The Design And Implementation Of Friends Recommendation System Based On Social Network

Nowadays,with the rapid development of the Internet,people are more and more inclined to obtain information from the network rather than other way,and at the same time,the way to make friends is also extended to the social network.The essence of social intercourse is human communication,and each user has their own communication circle in social network which can realize information share,pass in time and communion among each user.However,with the development of the Internet and the evolution of the social network,the user group in social network is becoming larger and larger,the relationship between users becomes more complex,and more and more data are generated by the users.It is difficult for users to find "like-minded" friends and build their communication circle quickly and accurately under these factors.In this background,the friend recommendation system arises at the historic moment and it can recommend friends with similar interests and hobbies for users.This paper takes the typical of social network——micro blog as the research object,and completed the design and implementation of the friends recommendation system based on social network.First of all,the system performs the collecting and pretreating of the data,and obtain useful data.Secondly,we use LDA subject model method to analyze the content of users' micro blog to get the user's Micro blog theme distribution information,and then calculate and show the user's interest preferences according to these information.Then,we calculate the similarity between different user interests according to the cosine similarity measure method,and select N users that have maximum similarity with the target users as the friends recommendation results.Finally,the system is evaluated by the precision index,and it's proved that our system has higher precision.Compared with the traditional friend recommendation system based on the user's personalized labels,educational background and geographic location or other information,we analyze the historical micro blog data of the users,mining user preferences from those data,and then make friends recommend,so we can get more representative description of user interest,and present to the user more "like-minded" result.