广电家庭用户个性化推荐的难点和解决思路  

Difficulties and Solutions for TV Recommender Systems

作　　者：李培琳[1] 赵明[1] 王炳飞 

机构地区：[1]国家新闻出版广电总局广播科学研究院,北京100866 [2]北京数码视讯科技股份有限公司,北京100085

出　　处：《电视技术》2015年第18期33-35,49,共4页Video Engineering

基　　金：2014年国家新闻出版广电总局科研项目"有线电视大数据融合分析平台设计及关键技术研究"

摘　　要：随着广电运营商双向业务的不断发展,为解决用户无法从海量内容中快速找到喜爱节目的问题,个性化推荐技术在广电领域得到了关注和应用。个性化推荐技术在互联网等领域已经有了较成熟的应用,并得到了良好的效果,但若具体落实在广电运营商的业务中并起到良好的作用,还需要解决直播节目元数据关联、推荐实效性、家庭用户行为分析等难点问题。对广电家庭用户个性化推荐存在的难点进行了总结,并提出解决思路。With the development of radio ＆ television operators two-way business, quickly finding a favorite program to solve the problem of facing mass content, and personalizing recommendation technology has been concerned and applied in the field of broadcasting. Persnnalized recommendation technology in the lnternet and other fields has been mature applications, and obtained good effect, but if the concrete implementation in radio ＆ television operators in the business and play a good rnle, still need to solve the difficulties of live broadcast metadata association, the effectiveness of recommendation, and problems such as family user behavior analysis. In this paper, on radio ＆ television home users personalized recommendation existing difficulties are summarized, and brought proposed solutions.

关 键 词：个性化推荐 用户行为分析 广电运营 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构] TN943[自动化与计算机技术—计算机科学与技术]