A Recommendation System for Travel Services Based on Cyber-Anima

Abstract

In recent years, recommendation systems have developed greatly, and is so widely used in online systems such as book, movie or friend recommendation. Current recommendation systems have problems with cold start and new entry. Cyber-Anima is a cyber-image of a user's anima, and we can derive one's cyber-anima by its user input and other observations of the user. With Cyber-Anima, we can build a recommendation system. The system take the concepts in cyber-anima into consideration and build relations between these concepts and user behavior to make further recommendations. We do some experiment with the dataset from travelhub.cn, which is a platform for travel services and is sponsored by the national technology plan. From the experiment, our implementation is better than FISM and Item-based collaborative and the time cost is reasonable.