Towards Enhancing Empathy Through Emotion Augmented Remote Communication

Abstract

Our ability to empathize stands at the core of our existence as humans. It has dramatically contributed to our evolution as species and remains key to our relationships with family, friends, and colleagues. Yet, we sometimes find it difficult to empathize, more so in the online space increasingly dominating our interactions. In this work, we explore the potential of videoconferencing platforms as avenues for augmenting empathy. We have developed Project Us, a system that uses machine learning to analyze interlocutors' tone of voice and facial expressions during online interactions, compiles their emotional valence and feeds it back to each participant as a real-time visual cue. Tested with 40 users, Project Us shows the emergence of empathy-related behaviours, such as a higher awareness of own and other's emotions, and an increased use of pro-social language, as evaluated through quantitative (natural language processing) and qualitative methods. We also analyze ethical challenges of empathy augmentation.