A Novel Ontology-Based Approach of Chinese Person Name Disambiguation

It is very common that various people share same name or a person has various names, which causes person name ambiguities. Existing methods group documents containing person names into various clusters and each cluster refers to an individual, while which target individual a person name occurring in a document refers to is still a challenged problem. In this paper, we firstly construct a person ontology with lots of person instances. Then, an approach of Chinese person name disambiguation based on the person ontology is presented: (1) Extract relevant personal features of a person name contained in a Chinese document, and (2) An improved person ontology tree matching algorithm is suggested to calculate similarities between the individual the person name represents and one of person instances sharing same name, rank the similarities and the top one is returned. To verify the validity of the proposed approach, three experiments are conducted on real web documents. Experimental results show that the proposed approach not only can effectively distinguish Chinese person name, but also can determine an individual for the person name.