Abstract: The existing distributed parallel compression coding algorithms for RDF data do not consider combining with the ontology file,resulting in encoded RDF data without any semantic information,which is not conducive to the distribu-ted query or reasoning. To solve these problems,a method named SCOM (Semantic Code with Ontology on MapReduce) was proposed to complete the semantic parallel coding for RDF data.Firstly,the algorithm combines the ontology of RDF data to build the class and attribute relationship model.The triple items are encoded and a dictionary table is generated after classifying and filtering triples.Finally,the coding for RDF data with semantic information and regularities is completed.In addition,SCOM algorithm can easily revert the encoded RDF data file to their original file.Experimental results show that SCOM algorithm can achieve the parallel coding of large-scale data efficiently.

Key words: RDF,Ontology,Semantic coding,MapReduce