FPIRPQ: Accelerating regular path queries on knowledge graphs

With the growing popularity and application of knowledge-based artificial intelligence, scale knowledge graph data is dramatically increasing. As an essential type query for RDF graphs, Regular Path Queries (RPQs) have attracted increasing research efforts, which explore graphs in a navigational manner. Moreover, path indexes proven successful semi-structured management. However, few techniques can be used effectively practice processing RPQ on large-scale graphs. In this paper, we propose novel indexing solution named FPIRPQ (Frequent Index Queries) by leveraging Frequent Mining (FPM). Unlike existing approaches to RPQs processing, takes advantage frequent paths, are statistically derived from accelerate RPQs. Furthermore, since there no explicit benchmark targeted over yet, design micro-benchmark including 12 basic queries synthetic real-world datasets. The experimental results illustrate that improves efficiency up orders magnitude compared state-of-the-art storage engine.