DOI QR코드

DOI QR Code

Development of the Potential Query Recommendation System using User's Search History

사용자 검색이력 기반의 잠재적 질의어 추천 시스템 개발

  • Park, Jeongbae (Dept. of Computer Science Education, Korea University) ;
  • Park, Kinam (Dept. of Computer Software Engineering, Soonchunhyang University) ;
  • Lim, Heuiseok (Dept. of Computer Science Education, Korea University)
  • 박정배 (고려대학교 컴퓨터교육과) ;
  • 박기남 (순천향대학교 컴퓨터소프트웨어공학과) ;
  • 임희석 (고려대학교 컴퓨터교육과)
  • Received : 2013.06.11
  • Accepted : 2013.07.20
  • Published : 2013.07.28

Abstract

In this paper, a user search history based potential query recommendation system is proposed to enable the user of information search system to represent one's potential desire for information in terms of query and to facilitate the desired information to be searched. The proposed system has analyzed the association with the existing users's search histories based on the users' search query, and it has extracted the users's potential desire for information. The extracted potential desire for information is represented in terms of recommended query and thereby made recommendations to users. In order to analyze the effectiveness of the system proposed in this paper, we conducted behavioral experiments by using search histories of 27656. As a result of behavioral experiments, the experiment subjects were found to show a statistically significant higher level of satisfaction when using the proposed system as compared to using general search engines.

Keywords

Query recommendation;Information retrieval;Relevant rule;Search history;Potential query

Acknowledgement

Supported by : 산학협동재단