[2E4-OS-1a] OS-1 (1)

[2E4-OS-1a-02] Analysis of Human Network Generation Process in SNS based on Difference of User's Values to SNS Interation

〇Keito Ishii¹, Yutaro Miura¹, Fujio Toriumi², Toshiharu Sugawara¹ (1. Waseda Univ., 2. Univ. of Tokyo)

Keywords:Social Networking Services, Multi-agent Simulation, Complex Networks, Meta-norms Game, Reinforcement Learning

This paper analyzes the effects of differences in values for social networking services (SNS) on the behaviors in SNS using the SNS-norms game. We think that such difference seems to be affect the attitude to or behavior in SNS activities, which was not taken into account in the previous studies. Therefore, we introduced such a difference into the model of SNS using a weighted directed graph by defining three types of behaviors in SNSs;positive, normal, and passive behavior. Then, we analyzed how it affects the degree of closeness to build friend relationships.Our simulated experiments indicates that there was differences in the behavior of posting articles, and there was a tendency for active people to be tied together and for passive people to be hardly connected each other.