Spark environment-based microblog network key user mining system and method

Abstract

The invention discloses a Spark environment-based microblog network key user mining system and method. The system integrates contents in two aspects of a social network topology structure and user comprehensive information, and proposes an integral framework of key user mining in a large-scale social network. The user topic similarity is introduced for representing the strength of a community relationship among users, and a Louvain community discovery algorithm is applied, so that high efficiency of community discovery is ensured and a division result is relatively accurate; according to fourevaluation indexes of authority, transmissibility, interactivity and linkability of microblog users, compared with an existing user influence evaluation method, the diversity of the evaluation indexesis expanded and the comprehensiveness of an evaluation result is enhanced; and a mainstream Spark big data computing platform is introduced in computing, so that the whole key user mining has not only a complete implementation process but also environment conditions of actual application.

Description

technical field [0001] The invention relates to the technical field of social network data mining, in particular to a system and method for mining key users of a microblog network based on a Spark environment. Background technique [0002] With the rapid development of the mobile Internet, social networking platforms have more and more functions and attributes, and the number of daily active users of social networking platforms represented by WeChat and Weibo is in the hundreds of millions. In-depth research on social network data, mining user relationships, community research, public opinion monitoring, and user recommendations have increasingly become the focus of academic and industrial circles. [0003] The Weibo platform has become an important platform in social media due to its strong dissemination, large number of users, and simple operation, and its high-value user ratio is as high as 76.3%. On February 13, 2018, Weibo released its financial report for the fourth q...

AI Technical Summary

Problems solved by technology

[0005] Existing mining methods for key users in social networks or because they seldom consider comprehensive user information, microblog network topology, etc., or are only applicable to specific small-scale scenarios and the calculation process is complex, resulting in inaccurate mining results and low efficiency. The mining effect of key users in a large-scale social network environment is not ideal

Images