Investigations in Possibilistic Reasoning and Multi-Context Systems
Many formalisms with various expressive power and reasoning properties have been proposed in knowledge representation (KR). Possibilistic logic is a weighted logic for dealing with incomplete and uncertain information by assigning weights to propositional formulas. A possibilistic knowledge base (KB) is a finite set of such formulas. The importance of incorporating possibilistic reasoning has been recognised by some researchers. However, some issues are still unsolved yet: (1) While
... c answer set programming (PASP) provides an elegant framework for combining possibilistic logic and answer set programming (ASP), PASP is unable to reason with inconsistent possibilistic logic programs; (2) The problem of revising a possibilistic KB by a possibilistic formula has been studied by several researchers. However, existing approaches to possibilistic revision suffer from the so-called drowning effect. Moreover, revisions by certain and uncertain formulas are handled separately and most revision operators are defined only for certain inputs; (3) Multi-context systems (MCSs) provide a unifying framework for representing distributed and heterogeneous information sources in an integrated way. An MCS is a set of component KR systems called contexts, which are interconnected via so-called bridge rules. It is still unclear how to incorporate possibilistic reasoning in MCSs. Besides, while MCSs address the heterogeneity of knowledge formalisms in an elegant theoretical manner, research on distributed and parallel reasoning systems for MCSs is still in its early development. Thus a more efficient reasoning framework is needed.