一种基于子周期社区挖掘的电视节目推荐方法

A Recommendation Methods of Television Programs Based on Sub-communities Mining

针对电视节目的互动性差、即时性要求较高,使得网络上通用的推荐电影、书籍的算法不适合电视节目的推荐这一现象,笔者根据用户观看电视的历史数据,得出用户观看电视的规律,根据当前时间生成推荐列表向用户推荐电视频道,提高电视节目的推荐效率。在真实的数据上做实验分析,并与其他主流推荐算法的命中率作对比,综合分析得出,本算法的推荐效果要高于其他算法。

Television programs have a poor interactivity and high real-time requirements.As a result,the current algorithm of the recommendation for the movies and books is not suitable for the television program.Aimed at this problem,this paper,according the historical data,obtained the regular pattern of TV watching,and produced a current list of the program for the users.Experiment and analysis were conducted with real data set,and the hit rate was compared between the algorithm in this paper and other major recommendation algorithms.Comprehensive analysis shows that,this algorithm had superior recommendation results compared with other algorithms.