The Implementation Of Video Recommendation System Based On Cloud Synchronization

Now we are in a period of video information overload, video recommendation system is to solve this problem, compared with the search engine requires the user to have clear goals and to provide search keywords, and video recommendation system is the application of recommendation algorithm for mining the user from user actions in the preferences and recommend videos. The development of cloud synchronization, recommended search science and technology becomes more mature, the development of science and technology has changed the way people use the network. In the process of Internet into people's life, gradually improve the customer requirements for network service. The Internet video resources is increasing, to watch the video of the user equipment is required for every kind of search for video, greatly increased the difficulty of doubt. Video recommendation system embodies the user of this service demand, video watch users hope that through their own interest to take to get the desired videorecommendation information results. But because of the traditional video recommendation technology limitations, video recommendation system most of the time it is difficult to make full and effective use of distributed technology advantage.In order to improve the processing ability of big data,to support Hadoop and provide a framework for distributed technology has developed as a basis for the recommended cloud synchronization technology.Based on the summary of theory research and application status of current domestic and foreign video recommendation system, the author carries out a specific commentary from four aspects: the collectionof user information, user modeling, the advantage ofmixed recommendation system, the design and implementation of video recommendation system, and explores the related key techniques, adoptsthe recommendation systemplatform with the technology of Hadoop based on mahout. The complected research works are as follows:(1) Analysed the demand of video recommendation system, confirmed the data model of recommendation system.(2) Analysed the traditional recommendation algorithm, carried out improvement to traditionalrecommendationalgorithm, carried out a mixed weighted hybrid recommendation algorithm according to the preference and duration in different equipmentand different time for application users.(3) Designed various modules of video recommendation system formed a structure of video recommendation system based on cloud synchronization.(4) Recommendationmodel achieved video recommendation system with the developmenttool of eclipse, and carried out basic function, browser compatibility, extendibility,performance test, verified the accuracy and availability of video recommendation system.Currently, video recommendation system has operated. Throughthe collection ofall viewing records of users by cloud synchronous, the author considers how to capture users' video preference in different time and different equipment with the aid of different recommended devices, and also makes a voluntary video recommendation for users. Finally, discusses the direction of improvement in the future for video recommendation system.