Analyzing the query performances of description logic based service matching using Hadoop

Extracting accurate information for the user input is the main drawback for information retrieval system in web. To overcome the matching problem, a description logic based matching technique is proposed. The description logic consists of TBox and ABox to separate the individual and universally quantified properties. The user input can be matched against ontology database and allows us to formalize information of domain as classes and instances. The DL-structured data format is processed through Hadoop to obtain the optimal matching service platform. The unstructured information is stored as files in the Hadoop Distributed File System and MapReduce operation is performed for simplified development of programs. The SPARQL query is used for querying data stored in Mongodb. Mongodb has collection of RDF datasets and are processed for matching using SPARQL query. The resultant dataset for the user input is given in the form of text document. The performance of the matching system is improved in terms of query response time and iteration time. Comparing with the existing system proves that the proposed system provide an optimal solution with minimal search time and maximum accuracy.