{"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/cotype-joint-extraction-of-typed-entities-and","title":"CoType: Joint Extraction of Typed Entities and Relations with Knowledge Bases","arxiv_id":"1610.08763","date":"2016-10-27","proceeding":null,"authors":["Xiang Ren","Zeqiu Wu","Wenqi He","Meng Qu","Clare R. Voss","Heng Ji","Tarek F. Abdelzaher","Jiawei Han"],"abstract":"Extracting entities and relations for types of interest from text is\nimportant for understanding massive text corpora. Traditionally, systems of\nentity relation extraction have relied on human-annotated corpora for training\nand adopted an incremental pipeline. Such systems require additional human\nexpertise to be ported to a new domain, and are vulnerable to errors cascading\ndown the pipeline. In this paper, we investigate joint extraction of typed\nentities and relations with labeled data heuristically obtained from knowledge\nbases (i.e., distant supervision). As our algorithm for type labeling via\ndistant supervision is context-agnostic, noisy training data poses unique\nchallenges for the task. We propose a novel domain-independent framework,\ncalled CoType, that runs a data-driven text segmentation algorithm to extract\nentity mentions, and jointly embeds entity mentions, relation mentions, text\nfeatures and type labels into two low-dimensional spaces (for entity and\nrelation mentions respectively), where, in each space, objects whose types are\nclose will also have similar representations. CoType, then using these learned\nembeddings, estimates the types of test (unlinkable) mentions. We formulate a\njoint optimization problem to learn embeddings from text corpora and knowledge\nbases, adopting a novel partial-label loss function for noisy labeled data and\nintroducing an object \"translation\" function to capture the cross-constraints\nof entities and relations on each other. Experiments on three public datasets\ndemonstrate the effectiveness of CoType across different domains (e.g., news,\nbiomedical), with an average of 25% improvement in F1 score compared to the\nnext best method.","url_abs":"http://arxiv.org/abs/1610.08763v2","url_pdf":"http://arxiv.org/pdf/1610.08763v2.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":"cotype-joint-extraction-of-typed-entities-and","repo_url":"https://github.com/shanzhenren/CoType","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"cotype-joint-extraction-of-typed-entities-and","repo_url":"https://github.com/ellenmellon/ReQuest","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null},{"paper_slug":"cotype-joint-extraction-of-typed-entities-and","repo_url":"https://github.com/ink-usc/request","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"joint-entity-and-relation-extraction","task_name":"Joint Entity and Relation Extraction"},{"task_slug":null,"task_name":"Relation"},{"task_slug":"relation-extraction","task_name":"Relation Extraction"},{"task_slug":"text-segmentation","task_name":"Text Segmentation"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/relation-extraction-on-nyt11-hrl","task":"Relation Extraction","dataset":"NYT11-HRL","model":"Cotype","rank_in_archive_order":11,"of":12,"metrics":{"F1":"43"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1610.08763","atlas_url":"https://app.syntology.ai/?focus=1610.08763","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}