Document similarity calculation model of CSLN

Abstract

CSLN is a citation semantic link network with semantic representation and reasoning abilities. The documents in citation semantic link network (CSLN) is a set of many objects. The similarity calculation of a single object can't accurately calculate the document similarity of citation semantic link network. This paper analyzes in depth the characteristics of each part of the document, and puts forward the document similarity calculation model of the citation semantic link network. This model not only improves the accuracy of document similarity calculation of the citation semantic link network, but also lays a foundation for the community discovery of citation semantic link network and the hot prediction of research.