Version 2.5.2.0 |
Abstract
Grant Number: 5R29LM005413-05 Project Title: BIOLOGICAL KNOWLEDGE-BASE MANAGEMENT SYSTEM
PI Information: Name Title KARP, PETER D. pkarp@ai.sri.com DIRECTOR Abstract: Biomedical researchers are generating vast amounts of data and knowledge, at an accelerating pace. Neither existing database systems nor existing frame knowledge representation systems have the capabilities required to support the development of large, shared repositories of biological information. Therefore, SRI International proposes to integrate database and knowledge-representation technology to develop a biological knowledge-base management system (KBMS) with unprecedented power to en- code biological knowledge. This KBMS will provide the expressive power needed to represent biological knowledge in all its complexity, will support the evolution of complex knowledge-base schemas, will enable multiple users to access large amounts of reliably stored information in a shared fashion, and will support inference over this knowledge. Our implementation efforts will build on an existing frame representation system called THEO. We will extend THEO so that its underlying storage system utilizes a database management system to facilitate the construction of frame knowledge bases containing large numbers (millions) of persistent frames. In addition, we will implement a collaboration subsystem that coordinates concurrent distributed development of large knowledge bases by multiple users. To exercise and validate our development efforts, this KBMS will be used in an application to construct a large biological knowledge base of intermediary metabolism.
Public Health Relevance:
This Public Health Relevance is not available.Thesaurus Terms:
artificial intelligence, information system, metabolism
enzyme activity, information system analysis
computer program /software
Institution: SRI INTERNATIONAL MENLO PARK, CA 940253493 Fiscal Year: 1997 Department: Project Start: 01-JAN-1993 Project End: 31-DEC-1997 ICD: NATIONAL LIBRARY OF MEDICINE IRG: BLR