Abstract:

This paper analyzes the deep interactive excavating of user information needs, and based on that, we propose the information retrieval method. By using the combination of interactive relevance feedback techniques of user participation and the deep excavating of user needs, we calculate the similarity of user demand vector with the document vector to achieve the accuracy matching of user requirements and retrieved results, which can effectively identify the users, at the same time, realize the precision and intelligent of information retrieval. The method is well applied in the intelligence retrieval system of the southern power grid center and does a good job.

Key words: Excavate to user needs, Relevance feedback, Intelligent information retrieval, Retrieval, Subject retrieval, User interaction, Rocchio algorithm