Loading RDF/OWL File into Oracle NoSQL Database Using a Bulk Loading and Parallelization Techniques

Abstract

Processing and loading large RDF/OWL data files often involves the transfer of large amounts of data from source operational systems to the target data store. It requires a proper loading techniques and more expressive methods to ensure the RDF/OWL data files are correctly loaded into the database, or the loading process will end up finishing all memory space which resulted in performance deterioration in the entire application. The Bulk loading is one of the predominant scientific techniques and with fastest loading method to load large amounts of data into the database; however, this study has found out that the bulk loading alone, cannot handle triples files containing object values with more than 4000 bytes. To solve these tidy problems, we propose to implement bulk loading techniques concurrently with parallelization methods to load RDF/OWL data file into the database. To accomplish this goal, the Oracle NOSQL database is chosen as the backend persistent storage for the RDF/OWL data files. In addition, loading model and techniques are provided to utilize the loading process, and finally algorithms, methods and parallelization techniques are given with the aim of increasing loading performance of RDF/OWL within the Oracle NOSQL database.