본문 바로가기
  • Home

Concept and Attribute based Answer Retrieval

  • Journal of The Korea Society of Computer and Information
  • Abbr : JKSCI
  • 2005, 10(3), pp.1-10
  • Publisher : The Korean Society Of Computer And Information
  • Research Area : Engineering > Computer Science

Yun Bo-Hyun 1 Changho Seo 2

1목원대학교
2공주대학교

Candidate

ABSTRACT

This paper presents the information retrieval system which can retrieve the most appropriate answer sentence for user queries by using the concept and the attribute for the knowledge retrieval. The system analyzes the user query into the Boolean queries with the concept and the attribute and then retrieve the relevant documents in the indexing set of answer documents. Users can retrieve the relevant answer sentences from the relevant documents. For this, the answer documents indexed by the concept and the attribute are segmented by each sentence respectively. Thus, the segmented sentences are analyzed into the concept and the attribute of which the relevance degree with indexing units of documents is evaluated. Then, the system indexes the location of answer sentences. In the experiment, we evaluate the performance of our answer retrieval system against 100 user queries and show the experimental results.

Citation status

* References for papers published after 2022 are currently being built.