Towards Agent-based Large-scale Decision Support System: The Effect of Facilitator

Abstract

Good discussions are essential for group decisions, especially when the number of people in thea group has many people is large. So, it is important to Pprovidinge good support is critical for having establishing and maintaining coherent discussions that avoid some of thesuch anti-social bad behaviors, like as flaming, that have, which has been observed in some large discussion groups. We have developed a large-scale online decision support system that has facilitator support functions, and deployed it in case studies for several real-world online discussion supports as case studies. In this paper, weWe propose a facilitator-mediated online discussion model in order to lead discussions to in a better direction for ato reach decisions. Our extreme ultimate goal is an to realize automated facilitator agent that can adequately leadhelp participants to achieve reach reasonable decisions. In reality, online discussion is often fails plagued byinto "flaming,", which is the act of posting or sending offensive messages during a discussion. Such flaming phenomena have been focused on as anti-social bad behavior of in online discussion forums. After several cases studies, we learned several lessons. Critically, The most important achievement is that in any all of our social experiments, no flaming has not been observed in our facilitator-mediated decision support system. Also, we obtained Our some insights also suggest in whichthat the social presence of a facilitator would have largegreatly aeffect for participants’ behavior.