面向互动电视的影视节目推荐系统研究与实现

发布时间：2018-03-29 00:21

本文选题：互动电视　切入点：推荐系统　出处：《复旦大学》2012年硕士论文

【摘要】：随着互联网、广播电视网络的不断发展,网络上的信息不断地增加。尤其是在WEB2.0技术的迅猛发展之下,互联网已经成为全球最大的信息库,它给我们带来极大便利的同时,也给带来了信息膨胀的问题。作为我国信息产业发展的战略目标(“三网融合”)正不断地改变着我们的生活。随着2010年国务院颁布《加快发展三网融合发展》的政策,最近几年下一代广播电视网络得到了快速的发展,具有互动点播功能的机顶盒走进了我们的生活,可供用户点播的视频也越来越多,电视用户获得了互联互通、个性化搜索、个性化推荐等全方面的服务。随着数字电视网络中中可供用户点播的视频数量越来越多,在数以万计的视频面前,用户感到迷茫,搜索引擎也只能解决一小部分的问题。所以而向互动电视的个性化推荐系统逐渐成为广大学者的研究重点。影视节目搜索点播是互动电视的一个重要功能,随着影视节目数量的日益增多,内容日益复杂,用户越来越对挑选节目感到疑惑,视频推荐技术可以作为一种理想的解决方案。在用户历史播放记录的基础上,通过分析用户的行为和喜好,再根据互动电视特有的信息(如多用户共用机顶盒、节目的时间特征),最后为用户产生一组推荐。本文主要研究了互动电视点播系统中的影视节目推荐技术,并且实现了一套面向互动电视的个性化推荐系统。该系统通过分析影视节目之间的关系和用户的历史记录,挖掘用户之间的相似度和偏好,进而给用户推荐一组个性化的影视节目列表,减少了用户选择视频的时间,提升了用户的体验。
[Abstract]:With the continuous development of the Internet and radio and television networks, the information on the network is constantly increasing. Especially with the rapid development of WEB2.0 technology, the Internet has become the largest information base in the world, which brings us great convenience at the same time. It has also brought the problem of information inflation. As a strategic goal of the development of our information industry ("three networks convergence"), it is constantly changing our lives. With the promulgation of the policy of "accelerating the Development of three Networks Integration" by the State Council in 2010, In recent years, the next generation radio and television network has been developing rapidly. The set-top box with interactive on-demand function has come into our life, and more and more videos can be delivered to users on demand. Television users have gained connectivity and personalized search. Personalized recommendation and other services. As more and more videos are available to users on demand in the digital television network, users feel confused in the face of tens of thousands of videos. Search engine can only solve a small part of the problem, so the personalized recommendation system to interactive television has gradually become the research focus of the majority of scholars. Video program search on demand is an important function of interactive television. With the increasing number of TV programs and the increasing complexity of the content, users are more and more confused about the selection of programs. Video recommendation technology can be used as an ideal solution. Based on the history of users, by analyzing the behavior and preferences of users, and based on the specific information of interactive TV (such as multi-user sharing set-top box, etc.), The time feature of the program, and finally a group of recommendations for the user. This paper mainly studies the technology of the video program recommendation in the interactive television on demand system. A personalized recommendation system for interactive TV is implemented, which analyzes the relationship between TV programs and users' historical records, and excavates the similarity and preference between users. Then we recommend a set of personalized TV program list to users, which reduces the time for users to choose video, and improves the user's experience.
【学位授予单位】：复旦大学
【学位级别】：硕士
【学位授予年份】：2012
【分类号】：TP391.3

【参考文献】