Term Recognition and Classi cation in Biological Science

In this paper we describe the application of automatic terminology recognition and classi-cation techniques for two bioinformatics projects: extraction of information about enzymes and metabolic pathways and extraction of information about protein structure, in both cases from scientiic journal papers. The techniques we use were adapted from already available Information Extraction (IE) technology that was developed for name recognition tasks deened in the US Message Understanding Conferences (MUC). Preliminary evaluation results of the terminological identiication components of our systems are quite encouraging and compare favourably with the state-of-the-art results in the MUC competitions. Further, this technology is quite generic and may be readily adapted to other subdomains in biological and medical science. Indeed, we are optimistic that the same methods will be applicable to term identiica-tion tasks in any technical domain.