EpiRL project aims at investigating the combination of epistemic planning and reinforcement learning (RL), by proposing new algorithms that are efficient, adaptive, and capable of computing decisions relying on theory of knowledge and belief. We expect from this approach an efficiency in the generation of epistemic plans, while decisions made RL algorithms will be explainable. Moreover, the algorithms of EpiRL will be tested and evaluated within a real application that exploits autonomous agents. The project will address the weaknesses of both epistemic planning and RL: on the one hand, existing epistemic planning algorithms are costly, do not adapt to the environment, and concepts are hand-crafted and are not learned; on the other hand, in reinforcement learning, agents adapt to their environments but are unable to reason about beliefs of other agents. The newly developed algorithms will combine the strengths of both fields. We propose four workpackages: 1. Study representations of states 2. Develop RL algorithms 3. Study representations of policies 4. Validating our algorithms with our industrial partner DAVI. In particular, we aim at developing a debunking chatbot whose use case will apply to raising awareness about environmental issues.
The general objective of the CoPains project is to develop an artificial agent which is capable of: (i) inferring the cognitive and affective states of the human user from her observable behaviors, (ii) influencing and persuading the human user to believe something and/or to behave in a certain way, and (iii) interacting with the human user through multimodal communication including textual expressions, facial expressions and gestures. The envisaged applications are in the area of caregiver assistance for elderly people and nutrition in which an artificial system such as an embodied conversational agent (ECA) has to take care of the user's health and wellbeing. In order to achieve its objectives, CoPains is expected to exploit theoretical and empirical approaches by combining, in a rather innovative way, corpus-based analysis with formal methods from different areas of artificial intelligence (AI). This includes logic, planning, sentiment analysis, and data-mining.