Social network opinion leader sorting algorithm based on multidimensional characteristic analysis

Abstract

The invention discloses a social network opinion leader sorting algorithm based on multidimensional characteristic analysis, comprising: a, analyzing user characteristics in the terms of structure, behavior and topic, and calculating their characteristic values; b, using the entropy weight method to calculate weights of the different characteristics; c, acquiring characteristic values of S(u), B(u) and T(u) in conjunction with the step a, acquiring weight wj in conjunction with the step b, calculating value of F(u) function according to the MFP (mining frequent patterns) algorithm, and generating a sorting result according to the value of the F(u) function. The social network opinion leader sorting algorithm based on multidimensional characteristic analysis comprehensively considers multiple characteristics that are influential to information spread, including structural characteristic, behavior characteristic and topic characteristic; the entropy weight method is used to calculate weights of their characteristics, the influence of subjective judgment is avoided, and sorting in the final result is more accurate.

Description

technical field [0001] The invention relates to the field of social networks, in particular to an algorithm for ranking social network opinion leaders based on multidimensional feature analysis. Background technique [0002] A social network is a relatively stable social structure formed by the mutual interaction among social members, with a complex network structure and a dynamic information dissemination mechanism. With the increase of the scale of social networks and the enhancement of standardization, the current large-scale social networks often have the characteristics of active users, relatively consistent interaction methods between users, and relatively complete user characteristics and behavior records, which brings great benefits to the analysis and research of social influence Great convenience. Therefore, there are currently many practical applications of social networks in social networks. Another important application of social influence: opinion leader mini...

AI Technical Summary

Problems solved by technology

[0006] The problem to be solved by the present invention is: to solve the problem of low ranking accuracy in the current opinion leader algorithm, comprehensively consider the characteristics of the three levels of structure, behavior and theme, and use the entropy weight method to calculate their weights separately, avoiding the influence of subjective judgments

Images