Large Scale Fuzzy RDF Reasoning Engine

In this paper , we design fuzzy pD * semantics , generalized pD * semantics to support the fuzzy RDF using OWL vocabulary on the reasoning and the development of a large-scale inference algorithm . We first define fuzzy RDF graph and fuzzy pD * translation . Then we have listed a set of fuzzy pD * inference rules , and define a set of triples optimal degree of sector ( Best Degree Bound BDB ) . We demonstrate that it BDB exist for any triples . We then generalized by introducing BDB part and completely pD * closure concept . We prove that the partial fuzzy closure of the existence of polynomial time computability . Finally, we demonstrate the soundness and integrity of the results of the reasoning relationship , and prove that the consistency check is P , the inference problem is NP- complete problem , and when the target is the underlying reasoning problems , compared with P . Therefore using fuzzy semantic extension pD * semantics and will not increase the computational complexity. MapReduce framework has proved very efficient for handling data intensive tasks . Previous work has to MapReduce applied to large-scale reasoning for pD * semantics , and show very promising results . We hope that one step further, consider the fuzzy pD * semantics large-scale inference problem . To our knowledge, this is the first study with MapReduce Research fuzzy reasoning . Although most optimized for pD * semantics equally fuzzy pD * , when we deal with fuzzy information , or encountered special difficulties . We first identified these difficulties , and then design a solution for each difficulty . In addition , we have realized a prototype system for testing . The experimental results show that the running time of our system can WebPIE comparable , while the latter is the most state - of - the - art processing pD * semantics mass inference engine .