Consequence Finding in ALC

Research on reasoning in description logics has traditionally focused on the standard inference tasks of (un)satisfiability, subsumption, and instance checking. In recent years, however, there has been a growing interest in so-called non-standard inference services (cf. [1]) which prove useful in the creation, evolution, and utilisation of description logic knowledge bases (KBs). In this paper, we propose consequence finding as a new non-standard reasoning task for description logics. As its name suggests, consequence finding is concerned with the generation of a subset of the implicit consequences of a KB. This topic has been extensively studied in propositional logic (cf. [2]) where it has been shown to be relevant to a number of areas of AI, among them knowledge compilation, abduction, and non-monotonic reasoning. In the context of description logics, we view consequence finding as a tool to enable knowledge engineers and end users alike to better understand and access the contents of a description logic KB. We consider two possible applications: