[2I1-GS-5a] エージェント：ゲーム理論

[2I1-GS-5a-05] Proposal of activation method by introducing articles on SNS

〇Yutaro Usui¹, Fujio Toriumi², Toshiharu Sugawara¹ (1. Waseda University, 2. University of Tokyo)

Keywords:Simulation, Multi-agent, Genetic algorithm

In this study, we propose a game-theoretic SNS model that extends the conventional SNS-norms game by introducing the mechanism of providing to article posters a number of related articles that are already posted by unknown third persons. In the SNS-norms game, users post articles, which are read and commented by other users, and the contributors of the articles reply to the comments. In this study, we add the structure of introducing related information (specifically, introducing similar articles) as a bonus for the post to this SNS model, in order to activate the SNS. This will increase the reward of substantive article submissions, and also increase the opportunity to be commented on articles, leading to a further increase in reward. As a result of the simulation-based experiment, we found that the proposed model, the mechanism for providing related articles effectively increased the rate of article submission for users with a small number of links. The mechanism of commenting back on related articles also contributed slightly to the increase in the posting rate.