EMPATHICSTORIES++
收藏资源简介:
EMPATHICSTORIES++是一个多模态数据集,旨在研究个人经历分享中的同理心。该数据集由麻省理工学院创建,包含41名参与者在一个月内的53小时视频、音频和文本数据,这些数据是在自然环境中与AI代理交互时收集的。数据集不仅包括原始的多模态数据,还包括心理测量调查,以推动计算同理心研究。EMPATHICSTORIES++是首个关注同理心的纵向数据集,通过在参与者家中部署社交机器人,捕捉了自然、同理心的故事讲述交互。该数据集的应用领域包括开发具有同理心的AI系统,以及在真实世界环境中理解人类同理心的复杂性。
EMPATHICSTORIES++ is a multimodal dataset designed to study empathy in personal experience sharing. Developed by the Massachusetts Institute of Technology (MIT), this dataset contains 53 hours of video, audio and text data from 41 participants over a one-month period, collected during natural in-the-wild interactions with AI Agents. In addition to raw multimodal data, the dataset also includes psychometric surveys to advance computational empathy research. EMPATHICSTORIES++ is the first longitudinal dataset focused on empathy, capturing natural and empathetic story-telling interactions by deploying social robots in participants' homes. Its application areas include the development of empathetic AI systems and the understanding of the complexity of human empathy in real-world environments.

- 1EmpathicStories++: A Multimodal Dataset for Empathy towards Personal Experiences麻省理工学院 · 2024年



