The Journal of the Korea institute of electronic communication sciences (한국전자통신학회논문지)

Korea Institute of Electronic Communication Science (한국전자통신학회)
권호 표지
DOI QR코드

DOI QR Code

Personalized Search Technique using Users' Personal Profiles

사용자 개인 프로파일을 이용한 개인화 검색 기법

  • 윤성희 (상명대학교 소프트웨어학과)
  • Received : 2019.05.16
  • Accepted : 2019.06.15
  • Published : 2019.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

This paper proposes a personalized web search technique that produces ranked results reflecting user's query intents and individual interests. The performance of personalized search relies on an effective users' profiling strategy to accurately capture their interests and preferences. User profile is a data set of words and customized weights based on recent user queries and the topic words of web documents from their click history. Personal profile is used to expand a user query to the personalized query before the web search. To determine the exact meaning of ambiguous queries and topic words, this strategy uses WordNet to calculate semantic similarities to words in the user personal profile. Experimental results with query expansion and re-ranking modules installed on general search systems shows enhanced performance with this personalized search technique in terms of precision and recall.

본 논문은 사용자의 검색 의도와 개별 관심을 반영한 순위화된 검색 결과 문서를 제공하는 개인화 검색 기법을 제안한다. 개인화 검색에서는 사용자의 개별 관심사와 선호도를 정확하게 판별하기 위한 사용자 프로파일을 생성하는 기술이 개인화 검색의 성능을 좌우한다. 개인 프로파일은 사용자의 최근 입력 질의어들과 검색과정에서 참조했던 문서들에 나타나는 주제어들의 가중치와 빈도가 기록된 데이터 집합이다. 사용자 프로파일은 웹 검색에 앞서 사용자의 입력 질의어를 개인화된 질의어들로 확장하기 위해 사용된다. 중의적 질의어의 정확한 의미를 결정하기 위해서 워드넷을 사용하여 프로파일에 등록된 단어들과 의미 유사도를 계산한다. 검색 시스템의 사용자 측에 질의확장 모듈과 순위 재계산 모듈을 확장모듈로 구축하여 진행한 실험에서 개인화 검색 기술을 적용한 실험 결과가 상위문서들에 대해서 정확률과 재현률이 크게 향상된 성능을 보이고 있다.

Keywords

KCTSAD_2019_v14n3_587_f0001.png 이미지

Fig. 1 Personal profile and modules

KCTSAD_2019_v14n3_587_f0002.png 이미지

Fig. 2 Architecture of personalized search with user profile

KCTSAD_2019_v14n3_587_f0003.png 이미지

Fig. 3 Enhanced precision of personalized search

KCTSAD_2019_v14n3_587_f0004.png 이미지

Fig. 4 Enhanced recall of personalized search

Table 1. Experimental environments summary

KCTSAD_2019_v14n3_587_t0001.png 이미지

References

  1. S. Park, "Analysis and Evaluation of Term Suggestion Services of Korean Search Portals: The Case of Naver and Google Korea," J. of the Korean Society of Information Management, vol. 30, no. 2, 2013, pp.297-315. https://doi.org/10.3743/KOSIM.2013.30.2.297
  2. S. Koratkar and S. A. Takale, "Deriving Concept Based User Profile for Search Engine Personalization," Int. J. of Science and Research, vol. 4, no. 6, 2013, pp.3086-3089.
  3. K. R. Remesh and P. Samuel, "Concept Networks for Personalized Web Search Using Genetic Algorithm," Int. Conf. on Information and Communication Technologies, Kochi, India, Dec. 2015, pp. 566-573.
  4. G. Park and S. Lee, "Personalized Search based on Community through Automatic Analysis of Query Patterns," J. of the Korean Institute of Information Scientists and Engineers: Database, vol. 36, no. 4, 2009, pp. 321-326.
  5. D. Kim, S. Kan, H. Kim, and B. Lee, "Folksonomy-based Personalized Web Search System," J. of Digital Contents Society, vol. 11, no. 1, 2010, pp. 105-116.
  6. S. Yoon, "Using Query Word Senses and User Feedback to Improve Precision of Search Engine," J. of the Korean Society of Information Management, vol. 26, no. 4, 2009, pp. 81-91. https://doi.org/10.3743/KOSIM.2009.26.4.081
  7. B. Kim, "Words Recommendation Algorithm for Similarity Connection based on Data Transmutability," J. of The Korea Institute of Electronic Communication Sciences, vol. 8, no. 11, 2013, pp. 1719-1724. https://doi.org/10.13067/JKIECS.2013.8.11.1719
  8. Y. Lee and Y. Chung, "An Experimental Study on an Effective Word Sense Disambiguation Model Based on Automatic Sense Tagging Using Dictionary Information," J. of the Korean Society for Information Management, vol .24, no. 1, 2007, pp. 321-342.
  9. Princeton University, "About WordNet," WordNet online Technical report, 2010.
  10. Y. Kim and Y. Kim, "A Question Example Generation System for Multiple Choice Tests by utilizing Concept Similarity in Korean WordNet," J. of the Korean Information Processing Society, vol. 15-A, no. 2, 2008. pp. 125-134.
  11. I. Lee, D. Hwang, Y. Hahm, and K. Choi, "Open Korean WordNet(KWN): Dictionary-based Semi-Automatic Development," The 26th Annual Conf. on Human & Language Technology, Busan, Korea, Oct. 2014. pp. 193-196.
  12. L. Meng, R. Huang, and J. Gu, "A Review of Semantic Similarity Measures in WordNet," Int. J. of Hybrid Information Technology. vol. 6, no. 1, 2013. pp. 1-12.
  13. T. Simpson and T. Dao, "WordNet-based semantic similarity measurement," Engineering Applications of Artificial Intelligence, vol. 36, Mar. 2016, pp.80-88.