A Method of Discovering Collaborative Users Based on Psychological Model in Academic Recommendation
Jie Yu, Haihong Zhao, Fangfang Liu, Geke Xie · 2012
Facing lager amount of web academic information and resources, collaborative filtering is an effective way to improve the efficiency of researchers information seeking. We can predict the researcher interest and needed by collaborative users which have similar interest to current researcher. So how to discover collaborative users is a key issue in collaborative filtering. But current existing methods cannot meet the needs of academic researchers from the user cognitive capacity and level. According to the current problems existing in the research, this paper proposes an approach which obtains the user's interest and discovers collaborative users based on psychology model. First, we proposed Browsing Interest Model for Personalized Service based on attitude behavior relationship model in psychology. Secondly, the paper presents a users similarity measure method based on the contrast model in psychology between users in the academic database which is Similarity Measure Mode for Personalized Service. Finally, according to acquired user interest which is represented by Concept-Relation Graph for Personalized Service and an improved user similarity measure method between users which is expressed by SMMPS that we can obtain collaborative users. Experimental results demonstrate that the proposed algorithm is better than the traditional algorithm for discovering collaborative users in academic database.