Abstract
The management of inconsistency between multiple
viewpoints has become a central problem in the
development of large software systems. In this paper we
argue that the same problem occurs in the development of
the semantic web, and indeed that this is the central
issue in semantic integration. A common approach is to
attempt to remove inconsistencies, if necessary by
discarding problematic information. We argue that this
approach will greatly limit the utility of the semantic
web. Instead, we argue the need for formal reasoning
systems that can tolerate inconsistent information. A key
observation is that the problem is essentially one of
model management. Rather than seeking to build a single
consistent model, the challenge is to reason about the
inconsistencies and dependencies between a set of interrelated
partial models, and to use paraconsistent logics
when reasoning with information from inconsistent
ontologies.
Users
Please
log in to take part in the discussion (add own reviews or comments).