FU Berlin, Fachbereich Mathematik und Informatik, Institut für Informatik

Vortrag des Informatik-Kolloquiums

Mining Biomedical Literature

Prof. Dr. Ulf Leser, Knowledge Management in Bioinformatics, Humboldt-Universität zu Berlin

Recent years have seen a steep increase of interest in Text Mining in the Life Sciences. Despite the many databases in this area, most knowledge is still published in scientific papers, i.e., in the form of essentially unstructured, natural language text. Researcher face app. 400.000 new publications in over 5.000 journals each year. At the same time, many results in modern biotechnology are based on high-throughput methods that generate very little information about very many objects. For instance, microarray-based gene expression profiling can measure the expression intensity of more than 20.000 genes on a single chip; but for robust results, these measurements need to be combined with additional information about the single genes. The talk will give an introduction into text mining in life sciences. The main application areas are presented. Special focus is put on two recent projects. First, we present a system for document classification which supports researches in Systems Biology. Second, we describe the specific problems of named entity-recognition in molecular biology and present our approach to this problem. The first project is joint work with the group of Edda Klipp, Max-Planck-Institute for Molecular Genetics. The second project is joint work with the group of Tobias Scheffer, Humboldt-Universitaet Berlin.