Over the last few years, RDF has been used as a knowledge representation model in a wide variety of domains. Some domains are full of uncertainty. Thus, it is desired to process and manage probabilistic RDF data. The core operation of queries on an RDF probabilistic database is computing the probability of the result to a query. In this paper, we describe a general framework for supporting SPARQL queries on probabilistic RDF databases. In particular, we consider transitive inference capability for RDF instance data. We show that the find operation for an atomic query with the transitive property can be formalized as the problem of computing path expressions on the transitive relation graph and we also propose an approximate algorithm for computing path expressions efficiently. At last, we implement and experimentally evaluate our approach.
展开▼
