Journal Search Engine
The Journal of The Korea Institute of Intelligent Transport Systems Vol.8 No.1 pp.119-126
사용자 상황을 이용한 추천 서비스 시스템의 필터링 기법에 관한 연구
A Study on a Filtering Method of Recommendation Service System Using User's Context
Abstract
In recent years, many recommendation service systems that search or recommend information automatically considering user's taste or property are developed. However, there is a weak point that correct recommendation is hard without considering the preference of user's context. This paper proposes a filtering method that gives correct recommendation considering the preference of user's context. To support this method, we get UCOP(User-Context Object Preference) using the preference of user's context and Pearson correlation coefficient. The results of the experiment show the improvement of 11%, 2% of precision and 8%, 4% of recall comparing with the existing service systems. Our recommendation service systems show 77% of precision and 53% of recall overall.