Name
Abstract | ||
|---|---|---|
ContactFinder is an intelligent agent whose approach to assisting users is valuable and innovative in the following four ways. First, ContactFinder operates proactively in reading and responding to messages on electronic bulletin boards rather than acting in response to user queries. Second, ContactFinder assists users by referring them to other people who can help them, rather than attempting to find information that directly answers the user's specific question. Third, ContactFinder categorizes messages and extracts their topic areas using a set of heuristics that are very efficient and demonstrably highly effective. Fourth, ContactFinder posts its referrals back to the bulletin boards rather than simply communicating with specific users, to increase the information density and connectivity of the system. This paper discusses these aspects of the system and demonstrates their effectiveness in over six months of use on a large-scale internal bulletin board. |
Year | Venue | Keywords |
|---|---|---|
1996 | National Conference on Artificial Intelligence | specific user,contactfinder agent,intelligent agent,bulletin board,large-scale internal bulletin board,bulletin board question,specific question,user query,electronic bulletin board,information density,topic area,contactfinder categorizes message |
Field | DocType | ISBN |
Information density,Intelligent agent,World Wide Web,Computer science,Heuristics,Artificial intelligence,Multimedia,Machine learning,Bulletin board | Conference | 0-262-51091-X |
Citations | PageRank | References |
34 | 12.65 | 4 |
Authors | ||
2 | ||
Name | Order | Citations | PageRank |
|---|---|---|---|
| Bruce Krulwich | 1 | 238 | 50.70 |
| Chad Burkey | 2 | 82 | 30.73 |