DISAMBIGUATION METHOD AND APPARATUS FOR AUTHOR OF PAPER, AND COMPUTER DEVICE

摘要

A disambiguation method for an author of a paper. The method comprises: respectively forming a name tree by means of the names of authors involved in all papers in a database and according to a preset rule (S1); acquiring an association relationship heterogeneous network corresponding to all the papers in the database (S2); acquiring paper semantic representations respectively corresponding to all the papers in the database (S3); on the basis of the name tree, the association relationship heterogeneous network and the paper semantic representations, constructing a similar matrix (S4); clustering the similar matrix to obtain paper clustering groups corresponding to all the papers in the database (S5); determining whether a paper clustering group corresponding to an author to be disambiguated belongs to a paper clustering group corresponding to a specified author (S6); and if not, determining that the author to be disambiguated is different from the specified author (S7). The names of authors are preprocessed to construct a name tree, and a clustering error caused by the different ways in which names can be written is then eliminated according to the name tree, thereby guaranteeing that the names of the same author are divided into the same group as much as possible, and improving the precision of name disambiguation.