The Design and Implementation of Minimal RDFS Backward Reasoning in 4store

Abstract

Abstract. This paper describes the design and implementation of Minimal RDFS semantics based on a backward chaining approach and implemented on a clustered RDF triple store. The system presented, called 4sr, uses 4store as base infrastructure. In order to achieve a highly scalable system we implemented the reasoning at the lowest level of the quad store, the bind operation. The bind operation runs concurrently in all the data slices allowing the reasoning to be processed in parallel among the cluster. Throughout this paper we provide detailed descriptions of the architecture, reasoning algorithms, and a scalability evaluation with the LUBM benchmark. 4sr is a stable tool available under a GNU GPL3 license and can be freely used and extended by the community 1 .

Statistical Analysis of the owl:sameAs Network for Aligning Concepts in the Linking Open Data Cloud

“…The first scenario explored the gain we have when we take into account (or not) the concept hierarchy when we compute the cardinality of the two sets, A and B respectively, as defined in Section 2. For doing this we used 4sr, a reasoner that efficiently implements the reasoning over rdfs:subClassOf axioms (sc 0 and sc 1 rules in [17]). …”

Large-Scale Storage and Reasoning for Semantic Data Using Swarms

“…As shown in the previous section, 2 Test data set available at: http://beast-reasoning.net/a.nt 3 FOAF vocabulary specification: http://xmlns.com/foaf/spec the potential amount of data that can be handled is directly dependent on the number and storage capacity of the nodes and the throughput of the network connections between them. Hence, a larger data set would not yield additional insight into the system's behavior.…”

“…Salvadores et al [2] have added support for reasoning for minimal RDFS rules in the distributed RDF storage system 4store. They include backward-chaining into the basic retrieval operation in a way very similar to the method presented in [29].…”

“…Many previous approaches rely on central instances orchestrating the reasoning process, e.g. in [2], which is undesirable since these central nodes are not protected against failure per se, and their failure would inhibit the entire system to perform its reasoning tasks. Alternatively, schema information is replicated in all participating nodes, which is unfeasible in cases where the schema information is very large.…”

QueryPIE: Backward Reasoning for OWL Horst over Very Large Knowledge Bases

“…In the context of RDF stores, in [10], backward-chaining reasoning for RDFS on 4Store is presented. The authors show how they perform RDFS reasoning on their architecture but do not report on more complex inferencing than RDFS.…”