THESIS
1998
xi, 76 leaves : ill. ; 30 cm
Abstract
The traditional solution to the problem of information retrieval in the World Wide Web is to build an index database of the documents from the WWW. The information retrieval system or the search engine uses traditional search and retrieval techniques to answer a query. One obvious problem is that the search engine may return an expired link to the user if the index database has not been updated for a considerable period of time. Furthermore, the search engine usually returns thousands of "matches" to answer one query. This can potentially confuse a user. Normally the user can only explore a very small part of the large Internet information space. We propose to solve the problem by using an intelligent agent to search for documents in the Web. In addition to its ability to interact with...[
Read more ]
The traditional solution to the problem of information retrieval in the World Wide Web is to build an index database of the documents from the WWW. The information retrieval system or the search engine uses traditional search and retrieval techniques to answer a query. One obvious problem is that the search engine may return an expired link to the user if the index database has not been updated for a considerable period of time. Furthermore, the search engine usually returns thousands of "matches" to answer one query. This can potentially confuse a user. Normally the user can only explore a very small part of the large Internet information space. We propose to solve the problem by using an intelligent agent to search for documents in the Web. In addition to its ability to interact with the dynamic environment, examine and retrieve documents, and select specific hyperlink to follow, the agent can learn from the search process and decide how to select a good link to pursue. Genetic algorithms and neural networks assist the agent to improve its performance. Finally, a simple local handler is used to steer the agent away from searching irrelevant links. The search is capable of retrieving a small number of relevant documents.
Post a Comment