Chinese Scientific And Technological Named Entity identification and Relation extraction based on knowledge tree

Abstract

In Scientific and Technological (S&T) papers, we first define Chinese S&T Named Entity (CSTNE) and Relation between Chinese S&T Named Entities (RCSTNE) that can offer help to solve the problem of deep knowledge mining in S&T literature database. Based on the definition of CSTNE and RCSTNE, we build their models which include semantic information. The relevant methods based on knowledge tree are also proposed to identify CSTNE and extract RCSTNE from S&T literatures. The evaluation on a large test set shows good result. Our experiments further demonstrate the practicality of our method.