Method for identifying network opinion leaders based on textual tendency analysis

Abstract

The present invention discloses a method for identifying network opinion leaders based on textual tendency analysis. The method specifically comprises: firstly, using an index analysis method to construct an index system of network opinion leader identification, and identifying potential opinion leaders; secondly, adding textual tendency analysis, introducing the Word2Vec algorithm, and removing negative emotions overweighed "pseudo opinion leaders"; and thirdly, carrying out comparison and verification on effects, comparing the identified opinion leaders with the WeiboRank opinion leader algorithm to verify the validity and credibility of the proposed algorithm in the present invention. According to the method disclosed by the present invention, three types of typical opinion leaders can be identified, all stages of public opinion development from the origin of emergency, the continuous expansion of the influence, to the change and stability of the attitude of the public opinions are covered, a theoretical basis is provided for monitoring the evolution of the public network opinion and capturing and predicting the livelihood and the opinion of the people, and the method can be taken as an important measure to prevent group emergencies.

Description

technical field [0001] The invention relates to an identification method, in particular to an identification method for network opinion leaders based on text tendency analysis. Background technique [0002] With the development of the Internet, the penetration rate of social networking sites has also been increasing. This network of interpersonal communication represented by microblogs and forums has gradually penetrated into people's lives, and has become a way for the masses to express their beliefs, attitudes, and opinions on social phenomena and social problems. One of the great platforms for opinions and sentiments. This kind of online expression, dissemination and interaction of public opinion is called Internet public opinion. Due to the spontaneity and freedom of the Internet, Internet public opinion contains not only rational speech, but also extreme speech and even rumors, so it is necessary to take certain measures to monitor and warn Internet public opinion. In...

AI Technical Summary

Problems solved by technology

[0003] Existing methods for identifying opinion leaders can rely on manual web browsing, comment collection, and opinion analysis to identify opinion leaders. However, this method is not only inefficient but also has inconsistent human evaluation standards. Strengthen the research on relevant information technology, and form an automatic network opinion leader identification system

However, the existing automated opinion leader identification systems mainly rely on social network analysis and cluster analysis methods. These methods mainly focus on the extraction and utilization of bloggers’ personal information and comment forwarding relationships, and lack the identification mechanism of public comment attitudes. It may lead to the appearance of bloggers among the selected opinion leaders who can trigger extensive discussions among netizens, but the contents of the discussions are all opposing or even abusive voices

These "pseudo-opinion leaders" do not meet the definition of opinion leaders, nor can they fulfill the basic role of opinion leaders

Images