My PhD research is looking at the role of agents on the World Wide Argument Web (WWAW), the infrastructure for which will be built by the Dialectical Argumentation Machines project.

My ultimate goal can be encapsulated as providing agents with the "skills" needed to be virtual participants on the WWAW, beyond what is currently offered by systems such as MAgtALO. In MAgtALO, the agents represent real people, and so each knowledgebase is limited to the existing knowledge elicited from the person they represent.

I'm looking to extend this concept so that agents are entities in their own right - so they can find new knowledge, form teams and actively engage in discussions, without any moral or ethical issues related to stating an opinion not held by the original contributor of the knowledge.

Currently, I'm looking at belief-based team formation, with an agent's beliefs being derived from Dung-style argumentation frameworks (AFs). Using AFs allows me to focus on the architecture required to facilitate team formation, without having to worry about issues related to linguistic arguments. There will come a time, of course, where "real" arguments will be integrated, but that's (relatively speaking) a long way off.

I'll update this page as I move into other areas.