Word Sense Discrimination by Clustering Contexts in Vector and Similarity Spaces

摘要: This paper systematically compares unsupervised word sense discrimination techniques that cluster instances of a target occur in raw text using both vector and similarity spaces. The context each instance is represented as high dimensional feature space. Discrimination achieved by clustering these vectors directly space also finding pairwise similarities among the then We employ two different representations which occurs. First order represent features context. Second are an indirect representation based on average words evaluate discriminated clusters carrying out experiments sense–tagged 24 SENSEVAL2 well known Line, Hard Serve corpora.