Modeling and extracting hyponymy relationships on Chinese electric power field content

Abstract

More and more electric power data are generated with the development of the electric power field. Hyponymy relationships extraction is an important task in building semantic knowledge base, information retrieval and other semantic applications etc. This paper uses sentences to match “is a” pattern as research subjects and proposes a novel hyponymy relation extraction approach, which combines lexical analysis and dependency parsing. In order to better complete experiment of the proposed method, the electric power field corpus is processed in this paper and more normative corpus are obtained. Some hyponymy relationships are obtained from the electric power concepts with proposed method. The experiment results confirm that the proposed approach of this paper has better performance in the process of relationship extraction.