|
Objective
The goal of the Scalable Agent-based Information Retrieval Engine (SAIRE)
project is to develop
software tools which incorporate intelligent methods in support of public
access of Earth and space
science data over the Internet.
Approach
Deploy SAIRE - A Scalable Agent-based Information Retrieval Engine as
a multi-agent search engine employing intelligent software agents, natural
language understanding, and conceptual search techniques to support public
access to earth and space science data over the Internet.
Accept natural language queries from web browsers on heterogeneous computer
platforms and permits clarification dialogues to support incomplete requests.
Significance
Demonstrates a technology for simplifying access to NASA and NOAA image
data sets.
Point of Contact
Jide Odubiyi, Principal Investigator
Lockheed Martin Space Mission Systems
7375 Executive Place
Seabrook, MD 20706
Email: jideo@groucho.sms.lmco.com
Phone: 301-805-0305
Fax: 301-805-0521
| Accomplishments
Developed collaborating intelligent software agents that support Internet
information retrieval from multiple databases in response to a single natural
language query. SAIRE's natural language module covers 13 topics, 721 terms,
and over 2,300 words in its Earth Science dictionary.
Implemented a mechanism for clarifying ambiguous words. This is achieved
through user-agent dialogues and a user modeling agent.
Developed on-line help and visual displays on the operational status of
the multi-agent network.
Demonstrated SAIRE at the EOSDIS workshops and ADL '96 conference. Our site
is visited by users from about 120 unique IP addresses per month.
Plans
Develop a spoken language interface. Address security issues. Deliver Build
3 of system to RSPAC for implementation. This agreement expires in Fiscal
Year 98.
|