Microblog specific event attention group identification method

Abstract

The invention discloses a microblog specific event attention group identification method. A key user of a specific event is identified through an improved PageRank algorithm; on the basis of identifying the key attention user of the specific event, feature extraction is carried out on key attention users, vector representation is carried out on user attribute features by adopting a mode based on feature mapping, then clustering is carried out on specific event attention users by adopting a fuzzy c-means clustering algorithm, and key attention crowd hierarchies under a certain specific event are identified. According to the method, the focused users of the specific event can be well identified, the user group is clustered and analyzed by adopting a feature mapping method, and the average clustering accuracy is at least 0.06 percent higher than that based on a hard matching rule.

Description

technical field [0001] The invention relates to a method for identifying groups concerned with specific microblog events, and belongs to the technical field of group identification methods. Background technique [0002] With the rapid development of the Internet, social media has quickly become an important carrier for people to express their opinions and views because of their convenience and timeliness. Weibo focuses on user interaction. Users tend to pay more attention to users who have the same hobbies as themselves or have real social relationships. Users with the same or similar attributes often form a group. The group phenomenon is common in social networks. For a specific group, the internal members are closely connected, while the relationship between the internal members and the external members is relatively sparse. [0003] Various user groups are active in Weibo, and internal members are often the publishers, communicators, and followers of a specific event. W...

AI Technical Summary

Problems solved by technology

This algorithm has two main defects: (1) It is impossible to find the boundary of the community accurately; (2) The complexity of the algorithm is high

However, the relevant research on the identification of groups of interest for specific events on Weibo is still limited, and the existing community discovery algorithms have high time complexity and ignore the characteristics of network interaction.

Images