9.1 Introduction

A growing number of groups are now working in text mining for biology. Two strands of research have converged on this problem: on the one hand, biology and bioinformatics, and on the other, computer science/natural language processing. The biology and bioinformatics research communities have felt an increasing need for improved tools to access and mine the rich textual information found in the biological literature, as well as the semistructured informa-tion in comment fields of the biological databases, such as GenBank (http://www.ncbi.nlm.nih.gov/Genbank/) or Swiss-Prot (http://www.us.expasy.org/sprot/). Computer science researchers, including people working in natural language processing, information retrieval, machine learning, and artificial intelligence, have long been interested in the general problem of accessing and mining textual information. What is new is the increased interaction among these communities.

On the natural language side, we have seen a series of workshops on biomedical text processing at the Association for Computation Linguistics (ACL), and Human Language Technology meetings since 2002 ^{[1] [2] [3]} In addition, the Text Retrieval Conference (TREC) ^[4] started a new Genomics Track in 2003 ^{[5] [6]}

On the biology side, there have been Special Interest Group (SIG) meetings on text mining for biology at the Intelligent Systems in Molecular Biology (ISMB) conference since 2001 ^[7] The Pacific Symposium on Biocomputing (http://www.psb.stanford.edu) has held sessions on text mining for biology since 2000. There also have been joint activities, including a workshop at the University of Pennsylvania in February 2001 ^[8], workshops at...