节点文献

车载多媒体系统与内容推荐研究

Research on Vehicle Multimedia System and Content Recommendation

【作者】 韩宇

【导师】 刘衍珩;

【作者基本信息】 吉林大学 , 软件工程硕士(专业学位), 2017, 硕士

【摘要】 随着我国高速铁路的普及,越来越多的用户选择高铁出行。在旅途中乘客对多媒体服务拥有广泛需求,而当前还缺少成熟的模式来为用户在高铁这种特殊场景下提供多媒体服务,这和高铁的特殊性有很大关系。首先高铁列车是全封闭车厢,信号的屏蔽效应很强,无法通过单点无线AP来覆盖整个列车的所有车厢;其次列车出厂后因为安全性的原因不允许进行改装,所以对已出厂列车而无法进行有线方式的布网。本课题研究组和相关领域公司合作,根据高铁列车特殊性进行了深入探讨,提出了通过无线串联的方式,实现各个车厢分布式的组网,设计实现了分布式车载多媒体运维系统。此系统主要分为三大模块:影音系统模块、内容推荐模块和运行维护模块。对于实现影音系统,作者首先分析了当前实现流媒体技术的主要方案,选择了效费比最高的方案进行了实际开发。影音系统模块是通过XAMPP服务包,HTML5,JAVASCRIPT,CSS技术和HTTP协议等,来实现客户端的影音播放功能;运行维护模块通过JAVA技术和SOCKET通信,来实现多个分布服务器间的信息传递、系统维护功能;内容推荐模块依托于影音系统模块和运行维护模块提供的基础数据与功能,通过协同过滤推荐算法,实现了根据用户历史浏览数据,为用户推荐喜爱的节目内容的功能,同时作者根据场景的特殊性,优化了推荐算法,提高了推荐系统在用户数据稀疏、用户对影片评分过少的情况下的准确性,为用户提供了更优的服务。

【Abstract】 With the popularity of high-speed railway in China,more and more users choose high-speed rail travel.In the long journey,passengers have a wide range of demand for multimedia services.However,there is a lack of mature models for users to provide multimedia services in the special scene of high-speed railway.It has a great relationship with the particularity of high-speed rail.First high-speed rail train is a closed car,the signal shielding effect is very strong.Therefore,single point wireless AP can’t cover all the carriages of the entire train.Second trains are not allowed to be modified after they leave factory because of the reasons for the safety.So we can’t install wired net on the train which has been left the factory.Through the in-depth discussion on the particularity of high-speed train,the author designs and realizes the distributed vehicle multimedia operation and maintenance system according to train’s networking characteristics.This system is divided into three modules: audio and video system module,the content recommended module and Operation-maintenance module.For the realization of audio-visual system,the author first analyzed the main program that achieved the current streaming media technology then chose the most cost-effective solution.Audio and video system module is through the XAMPP service pack,HTML5,JAVASCRIPT,CSS technology and HTTP protocol,to achieve the client’s audio and video playback.Operation and maintenance module through JAVA technology and SOCKET communication to achieve a number of distributed servers and system maintenance between the information transmission.Content recommendation module is the focus of this article.Through the collaborative filtering recommendation algorithm,it can recommend users to their favorite program content according to their historical browsing data.At the same time,the author optimizes the recommendation algorithm according to the particularity of the scene.And the author improves the recommendation system’s accuracy in the sparse user data and too little movie reviews and provides a better service for the users.

  • 【网络出版投稿人】 吉林大学
  • 【网络出版年期】2018年 01期
  • 【分类号】U463.6;TP311.52
  • 【下载频次】95
节点文献中: 

本文链接的文献网络图示:

本文的引文网络