Design and Implementation of a Proposed Model for a Semantic-Based Search Engine

Abstract

Conventional search engines that depend upon keyword-based retrieves
have resulted in inaccuracies and non-relevance with respect to context. This
research aims to overcome the drawbacks of traditional search methodologies
by presenting a semantic model based on search engine that improves
information retrieval using organized, ontology-oriented approach. This
research aims to attain a model that can understand the queries and give
accurate replies based on context. According to the company, this process
includes building an ontology in OWL (Ontology Web Language) that allows
and helps a semantic search engine understand nuanced academic
relationships within university course data. This model has been implemented
and tested in a well-controlled academic setting where the dataset was made
to befit its exact performance.The results obtained show that the proposed
model is indeed able to perform much better than regular search engines in
terms of precision, relevance and finding time. This study provides, to the
best of our knowledge, one of the first reported examples using domainspecific relatedness measures for rapid semantic search in specialized fields
where accurate data retrieval is a must-have. Major contributions also include
an efficient implementation framework that can be scaled up and adapted
towards specific scenarios involving large medical databases which use more
complex terminologies such as UMLS, ICD terms (e.g., BCM), etc. In an
attempt to fill this research gap, we provide a baseline model that can be
further developed and extended for other domains with contextual search
problem.