Behavioral analysis in social networks: an approach based on intelligent system

• Thiago R.P.M Rúbio CEFET-MG
• Rogério Martins Gomes CEFET-MG
• Adriano C.M. Pereira CEFET-MG

Resumo

This paper presents a new methodology to classify Twitter users based on Artificial Neural Networks and Fuzzy Logic. Simulations are carried out using a SOM (Self Organized Maps) neural network for classifying users into four distinct groups: (0)Unimpressive User; (1)Desired User: Follower; (2)Desired User: Follower and Publisher; (3)Desired User: Publisher. The proposed methodology was validated through an autonomous agent, whose interactions with others were modeled by means of Fuzzy Inference System. The results obtained show that neural networks can be used for user classification in social networks, and we observed that the interaction of the agent with other users occurred in a transparent way, i.e., showing typical behaviors of real users.This paper presents a new methodology to classify Twitter users based on Artificial Neural Networks and Fuzzy Logic. Simulations are carried out using a SOM (Self Organized Maps) neural network for classifying users into four distinct groups: (0)Unimpressive User; (1)Desired User: Follower; (2)Desired User: Follower and Publisher; (3)Desired User: Publisher. The proposed methodology was validated through an autonomous agent, whose interactions with others were modeled by means of Fuzzy Inference System. The results obtained show that neural networks can be used for user classification in social networks, and we observed that the interaction of the agent with other users occurred in a transparent way, i.e., showing typical behaviors of real users.