{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,4]],"date-time":"2026-06-04T05:23:36Z","timestamp":1780550616440,"version":"3.54.1"},"reference-count":33,"publisher":"China Science Publishing & Media Ltd.","issue":"1","license":[{"start":{"date-parts":[[2022,1,20]],"date-time":"2022-01-20T00:00:00Z","timestamp":1642636800000},"content-version":"vor","delay-in-days":19,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2022,2,3]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>Entity Linking (EL) aims to automatically link the mentions in unstructured documents to corresponding entities in a knowledge base (KB), which has recently been dominated by global models. Although many global EL methods attempt to model the topical coherence among all linked entities, most of them failed in exploiting the correlations among manifold knowledge helpful for linking, such as the semantics of mentions and their candidates, the neighborhood information of candidate entities in KB and the fine-grained type information of entities. As we will show in the paper, interactions among these types of information are very useful for better characterizing the topic features of entities and more accurately estimating the topical coherence among all the referred entities within the same document. In this paper, we present a novel HEterogeneous Graph-based Entity Linker (HEGEL) for global entity linking, which builds an informative heterogeneous graph for every document to collect various linking clues. Then HEGEL utilizes a novel heterogeneous graph neural network (HGNN) to integrate the different types of manifold information and model the interactions among them. Experiments on the standard benchmark datasets demonstrate that HEGEL can well capture the global coherence and outperforms the prior state-of-the-art EL methods.<\/jats:p>","DOI":"10.1162\/dint_a_00116","type":"journal-article","created":{"date-parts":[[2022,1,20]],"date-time":"2022-01-20T20:35:39Z","timestamp":1642710939000},"page":"20-40","update-policy":"https:\/\/doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":10,"title":["Integrating Manifold Knowledge for Global Entity Linking with\n                    Heterogeneous Graphs"],"prefix":"10.3724","volume":"4","author":[{"given":"Zhibin","family":"Chen","sequence":"first","affiliation":[{"name":"Wangxuan Institute of Computer Technology, Peking University, Beijing 100871, China"},{"name":"Center for Data Science, Peking University, Beijing 100871, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yuting","family":"Wu","sequence":"additional","affiliation":[{"name":"Wangxuan Institute of Computer Technology, Peking University, Beijing 100871, China"},{"name":"The MOE Key Laboratory of Computational Linguistics, Peking University, Beijing 100871, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Yansong","family":"Feng","sequence":"additional","affiliation":[{"name":"Wangxuan Institute of Computer Technology, Peking University, Beijing 100871, China"},{"name":"The MOE Key Laboratory of Computational Linguistics, Peking University, Beijing 100871, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Dongyan","family":"Zhao","sequence":"additional","affiliation":[{"name":"Wangxuan Institute of Computer Technology, Peking University, Beijing 100871, China"},{"name":"The MOE Key Laboratory of Computational Linguistics, Peking University, Beijing 100871, China"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"2026","published-online":{"date-parts":[[2022,2,3]]},"reference":[{"key":"2022020307030259100_ref1","first-page":"1321","volume-title":"Semantic parsing via staged query graph generation: Question\n                        answering with knowledge base","author":"Yih","year":"2015"},{"key":"2022020307030259100_ref2","first-page":"2915","volume-title":"Combining knowledge with deep convolutional neural networks for\n                        short text classification","author":"Wang","year":"2017"},{"key":"2022020307030259100_ref3","first-page":"541","volume-title":"Knowledge-based weak supervision for information extraction of\n                        overlapping relations","author":"Hoffmann","year":"2011"},{"key":"2022020307030259100_ref4","doi-asserted-by":"crossref","DOI":"10.18653\/v1\/D18-1360","volume-title":"Multi-task identification of entities, relations, and coreference\n                        for scientific knowledge graph construction","author":"Luan","year":"2018"},{"key":"2022020307030259100_ref5","volume-title":"Neural collective entity linking","author":"Cao","year":"2018"},{"key":"2022020307030259100_ref6","first-page":"438","volume-title":"Joint entity linking with deep reinforcement learning","author":"Fang","year":"2019"},{"key":"2022020307030259100_ref7","first-page":"640","volume-title":"High quality candidate generation and sequential graph attention\n                        network for entity linking","author":"Fang","year":"2020"},{"key":"2022020307030259100_ref8","first-page":"2619","volume-title":"Deep joint entity disambiguation with local neural\n                        attention","author":"Ganea","year":"2017"},{"key":"2022020307030259100_ref9","first-page":"1","article-title":"Robust named entity disambiguation with random\n                        walks","volume":"9","author":"Guo","year":"2017","journal-title":"Semantic Web"},{"key":"2022020307030259100_ref10","first-page":"4132","volume-title":"An attention-based graph neural network for heterogeneous structural\n                        learning","author":"Hong","year":"2020"},{"key":"2022020307030259100_ref11","first-page":"1595","volume-title":"Improving entity linking by modeling latent relations between\n                        mentions","author":"Le","year":"2018"},{"issue":"2","key":"2022020307030259100_ref12","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1145\/3300197","article-title":"A multi-view-based collective entity linking\n                        method","volume":"37","author":"Liu","year":"2019","journal-title":"ACM Transactions on Information\n                        Systems"},{"key":"2022020307030259100_ref13","first-page":"1375","volume-title":"Local and global algorithms for disambiguation to\n                    Wikipedia","author":"Ratinov","year":"2011"},{"key":"2022020307030259100_ref14","first-page":"1149","volume-title":"Dynamic graph convolutional networks for entity linking","author":"Wu","year":"2020"},{"key":"2022020307030259100_ref15","first-page":"16","volume-title":"Neural fine-grained entity type classification with hierarchy-aware\n                        loss","author":"Xu","year":"2018"},{"key":"2022020307030259100_ref16","doi-asserted-by":"crossref","first-page":"105620","DOI":"10.1016\/j.knosys.2020.105620","article-title":"Graph neural entity disambiguation","volume":"195","author":"Hu","year":"2020","journal-title":"Knowledge-Based Systems"},{"key":"2022020307030259100_ref17","volume-title":"Semi-supervised classification with graph convolutional\n                        networks","author":"Kipf","year":"2016"},{"key":"2022020307030259100_ref18","first-page":"1787","volume-title":"Relational inference for Wikification","author":"Cheng","year":"2013"},{"key":"2022020307030259100_ref19","first-page":"782","volume-title":"Robust disambiguation of named entities in text","author":"Hoffart","year":"2011"},{"key":"2022020307030259100_ref20","first-page":"2157","volume-title":"Evaluating the impact of knowledge graph context on entity\n                        disambiguation models","author":"Mulang","year":"2020"},{"key":"2022020307030259100_ref21","first-page":"5406","volume-title":"Deeptype: Multilingual entity linking by neural type system\n                        evolution","author":"Raiman","year":"2018"},{"key":"2022020307030259100_ref22","volume-title":"Zero-shot entity linking with dense entity retrieval","author":"Wu","year":"2019"},{"key":"2022020307030259100_ref23","volume-title":"Multilingual autoregressive entity linking","author":"Cao","year":"2021"},{"key":"2022020307030259100_ref24","first-page":"1025","volume-title":"Inductive representation learning on large graphs","author":"Hamilton","year":"2017"},{"key":"2022020307030259100_ref25","first-page":"1","volume-title":"Graph attention networks","author":"Veli\u010dkovi\u0107","year":"2018"},{"key":"2022020307030259100_ref26","first-page":"135","volume-title":"metapath2vec: Scalable representation learning for heterogeneous\n                        networks","author":"Dong","year":"2017"},{"key":"2022020307030259100_ref27","first-page":"593","volume-title":"Robust similarity measures for named entities matching","author":"Moreau","year":"2008"},{"key":"2022020307030259100_ref28","volume-title":"Gaussian error linear units (gelus)","author":"Hendrycks","year":"2016"},{"key":"2022020307030259100_ref29","first-page":"708","volume-title":"Large-scale named entity disambiguation based on Wikipedia\n                        data","author":"Cucerzan","year":"2007"},{"key":"2022020307030259100_ref30","first-page":"509","volume-title":"Learning to link with Wikipedia","author":"Milne","year":"2008"},{"key":"2022020307030259100_ref31","first-page":"1188","volume-title":"Distributed representations of sentences and documents","author":"Le","year":"2014"},{"key":"2022020307030259100_ref32","first-page":"1725","volume-title":"Simple and deep graph convolutional networks","author":"Chen","year":"2020"},{"key":"2022020307030259100_ref33","first-page":"770","volume-title":"Deep residual learning for image recognition","author":"He","year":"2016"}],"container-title":["Data Intelligence"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/direct.mit.edu\/dint\/article-pdf\/4\/1\/20\/1985039\/dint_a_00116.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/direct.mit.edu\/dint\/article-pdf\/4\/1\/20\/1985039\/dint_a_00116.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,3,14]],"date-time":"2025-03-14T07:42:47Z","timestamp":1741938167000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.sciengine.com\/doi\/10.1162\/dint_a_00116"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2022]]},"references-count":33,"journal-issue":{"issue":"1","published-print":{"date-parts":[[2022,2,3]]}},"URL":"https:\/\/doi.org\/10.1162\/dint_a_00116","relation":{},"ISSN":["2641-435X"],"issn-type":[{"value":"2641-435X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2022]]},"published":{"date-parts":[[2022]]}}}