Online group movie recommendation method

Abstract

The invention discloses an online group movie recommendation method. The method comprises the following steps: the step 1, calculating social influence of movie types; the step 2, performing statistics on the familiarity of group members to the movie types; the step 3, calculating the speaking right of the group members for a certain film type; the step 4, fusing preferences of group members; and the step 5, carrying out movie recommendation. Through the movie type social influence calculation method based on sampling, calculation is simplified, and the movie popularity trend is reflected; and different audience weights are considered by using an audience preference fusion method, so that reasonable group preference is obtained. According to the method, the influence of the inside and outside of the group on audience preference fusion can be considered, so that the recommendation effectiveness is ensured.

Description

technical field [0001] The invention belongs to the field of computer application technology, and relates to a method for recommending movie groups based on preference fusion (that is, multiple audiences with similar characteristics). method can achieve good results. Background technique [0002] In daily life, with the popularization of smart phones, mobile social service platforms have sprung up like mushrooms, greatly enriching people's daily life. The emergence of these platforms has expanded people's social radius and facilitated the organization and holding of group activities. In order to recommend items to groups of viewers that satisfy as many members of the group as possible, group recommendation systems emerge as the times require. Compared with the traditional recommendation system, the group recommendation system introduces the concept of group. The reason why the design of group recommendation systems is more challenging is mainly because it is very difficul...

AI Technical Summary

Problems solved by technology

[0004] Most existing audience preference fusion methods perform preference fusion based on the data characteristics of audience ratings. The rationality and interpretability of audience weights are ignored during preference fusion, and the accuracy of the obtained group preferences cannot be guaranteed. The performance bottleneck of group recommender systems

Moreover, the audience dynamics in the set U change during the online movie recommendation process, which further increases the difficulty of weight calculation.

Images