[3F4-GS-10k] AI応用：行動分析

[3F4-GS-10k-02] Demonstration experiments and user behavior analysis at online events recommended by multiple anthropomorphic agents

An attempt to increase value by using an interactive service that connects the real world and the Internet

〇Kazuya Yamashita¹, Mari Furuta^1,2, Megumi Inoue¹, Kota Takaoka¹, Koichi Shinoda², Yoichi Motomura¹ (1. Artificial Intelligence Research Center (AIRC),National Institute of Advanced Industrial Science and Technology (AIST), 2. Tokyo Institute of Technology)

Keywords:Service system, Anthropomorphic agent, Bayesian network

We have developed an interactive service system that can seamlessly handle the real and the Internet with the aim of increasing the value of stakeholders involved in the event, and at the Science Agora (2020), which became the first online event due to the influence of COVID-19, at the Real Science Museum. I tried to transfer the obtained persona and realized it in the form of "Web tour". In the Web tour, we will introduce a system that allows multiple anthropomorphic agents to recommend exhibitions and chat, and what kind of anthropomorphic agents can provide high value to various visitors at events that cannot be guided by actual people. I examined. This time, the type of anthropomorphic agent was created by combining attributes and tone, and it was possible to clarify which anthropomorphic agent's recommendation enhances the persuasive effect. As a result, the applicability of this system to other fields that emphasize persuasive effects, such as the health field, is also discussed.