Research And Application Of SNS Relation Chain Analysis System Based On Hadoop

With the rapid development of the Internet products, all kinds of data also show explosivegrowth, the amount of information is too large, also often bring trouble to users, users have tospend a lot of time and effort to identify and filter the information. At this time, the value ofthe recommendation engine can be revealed, through the data mining and analysis, at thesame time using machine learning and classification, clustering, for users to screening andfound that they might be interested in things, analysis of user behavior, to predict the user’spreferences, with all the growth of the Internet product, such demand also increasinglyapparent, recommendation engines also destined to lead the SNS to future more intelligent,more enrichment, more personalized development, in the future, simply provide a platformwill no longer be able to meet the needs of users, at the same time, for the Internet company, afew short years, they have accumulated a large amount of data, most are free, we know thatthe Internet products companies want to better profit must make use of these data, effectivelyexcavate vast amounts of user data, is not only beneficial to improve the user viscosity, butalso can quickly locate the product development direction in the future.In the whole analysis system, user relationship is a very important link, it is related to theestablishment and development of the whole ecosystem. User can only be better set up theirown circle of friends in the system to find what you need content, increasing the viscosity ofthe users presence as well as to the system.This system from two aspects of user relationshipchain and user attributes a systematic analysis was carried out on the relationship between theuser and recommended, among them, the chain of the relationship based on the analysis of theuser, using the vector space model, new methods, such as triangle closure law and based onthe Hadoop was realized on the big data platform, and the algorithm effect and performanceof the system are analyzed.