Horn-Rule Based Compression Technique for RDF Data

With the growing number of RDF datasets being published, RDF compression techniques have attracted a lot of research attention. A number of structural compression techniques exist that take into account the structural redundancies such as blank nodes. Few other works focus on providing a compact representation. In our work, we utilize the various semantic associations that can be learned from RDF graphs to compress them. We mine logical Horn rules from the RDF datasets and utilize them for achieving better compression. The technique employed is to store just the triples matching the antecedent part. We delete the triples that match the head part of the rules, as they can be inferred by applying the rules. The experimental evaluation of our approach shows that greater compression can be achieved compared to the existing technique.