2016. User embedding for scholarly microblog recommenda- tion. In Proceedings of the 54th Annual Meeting of the Association for Computational Linguistics (Vol- ume 2: Short Papers).",
                "links": null
            },
            "BIBREF52": {
                "ref_id": "b52",
                "title": "ELISA-EDL: A cross-lingual entity extraction, linking and localization system",
                "authors": [
                    {
                        "first": "Boliang",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Ying",
                        "middle": [],
                        "last": "Lin",
                        "suffix": ""
                    },
                    {
                        "first": "Xiaoman",
                        "middle": [],
                        "last": "Pan",
                        "suffix": ""
                    },
                    {
                        "first": "Di",
                        "middle": [],
                        "last": "Lu",
                        "suffix": ""
                    },
                    {
                        "first": "Jonathan",
                        "middle": [],
                        "last": "May",
                        "suffix": ""
                    },
                    {
                        "first": "Kevin",
                        "middle": [],
                        "last": "Knight",
                        "suffix": ""
                    },
                    {
                        "first": "Heng",
                        "middle": [],
                        "last": "Ji",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "Proc. The 16th Annual Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (NAACL-HLT2018)",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Boliang Zhang, Ying Lin, Xiaoman Pan, Di Lu, Jonathan May, Kevin Knight, and Heng Ji. 2018a. ELISA-EDL: A cross-lingual entity extraction, link- ing and localization system. In Proc. The 16th An- nual Conference of the North American Chapter of the Association for Computational Linguistics: Hu- man Language Technologies (NAACL-HLT2018).",
                "links": null
            },
            "BIBREF53": {
                "ref_id": "b53",
                "title": "Heterogeneous graph neural network",
                "authors": [
                    {
                        "first": "Chuxu",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Dongjin",
                        "middle": [],
                        "last": "Song",
                        "suffix": ""
                    },
                    {
                        "first": "Chao",
                        "middle": [],
                        "last": "Huang",
                        "suffix": ""
                    },
                    {
                        "first": "Ananthram",
                        "middle": [],
                        "last": "Swami",
                        "suffix": ""
                    },
                    {
                        "first": "Nitesh V",
                        "middle": [],
                        "last": "Chawla",
                        "suffix": ""
                    }
                ],
                "year": 2019,
                "venue": "Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining",
                "volume": "",
                "issue": "",
                "pages": "793--803",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Chuxu Zhang, Dongjin Song, Chao Huang, Ananthram Swami, and Nitesh V Chawla. 2019. Heterogeneous graph neural network. In Proceedings of the 25th ACM SIGKDD International Conference on Knowl- edge Discovery & Data Mining, pages 793-803.",
                "links": null
            },
            "BIBREF54": {
                "ref_id": "b54",
                "title": "Anrl: Attributed network representation learning via deep neural networks",
                "authors": [
                    {
                        "first": "Zhen",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Hongxia",
                        "middle": [],
                        "last": "Yang",
                        "suffix": ""
                    },
                    {
                        "first": "Jiajun",
                        "middle": [],
                        "last": "Bu",
                        "suffix": ""
                    },
                    {
                        "first": "Sheng",
                        "middle": [],
                        "last": "Zhou",
                        "suffix": ""
                    },
                    {
                        "first": "Pinggang",
                        "middle": [],
                        "last": "Yu",
                        "suffix": ""
                    },
                    {
                        "first": "Jianwei",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Martin",
                        "middle": [],
                        "last": "Ester",
                        "suffix": ""
                    },
                    {
                        "first": "Can",
                        "middle": [],
                        "last": "Wang",
                        "suffix": ""
                    }
                ],
                "year": 2018,
                "venue": "In IJCAI",
                "volume": "18",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Zhen Zhang, Hongxia Yang, Jiajun Bu, Sheng Zhou, Pinggang Yu, Jianwei Zhang, Martin Ester, and Can Wang. 2018b. Anrl: Attributed network represen- tation learning via deep neural networks. In IJCAI, volume 18.",
                "links": null
            },
            "BIBREF55": {
                "ref_id": "b55",
                "title": "Userbased collaborative-filtering recommendation algorithms on hadoop",
                "authors": [
                    {
                        "first": "Dan",
                        "middle": [],
                        "last": "Zhi",
                        "suffix": ""
                    },
                    {
                        "first": "Ming-Sheng",
                        "middle": [],
                        "last": "Zhao",
                        "suffix": ""
                    },
                    {
                        "first": "",
                        "middle": [],
                        "last": "Shang",
                        "suffix": ""
                    }
                ],
                "year": 2010,
                "venue": "2010 Third International Conference on Knowledge Discovery and Data Mining",
                "volume": "",
                "issue": "",
                "pages": "478--481",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Zhi-Dan Zhao and Ming-Sheng Shang. 2010. User- based collaborative-filtering recommendation algo- rithms on hadoop. In 2010 Third International Con- ference on Knowledge Discovery and Data Mining, pages 478-481. IEEE.",
                "links": null
            },
            "BIBREF56": {
                "ref_id": "b56",
                "title": "Entity linking for biomedical literature",
                "authors": [
                    {
                        "first": "Jin",
                        "middle": [
                            "Guang"
                        ],
                        "last": "Zheng",
                        "suffix": ""
                    },
                    {
                        "first": "Daniel",
                        "middle": [],
                        "last": "Howsmon",
                        "suffix": ""
                    },
                    {
                        "first": "Boliang",
                        "middle": [],
                        "last": "Zhang",
                        "suffix": ""
                    },
                    {
                        "first": "Juergen",
                        "middle": [],
                        "last": "Hahn",
                        "suffix": ""
                    },
                    {
                        "first": "Deborah",
                        "middle": [],
                        "last": "Mcguinness",
                        "suffix": ""
                    },
                    {
                        "first": "James",
                        "middle": [],
                        "last": "Hendler",
                        "suffix": ""
                    },
                    {
                        "first": "Heng",
                        "middle": [],
                        "last": "Ji",
                        "suffix": ""
                    }
                ],
                "year": 2014,
                "venue": "BMC Medical Informatics and Decision Making",
                "volume": "",
                "issue": "",
                "pages": "",
                "other_ids": {},
                "num": null,
                "urls": [],
                "raw_text": "Jin Guang Zheng, Daniel Howsmon, Boliang Zhang, Juergen Hahn, Deborah McGuinness, James Hendler, and Heng Ji. 2014. Entity linking for biomedical literature. In BMC Medical Informatics and Decision Making.",
                "links": null
            }
        },
        "ref_entries": {
            "FIGREF1": {
                "text": "An illustration of the cross-source heterogeneous customer-product graph.",
                "uris": null,
                "type_str": "figure",
                "num": null
            },
            "FIGREF2": {
                "text": "An illustration of our framework.",
                "uris": null,
                "type_str": "figure",
                "num": null
            },
            "FIGREF3": {
                "text": "Distribution of remaining Errors.",
                "uris": null,
                "type_str": "figure",
                "num": null
            },
            "TABREF0": {
                "text": "",
                "html": null,
                "content": "<table><tr><td>Query Embedding</td><td>History Encoder</td><td/></tr><tr><td>Query Encoder</td><td/><td>Candidate Encoder</td></tr><tr><td>Query</td><td>Customer</td><td>Candidate</td></tr></table>",
                "type_str": "table",
                "num": null
            },
            "TABREF1": {
                "text": "Relative gains compared to QUARTS. (%)",
                "html": null,
                "content": "<table><tr><td>Method</td><td>Dev Acc</td><td>Test Acc</td></tr><tr><td>4 Our Model</td><td>+32.9</td><td>+24.6</td></tr><tr><td>5 w/o Ranking</td><td>-17.1</td><td>-20.4</td></tr><tr><td>6 w/o Personalized Features</td><td>-10.5</td><td>-18.0</td></tr><tr><td>7 w/o Product Embedding</td><td>+25.2</td><td>+19.0</td></tr><tr><td>8 w/o Joint Embedding</td><td>+28.1</td><td>+20.4</td></tr></table>",
                "type_str": "table",
                "num": null
            },
            "TABREF2": {
                "text": "Ablation study. (%, relative gains compared to QUARTS.) ] instanatural vitamin c serum with hyaluronic acid & vit e -natural & organic anti wrinkle ... * foundation makeup brush flat top kabuki for faceperfect for blending liquid, cream or flawless powder * [1] truskin vitamin c serum for face, topical facial serum with hyaluronic acid, vitamin e, 1 fl oz * women's rogaine 5% minoxidil foam for hair thinning and loss, topical treatment for women's hair ... * [2] vitamin c serum for face -anti aging facial serum * vita liberata advanced organics fabulous self-tanning gradual lotion with marula oil, 6.76 fl oz * [4] vitamin c serum plus 2% retinol, 3.5% niacinamide, 5% hyaluronic acid, 2% salicylic acid ...",
                "html": null,
                "content": "<table><tr><td>Query</td><td>Candidates</td><td>History</td></tr><tr><td>#1 vitamin c serum</td><td colspan=\"2\">* [3* instanatural vitamin c serum with hyaluronic acid &amp;</td></tr><tr><td/><td/><td>vit e -natural &amp; organic anti wrinkle reducer ...</td></tr><tr><td/><td colspan=\"2\">Our model promotes candidate 3 as this product was purchased by the customer.</td></tr><tr><td colspan=\"2\">#2 toothpaste * [2] crest 3d white whitening toothpaste, radiant mint,</td><td>* crest 3d white toothpaste radiant mint (3 count of 4.1</td></tr><tr><td/><td>3.5oz, twin pack</td><td>oz tubes), 12.3 oz packaging may vary</td></tr><tr><td/><td>* [1] crest + scope complete whitening toothpaste,</td><td>* skindinavia the makeup of countrol finishing spray,</td></tr><tr><td/><td>minty fresh, 5.4 oz, pack of 3</td><td>8 fluid ounce</td></tr><tr><td/><td>* [3] pronamel gentle whitening enamel toothpaste for</td><td>* crest 3d white toothpaste radiant mint (3 count of 4.1</td></tr><tr><td/><td>sensitive teeth, alpine breeze-4 ounces (pack of 3)</td><td>oz tubes), 12.3 oz packaging may vary</td></tr><tr><td/><td>* [4] colgate cavity protection toothpaste with fluoride</td><td>* nivea shea daily mointure body lotion -48 hour</td></tr><tr><td/><td>-6 ounce (pack of 6)</td><td>moisture for dry skin -16.9 fl. oz. pump bottle, ...</td></tr><tr><td/><td colspan=\"2\">Although the previously purchased item is no longer available, with entity embedding learned from the</td></tr><tr><td/><td colspan=\"2\">cross-source graph, our model successfully promotes the most similar product.</td></tr><tr><td>#3 sun dried</td><td>* [3] 365 everyday value, organic sundried tomatoes</td><td>* #1 usda organic aloe vera gel -no preservatives, no</td></tr><tr><td>tomatoes</td><td>in extra virgin olive oil, 8.5 oz</td><td>alcohol -from freshly cut usa grown 100% pure ...</td></tr><tr><td/><td>* [1] 35 oz bella sun luci sun dried tomatoes julienne</td><td>* organic aloe vera gel with 100% pure aloe from</td></tr><tr><td/><td>cut in olive oil (original version)</td><td>freshly cut aloe plant, not powder -no xanthan ...</td></tr><tr><td/><td colspan=\"2\">* [2] julienne sun-dried tomatoes -16oz bag (kosher) * wicked joe organic coffee wicked italian ground</td></tr><tr><td/><td>* [4] organic sun-dried tomatoes with sea salt, 8</td><td>*thayers alcohol-free original witch hazel facial toner</td></tr><tr><td/><td>ounces -salted, non-gmo, kosher, raw, vegan, ...</td><td>with aloe vera formula, clear, 12oz</td></tr></table>",
                "type_str": "table",
                "num": null
            },
            "TABREF4": {
                "text": "Negative examples in the data set.",
                "html": null,
                "content": "<table/>",
                "type_str": "table",
                "num": null
            }
        }
    }
}