SPARQL2Hive: Translating SPARQLqueries on Hive using A meta-model-based techniques

Abstract

Abstract— As a result of correctly querying RDF data, new problems have been created by the extension of Web documents. Traditional RDBMS can successfully adapt and search for scattered data. With the development of Hdfs and its use of the Mapreduce Model with the Hive database engine, the ethics of data collection and retrieval have altered. In this work, we introduce SPARQL2Hive, a MapReduce-based, cost-effective SPARQL querying programme that enables ad hoc SPARQL querying parsing on massive RDF networks. Instead of directly translating from one language to another, SPARQL2Hive uses Hive's parser as a bridge between SPARQL and MapReduce. A data warehousing platform called Hive is built on top of parallel processing and uses Hadoop to search computers. This extra degree of virtualization renders our method agnostic of the present version of Hdfs, ensuring interoperability with potential updates to the Hadoop platform since they will be handled by the underpinning Hive level. Our strategy is to employ the SPARQL and Hive conceptual and suggest a conversion between them using the ATL vocabulary. SPARQL2Hive is compared to Apache hadoop SPARQL solutions.