Designers should enhance the chatbot's character profile to ensure consistency in interactions, thus maintaining children's trust and encouraging emotional sharing.
About this paper
The author introduces ChaCha, a chatbot that aids children in sharing personal events and emotions by combining a state machine and large language models.
An exploratory study with 20 children found that ChaCha was perceived as a close friend and effective in prompting emotional sharing on various topics.
Here are some methods used in this study:
Which part of the paper did the design guideline come from?
“Based on the analysis of the participant debriefs, we identified how they perceived ChaCha's persona and conversation skills about sharing their emotions. 6.3.1 ChaCha's Persona. Most participants perceived ChaCha as a close friend with whom children can share their emotions and secrets. Some participants perceived ChaCha as older or even younger than themselves, even though we set ChaCha the same age as each participant. For instance, C1 thought ChaCha would be an older friend since ChaCha (...)” (Section 6.3: RESULTS)
Seo, W., Yang, C., & Kim, Y.-H. (2024). ChaCha: Leveraging Large Language Models to Prompt Children to Share Their Emotions about Personal Events. Proceedings of the CHI Conference on Human Factors in Computing Systems, 24, 1–20.