DRUMS II is a continuation of the DRUMS I Basic Research Action (3085), in which defeasible reasoning, uncertainty and incompleteness were studied. The overall objective is to investigate symbolic and quantitative methods of representation and reasoning able to deal with knowledge that is uncertain and defeasible. Six topics have been selected that are of fundamental importance in the domain of defeasible and uncertain reasoning:
- Belief change Organisation of the body of existing results for modelling belief changes so as to assess what is common to these belief change schemes.
- Non-monotonic deduction Integration of four general frameworks for deduction, namely Gentzen systems, labelled deductive systems, the axiomatisations of the consequence relation, and category theory; in order to provide appropriate insights into the nature and mechanism of non-monotonic deduction.
- Inconsistency in reasoning Development of a logic-based model that accounts for inconsistency in reasoning.
- Abduction Clarification of the relationships and actual differences of numerical, logical and machine learning perspectives of abduction.
- Algorithmic problems Study of the algorithmic problems of uncertain reasoning, taking as general a viewpoint as possible.
- Dynamic reasoning with partial models Study of the dynamics of non-monotonic reasoning using information structures as a vehicle for representation.
The symbolic and quantitative methods of representation and reasoning able to deal with knowledge that is uncertain and defeasible have been investigated. The 6 topics under study are: belief change, nonmonotonic deduction, inconsistency in reasoning, abduction, algorithmic problems and dynamic reasoning with partial models.
The major achievement is the integration of both the symbolic and numerical approaches for handling uncertainty problems, bridging the gaps between 2 usually diametrically opposed approaches. Strong cross fertilization is apparent. The results will be presented in a series of handbooks.
APPROACH AND METHODS
The project is centred on topics, and not on methods as was the case in DRUMS I. Partners are distributed amongst the six topics, creating a context favourable for strong integration between numerical and symbolic approaches as well as between various schools of thought.
The project will provide a better understanding of the models proposed to represent and to handle uncertainty and defeasible reasoning, two topics that seem to have become major issues in the development of information systems.
9220 Aalborg East
SW7 2BZ London
E1 4NS London
1007 MC Amsterdam
221 00 Lund