Integration of Jason Reinforcement Learning Agents into an Interactive Application

The context of this paper is research on the application of agent-oriented programming to experiment with reinforcement learning algorithms. Traditionally, reinforcement learning fits well within the theory of agent systems. Nevertheless, most experimental approaches employ standard software engineering tools, rather than specific agent-oriented technologies developed within the agent community. The goal of our work is to stimulate synergies and development of agent-oriented programming technologies to better fit the context of agent systems research. In particular in this paper we focus on the development of an experimental interactive application that incorporates reinforcement learning agents created using the Jason agent-oriented programming language.