Similarity-Based Query Caching

With the success of the semantic web infrastructures for storing and querying RDF data are gaining importance. A couple of systems are available now that provide basic database functionality for RDF data. Compared to modern database systems, RDF storage technology still lacks sophisticated optimization methods for query processing. Current work in this direction is mainly focussed on index structures for speeding up the access at triple level or for special queries. In this paper, we discuss semantic query caching as a high level optimization technique for RDF querying to supplement existing work on lower level techniques. Our approach for semantic caching is based on the notion of similarity of RDF queries determined by the costs of modifying the results of a previous query into the result for the actual one. We discuss the problem of subsumption for RDF queries, present a cost model and derive a similarity measure for RDF queries based on the cost model and the notion of graph edit distance.