作者: Stefan Thater , Georgiana Dinu , Soeren Laue
DOI:
关键词:
摘要: This paper compares a number of recently proposed models for computing context sensitive word similarity. We clarify the connections between these models, simplify their formulation and evaluate them in unified setting. show that are essentially equivalent if syntactic information is ignored, substantial performance differences previously reported disappear to large extent when simplified variants evaluated under identical conditions. Furthermore, our reformulation allows design straightforward fast implementation.