{"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/wic-10000-example-pairs-for-evaluating","title":"WiC: the Word-in-Context Dataset for Evaluating Context-Sensitive Meaning Representations","arxiv_id":"1808.09121","date":"2018-08-28","proceeding":"NAACL 2019 6","authors":["Mohammad Taher Pilehvar","Jose Camacho-Collados"],"abstract":"By design, word embeddings are unable to model the dynamic nature of words'\nsemantics, i.e., the property of words to correspond to potentially different\nmeanings. To address this limitation, dozens of specialized meaning\nrepresentation techniques such as sense or contextualized embeddings have been\nproposed. However, despite the popularity of research on this topic, very few\nevaluation benchmarks exist that specifically focus on the dynamic semantics of\nwords. In this paper we show that existing models have surpassed the\nperformance ceiling of the standard evaluation dataset for the purpose, i.e.,\nStanford Contextual Word Similarity, and highlight its shortcomings. To address\nthe lack of a suitable benchmark, we put forward a large-scale Word in Context\ndataset, called WiC, based on annotations curated by experts, for generic\nevaluation of context-sensitive representations. WiC is released in\nhttps://pilehvar.github.io/wic/.","url_abs":"http://arxiv.org/abs/1808.09121v3","url_pdf":"http://arxiv.org/pdf/1808.09121v3.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":[],"tasks":[{"task_slug":"word-embeddings","task_name":"Word Embeddings"},{"task_slug":"word-sense-disambiguation","task_name":"Word Sense Disambiguation"},{"task_slug":"word-similarity","task_name":"Word Similarity"}],"methods":[],"datasets_introduced":[{"slug":"wic","name":"WiC","full_name":"Words in Context"}],"methods_introduced":[],"results":[{"leaderboard":"/sota/word-sense-disambiguation-on-words-in-context","task":"Word Sense Disambiguation","dataset":"Words in Context","model":"BERT-large 340M","rank_in_archive_order":14,"of":37,"metrics":{"Accuracy":"65.5"},"uses_additional_data":false},{"leaderboard":"/sota/word-sense-disambiguation-on-words-in-context","task":"Word Sense Disambiguation","dataset":"Words in Context","model":"Context2vec","rank_in_archive_order":17,"of":37,"metrics":{"Accuracy":"59.3"},"uses_additional_data":false},{"leaderboard":"/sota/word-sense-disambiguation-on-words-in-context","task":"Word Sense Disambiguation","dataset":"Words in Context","model":"DeConf","rank_in_archive_order":18,"of":37,"metrics":{"Accuracy":"58.7"},"uses_additional_data":false},{"leaderboard":"/sota/word-sense-disambiguation-on-words-in-context","task":"Word Sense Disambiguation","dataset":"Words in Context","model":"SW2V","rank_in_archive_order":19,"of":37,"metrics":{"Accuracy":"58.1"},"uses_additional_data":false},{"leaderboard":"/sota/word-sense-disambiguation-on-words-in-context","task":"Word Sense Disambiguation","dataset":"Words in Context","model":"ElMo","rank_in_archive_order":20,"of":37,"metrics":{"Accuracy":"57.7"},"uses_additional_data":false},{"leaderboard":"/sota/word-sense-disambiguation-on-words-in-context","task":"Word Sense Disambiguation","dataset":"Words in Context","model":"Sentence LSTM","rank_in_archive_order":24,"of":37,"metrics":{"Accuracy":"53.1"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1808.09121","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}