Extracting Glosses to Disambiguate Word Senses

Like most natural language disambiguation tasks, word sense disambiguation (WSD) requires world knowledge for accurate predictions. Several proxies for this knowledge have been investigated, including labeled corpora, user-contributed knowledge, and machine readable dictionaries, but each of these proxies requires significant manual effort to create, and they do not cover all of the ambiguous terms in a language. We investigate the task of automatically extracting world knowledge, in the form of glosses, from an unlabeled corpus. We demonstrate how to use these glosses to automatically label a training corpus to build a statistical WSD system that uses no manually-labeled data, with experimental results approaching that of a supervised SVMbased classifier.

• Type: Journal
• Year: 2010
• Where: NAACL
• Authors: Weisi Duan, Alexander Yates
• Keywords: Computational Linguistics, Language Disambiguation Tasks, NAACL 2010, Word Sense Disambiguation, World Knowledge