NLP for Bioinformatics --- Toward Knowledge Integration and Management

Toru Hisamitsu

Business situation of bioinformatics has been drastically changing. Knowledge management, however, has been consistently playing an important role and is becoming one of a central technology for biomedical research. Confronted with the exponential growth of text data, text management is a major part of knowledge management. Among various technologies, our emphasis is on associative search and text mining: associative search dynamically relates documents in different databases that contain a hundred millions of documents, and our text mining system extracts, for instance, protein-protein interactions from a large collection of documents such as MEDLINE, which contains twelve million abstracts. The extracted data are visualized in various ways so that a human can comprehend them intuitively. Lastly we briefly introduce research of computational terminology, which is critical to the quality of the text management system.