I try to build intelligent systems, known as cognitive agents, which implement the "perception," "reasoning," "action" cycle that constitutes the decision-making process. These intelligent agents must work as a team to solve communication tasks.  The goal is to build a common language to refer to a situation, objects, etc. This is done in the form of a game according to an established protocol. An agent chooses a topic of conversation and must ask themselves: what do I need to say so that the other person knows what I'm talking about? They have perception abilities but see the world or segments of the world without labels and have no linguistic system.

The AI method for language games that I am working on involves a multitude of agents who will introduce many ideas and concepts in order to ultimately find a common language and thus a robust, adaptive, and productive communication protocol.

I enjoy working in a team. It takes more than one person to learn! It is we, as individuals interacting with others, who have ultimately built our language. In fact, I am like my agents.

Prof. Katrien Beuls, Faculty of Computer Science, NaDI Institute