Knowledge-based WSD on specific domains: performing better than generic supervised WSD

摘要: This paper explores the application of knowledge-based Word Sense Disambiguation systems to specific domains, based on our state-of-the-art graph-based WSD system that uses information in WordNet. Evaluation was performed over a publicly available domain-specific dataset 41 words related Sports and Finance, comprising examples drawn from three corpora: one balanced corpus (BNC), two corpora (news Finance). The results show all algorithm improves previous results, also supervised trained SemCor, largest annotated corpus. We using as context, instead actual occurrence contexts, yields better domain datasets, but not general one. Interestingly, are higher for than corpus, raising prospects improving current when applied domains.