Keywords: conversational memory, conversational memorability, social signal processing, multi-modal interaction, group dynamics, multiparty interaction
Working in a group requires understanding each other's perspectives for better group unity and meeting efficiency. This can be difficult, leading to misunderstandings and repetitive discussions. Research on multimodal interactions has focused on topics such as detecting dominance, improving engagement and boosting creativity, but tracking what each member or the group as a whole finds memorable from the conversation is a widely unexplored area. In this project Maria Tsfasman and Catharine Oertel are exploring what participants recall from meetings and their (non)verbal cues can be used to predict the most memorable moments. For that, they recorded a multimodal longitudinal meeting corpus (MEMO) annotated by what participants remembered from each conversation and why. The first results show that memorable conversations are strongly indicated through group eye gaze behaviour. Watch the video and the paper describing our first results in more detail.
Tsfasman, M., Fenech, K., Tarvirdians, M., Lorincz, A., Jonker, C., & Oertel, C. (2022). Towards creating a conversational memory for long-term meeting support: predicting memorable moments in multi-party conversations through eye-gaze. In ICMI 2022 - Proceedings of the 2022 International Conference on Multimodal Interaction (pp. 94-104). (ACM International Conference Proceeding Series). Association for Computing Machinery (ACM). https://doi.org/10.1145/3536221.3556613