Page 31 - 2019

Ph.D.
(Engineering & Technology)
PERFORMANCE ENHANCEMENT OF SPARQL QUERY
OVER THE SEMANTIC WEB

Ph.D. Scholar : Nagar Bhaumik Chhotalal
Research Supervisor : Dr. Naresh D. Jotwani

Regi. No.: 11146051001
Abstract :
The internet is increasingly turning into a global information lake which contains not only
linked documents, but also linked data. Querying linked data efficiently is essential with
the growth of the inked data cloud. Linked data is always in the form of Resource
Description Framework (RDF) format. After the development of different prototypes for
querying this cloud linked structured data, SPARQL protocol and RDF query Language
was developed and accepted by the World Wide Web consortium as the query language
to query Linked data. Considering the growth of linked data, optimization of SPARQL
query is essential as linked data cloud will further grow with more websites converging
towards the Semantic Web.

Tackling the issue of efficient data processing over the Semantic Web, we investigated
the challenges that arises in the context of the standard Semantic Web data format RDF
[rdfs] and, in particular, the PARQL query language for RDF. The goal was to bring
together classical database research and the novel ideas behind the Semantic Web to
enhance the performance of SPARQL query on Semantic Web. With the fast growth of the
Semantic Web, a large amount of resource description framework and ontologies are
reated and published for knowledge sharing and integration for linked open data. SPARQL
query optimization for querying large-scale resource description framework (RDF) triples
is key part of semantic web data management. Performance of SPARQL query over the
Semantic Web is always a concern because of the size of ontology. Before the SPARQL
query is executed, loading the ontology, SPARQL query execution uses in-memory
standard leviathan query processor and retrieve the required results from the ontology
which takes lot of time. There is a possibility of enhancing and improving the
performance of SPARQL query. Due to inherent graph-structure of ontologies, querying
large RDF datasets requires efficient mechanisms to speed up the retrieval of
information. However, these approaches involve many self-joins or cross-table-joins when
processing queries, which greatly slows the query performance.

We proposed an optimization technique on RDF or OWL i.e. by reducing Search Space
using semantic similarity in SPARQL query comparison. This can be achieved by Then
converting the SPARQL query into equivalent graph and use the semantic similarity
measures to compare SPARQL query under execution with already executed SPARQL
query in the system Scope of search space can be reduced in ontology by marking
12