The Research Of Social Network Friend Recommendation System In Cloud Computing Environment

Social networks tend to have a huge user groups, and therefore will have a very large data set. It’s an important research field of social network and a challenge in the age of big data to help users find useful information from massive data. The massive data ask for huge, high-speed storage. In order to calculate the mass data effectively, it’s necessary to maximize the use of computing and network resources, and computing virtualiza t io n and network virtualization are essential.These distributed and heterogeneous computing, storage and network resources are difficult to manage for traditional system software. However, cloud computing is a solution to address these chal enges. Cloud computing technology can be very effective for large-scale data storage and computing, and is an effective means to deal with the issues of informatio n overload and scalability of recommendation system. The open-source software Hadoop provides a software platform for cloud computing infrastructure, including HDFS and MapReduce, and the integrat io n of many components such as databases. Hadoop has the ability of large-scale data distributed processing, and many advantages such as high efficiency, high reliability, high scalability, relatively cheap and so on. Hadoop is widely used in industry and academia.This paper studies friend recommendation system of social network on Hadoop cloud platform.According to the data characteristics of social networks, the main basis of friend recommendation is determined to be the information of the friend relations hip network, interactions and interests. The traditiona l collaborative filtering technology is applied to the field of friend recommendation to get the tendency in the social dimension.The computing method of text similarity in data mining is used to evaluate the interest similarity between users.And then recommendation result is obtained by integrating the above information. In order to analyze and deal with large-scale data, the algorithm is improved according to MapReduce computing framework.The system includes data acquisition and preprocessing module, data storage module and data processing module.The data acquisition module use web crawler or open API of social networking site to collect user data,and the data is pre-processed so that it can meet the input format requirements of recommendation algorithm.The data storage module determine the form of data files and operations of database.The data processing module deal with the data in file system or database on the Hadoop platform to get recommended results. The data processing module on the Hadoop platform, according to the distributed algorithm to deal with the data file system or database, recommended results.At last, Hadoop cloud computing platform is set up to make experiments. The results show that the algorithm in this paper compared with the traditional friend recommendation method has certain promotion, and also can effectively operate in a distributed cluster.