| Abstract |
| Although generative AI has reshaped human-AI collaboration, empirical research has primarily focused on outcome-based evaluations, neglecting interaction processes. A clear understanding of user-AI interaction mechanisms requires investigating the linguistic speech acts of prompts. Therefore, this study examined how personality traits and task types shape prompt speech act tendencies. Controlled experiments were performed with 40 college students across three tasks. Users’ personalities were clustered using the Big Five Inventory, and prompt logs were classified using GPT, Gemini, and Claude. Binary logistic regression revealed that both personality and task types significantly affected speech act tendencies. Remarkably, conscientiousness and neuroticism emerged as key determinants of these differences. This study provides academic value by shifting the analytical focus to the interaction process using empirical conversation data, serving as a foundational reference for designing personalized human-AI systems. |
|
|
| Key Words |
| Generative AI, Big Five Personality, Task Type, Human-AI Collaboration, Interaction, Prompt, Speech Act, 생성형 AI, 빅 파이브 성향, 과업 유형, 인간-AI 협업, 상호작용, 프롬프트, 화행 |
|
|
 |
|
