Research Interests

I am interested in the reasoning capabilities of agents, both human and artificial. My research is currently focused on inconsistency tolerance, reasoning with ontologies, and dealing with preferences.

Belief change (which I take to include belief revision and belief merging) is a specific approach to inconsistency tolerance to which I have devoted a sizable chunk of my time. A good reference is Sven Ove Hansson's A Textbook of Belief Dynamics: Theory Change and Database Updating. Although a bit dated by now, Peter Gärdenfors's Knowledge in Flux: Modeling the Dynamics of Epistemic States is also well worth having a look at. For a quick introduction to belief change basics, have a look at the following two papers:

  1. Hirofumi Katsuno and Alberto Mendelzon. Propositional knowledge base revision and minimal change. Artificial Intelligence, 52(3), 263-294, 1991.
  2. Adnan Darwiche and Judea Pearl. On the logic of iterated belief revision. Artificial Intelligence, 89(1-2):1-29, January, 1997.
Other approaches to inconsistency tolerance that I am interested in include nonmonotonic reasoning and paraconsistent reasoning.

Ontology research has the goal of providing meaning to information in a way that can be understood by both humans and machines. Dieter Fensel's Ontologies: A Silver Bullet for Knowledge Management and Electronic Commerce is a good starting point. A particularly exciting application of ontologies is the development of the Semantic Web, the next generation Web which will enable software agents to reason about Web content. For a look at the vision of the Semantic Web as expressed by Tim Berners-Lee and others in their 2001 Scientific American article, go here. For more information on the Semantic Web, go here and here. For a roadmap of the Semantic Web, go here. And for what the Semantic Web is not, go here.

I am mostly interested in the reasoning aspects of ontologies. One of the most successful formalisms for ontology reasoning is the class of description logics. The Description Logics Handbook by Franz Baader and others is an excellent introduction to the topic. For a comprehensive online resource about description logics, go here.

My interest in preferences is mostly concerned with cases where they can be expressed qualitatively, usually as a binary relation of some kind. A good example is voting, where voters get to indicate their preferences for the available candidates, which are then combined in some way to determine the most preferred candidate. For an introduction to preferences, and how they are be dealt with, it is a good idea to consult the vast literature on game theory and social choice theory. My interest lies in determining to what extent existing results can be applied to settings such a belief merging, agent negotiation and constraint satisfaction.

Not surprisingly, perhaps, these areas of interest overlap substantially. For example. it turns out that the specification of preferences is an excellent way of guiding inconsistency tolerance. And inconsistency tolerance, in turn, has been identified as an important problem to be solved if the Semantic Web is to become a reality. For more details, go to to my link for potential PhD students.

My published research has so far focused on what is usually referred to as normative reasoning, broadly defined as the type of reasoning an agent ought to exhibit, as opposed to the type of reasoning that humans frequently do exhibit. Human reasoning is more the domain of Cognitive Science than Artificial Intelligence, and while the two types of reasoning are related they are, in my opinion, not exactly the same. In this respect Mark Johnson and George Lakoff's Women, Fire and Dangerous Things is must-read for anyone interested in the topic. And for all you Platonists out there, so is Where Mathematics Comes From: How the Embodied Mind Brings Mathematics into Being by George Lakoff and Rafael E. Núňez.

Last update:

[]Tommie's home page