Partitioning ABoxes Based on Converting DL to Plain Datalog

To make ABox reasoning scalable for large ABoxes in description logic (DL) knowledge bases, we develop a method for partitioning the ABox so that specific kinds of reasoning can be performed separately on each partition and the results trivially combined can achieve complete answers. Our method applies to SHIQ(D) knowledge bases. It first converts a DL knowledge base KB to a plain datalog program H(KB), and then computes the least fixpoint of the definite part of H(KB) while generating ABox partitions. Its time data complexity is polynomial in the ABox size, under some general assumption on concrete domains. Experimental results further demonstrate the advantages