{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/imposing-label-relational-inductive-bias-for","title":"Imposing Label-Relational Inductive Bias for Extremely Fine-Grained Entity Typing","arxiv_id":"1903.02591","date":"2019-03-06","proceeding":"NAACL 2019 6","authors":["Wenhan Xiong","Jiawei Wu","Deren Lei","Mo Yu","Shiyu Chang","Xiaoxiao Guo","William Yang Wang"],"abstract":"Existing entity typing systems usually exploit the type hierarchy provided by\nknowledge base (KB) schema to model label correlations and thus improve the\noverall performance. Such techniques, however, are not directly applicable to\nmore open and practical scenarios where the type set is not restricted by KB\nschema and includes a vast number of free-form types. To model the underly-ing\nlabel correlations without access to manually annotated label structures, we\nintroduce a novel label-relational inductive bias, represented by a graph\npropagation layer that effectively encodes both global label co-occurrence\nstatistics and word-level similarities.On a large dataset with over 10,000\nfree-form types, the graph-enhanced model equipped with an attention-based\nmatching module is able to achieve a much higher recall score while maintaining\na high-level precision. Specifically, it achieves a 15.3% relative F1\nimprovement and also less inconsistency in the outputs. We further show that a\nsimple modification of our proposed graph layer can also improve the\nperformance on a conventional and widely-tested dataset that only includes\nKB-schema types.","url_abs":"http://arxiv.org/abs/1903.02591v1","url_pdf":"http://arxiv.org/pdf/1903.02591v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"imposing-label-relational-inductive-bias-for","repo_url":"https://github.com/xwhan/Extremely-Fine-Grained-Entity-Typing","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"entity-typing","task_name":"Entity Typing"},{"task_slug":"inductive-bias","task_name":"Inductive Bias"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/entity-typing-on-ontonotes-v5-english","task":"Entity Typing","dataset":"Ontonotes v5 (English)","model":"LabelGCN Xiong et al. (2019)","rank_in_archive_order":3,"of":4,"metrics":{"F1":"36.9","Precision":"50.3","Recall":"29.2"},"uses_additional_data":false},{"leaderboard":"/sota/entity-typing-on-open-entity-1","task":"Entity Typing","dataset":"Open Entity","model":"LabelGCN","rank_in_archive_order":12,"of":13,"metrics":{"F1":"36.9"},"uses_additional_data":false}],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=1903.02591","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}