Interactive With Rcommender Systems

Recommender system has been developed as a field of study, and its role in solving the information overload and personalized information services has been widely recognized. Human-computer interaction focuses on the interaction between user and system in a situation, user interface is the interactive entrance of a system, which provides the communication between user and system, good interaction design enables user to enjoy intelligence services comfortably and efficiently. Throughout the research history of recommender system, recommender system which combines with human-computer interaction techniques has still rarely been studied. This article completed the following work from three important aspects:critiquing, recommend explanation and the social tag technology:1. Critiquing is introduced to the system design. The critiquing is a new concept, it aims to stimulate users to conduct tradeoff navigation and freely specify feedback. It can help set up the user preference model with higher accuracy and improve the cold start problem.2. Introduced tag technology to recommend explanation. Recommend explanation plays an active role in the success of recommendation, this article adds tag technology into explanation. The tags provide very good expression and discovery, effective tags can enhance the recommend interpretabiiity, raise the transparency of the recommendation and users’trust on the system.3. Based on the tag technology and user critiquing, one recommender system is designed. The system’s overall framework, the data model and recommendation engine are implemented. The system’s interfaces are also displayed.Critiquing-based recommendation is an innovative concept, it makes recommendation logic more transparent to users, helps users build confidence and trust on the system, and assists users make more efficient decisions.