Approach to extracting non-taxonomic relationships for Chinese ontology from web
MA Wei-wei · Jisuanji gongcheng yu sheji · 2010
To learn non-taxonomic relationships of ontology effectively to help knowledge engineers build domain ontologies,a method for automatic extraction of non-taxonomic relationship between concepts in the Chinese domain ontology learning environment is presented,which uses web as the data source to extract candidate relations and to calculate the demographic characteristics of the information distribution.It uses verb as the pivot to find the non-taxonomic relationship,the domain-related verb as the seed to search related concepts,and which is also used to tag the corresponding relationship.The learning result of this method is a semantic system composed of multilevel taxonomic and non-taxonomic relationships.Finally,the result and performance analysis of the proposed method are demonstrated by extracting corresponding relationships of ontologycancer.