The CMU BioKB Project

Overview:

The goals of our project are twofold:

To develop methods for extracting information from on-line, scientific text sources (such as Medline), and mapping this information into a structured format (i.e. a knowledge base) so that it can be used for query answering and problem solving.
To develop a knowledge base characterizing subcellular protein localization patterns.

Towards this end, we are currently developing new information-extraction methods, developing an ontology of protein localization patterns, and populating our knowledge base with facts extracted from sources such as Medline and SwissProt.

This project is closely related to the WebKB project, which is devloping methods for extracting knowledge bases from the Web, and to the Murphy lab, which is analyzing fluorescence microscopy images and investigating the endocytic pathway.

Project Members:

Mark Craven
Toby Goldstein
Chris Hauser
Kiarri Kershaw
Bob Murphy
Francisco Pereira

Friends of the Project:

Susan Henry
Johan Kumlien
Tom Mitchell
Mark Perlin
David Whitcomb

First Publication:

Constructing Biological Knowledge Bases by Extracting Information from Text Sources.
Mark Craven and Johan Kumlien.
Proceedings of the 7th International Conference on Intelligent Systems for Molecular Biology (ISMB-99).

last update: April 21, 1999 by Mark Craven