Design And Implementation Of Recommendation Subsystem In Information Sharing Social Platform

With the advent of web2.0era, Internet develops so fast that the size of information which users can get is much bigger than ever. In facing a flood of information, users are difficult to find the content that they are interested in and the time cost of information acquisition becomes out of control. Although search engine technology do help users find items, but user tend to get pushed content rather than to search the content using search engine. Recommendation system has made great progress in this context.With the rapid developing of Internet, applications based on social network emerge with a rush and more and more knowledge sharing applications based on social network appear. But with the development of applications, the information in the application begins to appear serious redundancy so that people have to face the problem of selection and information obtained confusion. Therefore, there is an urgent requirement for people to have a recommendation system to get valuable recommended content.In this paper, the recommendation system in social networks and related technologies are mainly researched. By reconstructing the recommendation algorithm, the algorithm can be combined with the special nature of the social network and can make a depth analysis on user behavior to get a more accurate user preferences, improving the accuracy of the algorithm; in the recommended system development, the pattern of the plug-in based and the ability based were used so that the architecture could be extended easily; by using distributed computing and distributed storage, the system become efficient.In this papery the background about knowledge sharing system and recommendation system was introduced first. Then, it comes to the essence of this paper. The design and implementation of the recommendation system were elaborated in detail, introducing the implementation of each module and related technologies. In the end, the effectiveness and availability of the recommendation system is verified by a set of tests.