Information Extraction from Biological Papers

Abstract

In recent years, the rapid development of computer technologies has facilitated the proliferation of bioinformatics. Except hundreds of on-line databases characterizing biological information such as sequences, structures, molecular interactions, ?€?, etc., another rich source of on-line information is the scientific literature. However information in text form, such as MEDLINE records, remains a greatly underutilized source of biological information. Hence there is growing need to develop empirical information extraction methods that can inexpensively and accurately map information scientific text sources into a structured representation, such as knowledge base. In this talk, the information extraction task will cover the issues such as the identification of biological objects, keyword prediction, term clustering and the identification of domain-specific relations. Some empirical methods useful to deal with these issues will be introduced.