{"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/deeptype-multilingual-entity-linking-by","title":"DeepType: Multilingual Entity Linking by Neural Type System Evolution","arxiv_id":"1802.01021","date":"2018-02-03","proceeding":null,"authors":["Jonathan Raiman","Olivier Raiman"],"abstract":"The wealth of structured (e.g. Wikidata) and unstructured data about the\nworld available today presents an incredible opportunity for tomorrow's\nArtificial Intelligence. So far, integration of these two different modalities\nis a difficult process, involving many decisions concerning how best to\nrepresent the information so that it will be captured or useful, and\nhand-labeling large amounts of data. DeepType overcomes this challenge by\nexplicitly integrating symbolic information into the reasoning process of a\nneural network with a type system. First we construct a type system, and\nsecond, we use it to constrain the outputs of a neural network to respect the\nsymbolic structure. We achieve this by reformulating the design problem into a\nmixed integer problem: create a type system and subsequently train a neural\nnetwork with it. In this reformulation discrete variables select which\nparent-child relations from an ontology are types within the type system, while\ncontinuous variables control a classifier fit to the type system. The original\nproblem cannot be solved exactly, so we propose a 2-step algorithm: 1)\nheuristic search or stochastic optimization over discrete variables that define\na type system informed by an Oracle and a Learnability heuristic, 2) gradient\ndescent to fit classifier parameters. We apply DeepType to the problem of\nEntity Linking on three standard datasets (i.e. WikiDisamb30, CoNLL (YAGO), TAC\nKBP 2010) and find that it outperforms all existing solutions by a wide margin,\nincluding approaches that rely on a human-designed type system or recent deep\nlearning-based entity embeddings, while explicitly using symbolic information\nlets it integrate new entities without retraining.","url_abs":"http://arxiv.org/abs/1802.01021v1","url_pdf":"http://arxiv.org/pdf/1802.01021v1.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":"deeptype-multilingual-entity-linking-by","repo_url":"https://github.com/openai/deeptype","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"entity-disambiguation","task_name":"Entity Disambiguation"},{"task_slug":"entity-embeddings","task_name":"Entity Embeddings"},{"task_slug":"entity-linking","task_name":"Entity Linking"},{"task_slug":"heuristic-search","task_name":"Heuristic Search"},{"task_slug":"stochastic-optimization","task_name":"Stochastic Optimization"},{"task_slug":"type","task_name":"Vocal Bursts Type Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/entity-disambiguation-on-aida-conll","task":"Entity Disambiguation","dataset":"AIDA-CoNLL","model":"DeepType","rank_in_archive_order":3,"of":20,"metrics":{"In-KB Accuracy":"94.88"},"uses_additional_data":false},{"leaderboard":"/sota/entity-disambiguation-on-tac2010","task":"Entity Disambiguation","dataset":"TAC2010","model":"DeepType","rank_in_archive_order":1,"of":4,"metrics":{"Micro Precision":"90.85"},"uses_additional_data":false},{"leaderboard":"/sota/entity-linking-on-conll-aida","task":"Entity Linking","dataset":"CoNLL-Aida","model":"Raiman & Raiman 2018","rank_in_archive_order":2,"of":3,"metrics":{"Accuracy":"94.9"},"uses_additional_data":false},{"leaderboard":"/sota/entity-linking-on-tac-kbp-2010","task":"Entity Linking","dataset":"TAC-KBP 2010","model":"Raiman & Raiman 2018","rank_in_archive_order":1,"of":3,"metrics":{"Accuracy":"90.9"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1802.01021","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1802.01021"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/openai/deeptype","reach":null}],"summary":{"ran_honours":2},"by_repo_kind":{"official":{"samples":2,"ran":2,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":2,"samples":[{"code_sha256_prefix":"98dfc6c9fb866028","entry":"wkd","repo":"openai/deeptype","repo_kind":"official","path":"extraction/classifiers/type_classifier.py","file_url":"https://github.com/openai/deeptype/blob/HEAD/extraction/classifiers/type_classifier.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"98dfc6c9fb866028"}},{"code_sha256_prefix":"f1229bc17f655a23","entry":"wkp","repo":"openai/deeptype","repo_kind":"official","path":"extraction/classifiers/type_classifier.py","file_url":"https://github.com/openai/deeptype/blob/HEAD/extraction/classifiers/type_classifier.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"f1229bc17f655a23"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}