Arthur and Bella: multi-purpose empathetic AI assistants for daily conversations

The paper presents a novel approach to developing an embodied conversation agent (ECA) that is capable of displaying empathy toward its human partner during interactions. The virtual agents are equipped with both memory and empathy capabilities, with the main focus being on modeling an empathy model associated with the ECA's memory. The paper presents the proposed model of empathy, as well as its connection with memory, and evaluates how this relationship affects the user's experience (UX) through experiments with volunteers who participated in long and short-term interactions. The results of the experiments show that the association of memory with the empathy model makes interactions with embodied conversational agents more enjoyable and to the user. This suggests that the ECA's ability to display empathy can have a positive impact on the user's experience, which is an important factor to consider when designing conversational agents for various purposes. Overall, the paper presents an interesting and valuable contribution to the field of embodied conversational agents and human-computer interaction. The incorporation of empathy and memory capabilities into an ECA has the potential to improve the user's experience and make interactions with machines more human-like.