Frontiers of Information Technology & Electronic Engineering
>> 2015,
Volume 16,
Issue 10
doi:
10.1631/FITEE.1500025
Time-series prediction based on global fuzzy measure in social networks
Department of Computer, Nanjing University of Posts and Telecommunications, Nanjing 210003, China
Available online: 2015-10-12
Next
Previous
Abstract
Social network analysis (SNA) is among the hottest topics of current research. Most measurements of SNA methods are certainty oriented, while in reality, the uncertainties in relationships are widely spread to be overridden. In this paper, fuzzy concept is introduced to model the uncertainty, and a similarity metric is used to build a fuzzy relation model among individuals in the social network. The traditional social network is transformed into a fuzzy network by replacing the traditional relations with fuzzy relation and calculating the global fuzzy measure such as network density and centralization. Finally, the trend of fuzzy network evolution is analyzed and predicted with a fuzzy Markov chain. Experimental results demonstrate that the fuzzy network has more superiority than the traditional network in describing the network evolution process.
京公网安备 11010502051620号
