A Method of Mining Opinion Leaders Based on PageRank

Abstract

The invention proposes an opinion leader mining method based on PageRank, which belongs to the field of data mining. The present invention is based on the core idea of ​​the PageRank algorithm, which is introduced into the analysis of social network user influence. The present invention not only considers the relation topological structure information of social network users, but also integrates some attribute information of social network users reflecting their influence. In terms of social network topology information, the algorithm constructs a weighted user relationship network based on the forwarding relationship of users in the event. The event is of thematic nature and a large number of users participate in it. Using the forwarding mechanism to build a relationship network based on events not only has information The pertinence and targetness of the network can obviously reduce the use of social media information data, and at the same time, this kind of network is dynamic due to different events. For social network user attribute information, the algorithm uses user attribute information that can directly reflect user influence, and can more accurately describe user influence.

Description

technical field [0001] The invention belongs to the field of data mining, in particular to a method for mining opinion leaders based on PageRank. Background technique [0002] With the rapid development of Internet-based social media, social networks as a reflection and extension of the real world have become an inseparable part of people's daily life. As the main body of social networks, people actively participate in Internet-based social activities and dominate the interaction behavior of the entire social network. People can not only post new things they encounter anytime and anywhere on social platforms, but also Browse hot events in a timely manner, and participate in the discussion of events to express your views and emotions, and you can also find things you are interested in, and communicate and communicate with your friends and followers in a timely manner. [0003] In such a huge user base, users with different backgrounds and composition structures have differen...

AI Technical Summary

Problems solved by technology

[0004] The traditional influence node analysis method relies on the topological structure of the complex network to measure the key nodes. Several commonly used statistical indicators for the key ranking of undirected network nodes are degree centrality, node betweenness centrality, proximity centrality, k-shell, k-core, eigenvector centrality, etc. These indicators for measuring key nodes have a single applicability and a narrow range, and are difficult to apply to new social networks. New social networks are user-centered and between users The various relationships in the network are constructed by edges, and also include text information generated by users, social behaviors between users, etc. Therefore, the traditional measurement method of node importance is difficult to apply, and new key points suitable for new social networks are needed. Algorithms for Node Metrics

Images