Study Of Algorithm Of Social Network Users' Influence And Its Implementation

In recent years, the rapid development of social networks has more and more far-reaching influence on people's lives. Social networks attract a large number of users because of their convenience, quickness, timeliness and other features. Deeply analyzing and mining the social network users is of great significance in the area of controlling public opinion,information propagation and advertising, and there have been increasing research efforts on social networks. But today's social networks are complex, diverse and have large amount of data. How to analyze the social network users accurately and efficiently based on these characteristics is very important, and this is also the main research content of this thesis.First, we analyzed the characteristics of micro-blog which is the representative of the social networks, took into account the number of likes,forwards and comments based on CASINO algorithm, and proposed an improved algorithm—TPURANK algorithm. We used this algorithm to analyze the users' influence indexes, and the experimental results show that the TPURANK algorithm is more scientific and reasonable.Second, we proposed a parallel scheme of TPURANK algorithm based on the analysis of the TPURANK algorithm and the MapReduce programming model, and implemented it on the Hadoop platform. The experimental results show that compared to TPURANK algorithm, the parallel algorithm has a better performance and a faster speed. This can help us to analyze the massive data of social networks and adapt to their growing number users and increasing scale.Third, On the basis of the previous work, we designed and implemented the social network analysis system, including requirements analysis, overall design, function module design and its implementation.The system can rank influence index and conformity index, and search the information of users.Finally, we analyzed and compared the results calculated by the social network analysis system, and found that the influence indexes of users are related to the number of in-links, likes, forwards and comments. This is very consistent with the actual situation and of great reference. So the system is very suitable for the analysis of social network users. In addition,we summarized our work and forecasted the future work.As one of the important platforms of modern information communication, social networks have a very high research value and broad application prospects. Deeply analyzing the social network users can not only help us find more valuable information, but also can promote the development of social networks, and this is also an important goal of our research work.