A Survey of Word-level Semantic Relatedness Computation based on Distribution
Sun Shuq · Intelligent Computer and Applications · 2014
In the domain of digital intelligent information processing,the semantic relationship between word- level objects provides effective evidences for a variety of research questions. Semantic relatedness computation is the quantification manner of semantic relationships,among which the typical one is the distribution based approach. It converts linguistic objects to distributions over a semantic space,and evaluates two objects' semantic relatedness by examining the similarity between their corresponding distributions. This paper introduces in detail two representative approaches,such as the method based on context distribution,and knowledge resource element distribution,therefore summarizes them from the viewpoints of their fundamental resource's scale,quality and expandability.