ChatNekoHacker: Real-Time Fan Engagement with Conversational Agents
Takuya Sera, Yusuke Hamano
TL;DR
This paper tackles real-time fan engagement for musicians by deploying a low-resource conversational-agent system that combines Amazon Bedrock Agents, Unity-based 3D livestreams, and VOICEVOX Japanese TTS to realize two personas, Neko-Chan and Hacker-Chan, during live YouTube broadcasts. It ingests YouTube comments and leverages a knowledge base built from social posts and Wikipedia to generate autonomous dialogue with Kansai dialect prompts, testing the approach in a one-hour, 30-participant study. Regression analysis shows that perceived fun in conversation is the key driver of increased interest in the artists, with an adjusted $R^2$ of 0.56 and a $p$-value of 0.00005, while other factors are not significant. The findings highlight the potential of entertaining, interactive broadcasts for fandom growth, while also noting limitations such as latency and misinformation risk, and outlining directions for broader response diversity and tighter fact-checking.
Abstract
ChatNekoHacker is a real-time conversational agent system that strengthens fan engagement for musicians. It integrates Amazon Bedrock Agents for autonomous dialogue, Unity for immersive 3D livestream sets, and VOICEVOX for high quality Japanese text-to-speech, enabling two virtual personas to represent the music duo Neko Hacker. In a one-hour YouTube Live with 30 participants, we evaluated the impact of the system. Regression analysis showed that agent interaction significantly elevated fan interest, with perceived fun as the dominant predictor. The participants also expressed a stronger intention to listen to the duo's music and attend future concerts. These findings highlight entertaining, interactive broadcasts as pivotal to cultivating fandom. Our work offers actionable insights for the deployment of conversational agents in entertainment while pointing to next steps: broader response diversity, lower latency, and tighter fact-checking to curb potential misinformation.