Evaluating an optimized backward chaining ontology reasoning system with innovative custom rules
| Pages | 45-56 |
| Published date | 19 February 2018 |
| Date | 19 February 2018 |
| DOI | https://doi.org/10.1108/IDD-10-2017-0070 |
| Author | Hui Shi,Dazhi Chong,Gongjun Yan |
| Subject Matter | Library & information science,Library & information services,Lending,Document delivery,Collection building & management,Stock revision,Consortia |
Evaluating an optimized backward chaining
ontology reasoning system with
innovative custom rules
Hui Shi
California State Polytechnic University, Pomona, California, USA
Dazhi Chong
MSIT, California Lutheran University, Thousand Oaks, USA, and
Gongjun Yan
Management and Information Sciences, University of Southern Indiana, Evansville, Indiana, USA
Abstract
Purpose –Semantic Web is an extension of the World Wide Web by tagging content with “meaning”. In general, question answering systems
based on semantic Web face a number of difficult issues. This paper aims to design an experimental envi ronment with custom rules and scalable
data sets and evaluate the performance of a proposed optimized backward chaining ontology reasoning system. This study also compa res the
experimental results with other ontology reasoning systems to show the performance and scalability of this ontology reasoning syste m.
Design/methodology/approach –The authors proposed a semantic question answering system. This system has been built using ontological
knowledge base including optimized backward chaining ontology reasoning system and custom rules. With custom rules, the proposed semantic
question answering system will be able to answer questions that contain qualitative descriptors such as “groundbreaking”resesarch and “tenurable
at university x”. Scalability has been one of the difficult issues faced by an optimized backward chaining ontology reason ing system and semantic
question answering system. To evaluate the proposed ontology reasoning system, first, the authors design a number of innovative custom rule sets
and corresponding query sets. The innovative custom rule sets and query sets will contribute to the future research on evaluating ontology reasoning
systems as well. Then they design an experimental environment including ontologies and scalable data sets and metrics. Furthermore , they evaluate
the performance of the proposed optimized backward chaining reasoning system on supporting custom rules. The evaluation results have been
compared with other ontology reasoning systems as well.
Findings –The proposed innovative custom rules and query sets can be effectively employed for evaluating ontology reasoning syste ms. The
evaluation results show that the scalability of the proposed backward chaining ontology reasoning system is better than in-memory reasoning
systems. The proposed semantic question answering system can be integrated in sematic Web applications to solve scalability issues. For light
weight applications, such as mobile applications, in-memory reasoning systems will be a better choice.
Originality/value –This paper fulfils an identified need for a study on evaluating an ontology reasoning system on supp orting custom rules with
and without external storage.
Keywords Semantic web, Benchmark, Ontology, Backward chaining reasoner, Innovative custom rules, Ontology reasoning system
Paper type Research paper
1. Introduction
There is a variety of structured and semi-structured information
increasingly available on the Internet, which can be mined,
organized and queried in a collaborative environment. As a
result, there have been more and more research on the
intersections of semantic web, collaborative work and social
media research (Breslin et al.,2005,2009;Schaffert, 2006;Sure
et al.,2002;Wennerberg, 2005;Carminati et al.,2011;He and
Tian, 2017). For example, IkeWiki, as a semantic wiki, is mainly
developed for collaborative knowledge engineering with support
for different formalization (Schaffert, 2006). Another example is
SIOC (Semantically Interlinked Online Communities), which
interconnect present online communities for data integration and
cross-site structural queries (Breslin et al.,2005).
Despite the growing number of work in this area, there is no
much research related to semantic Web systems that can
provide answers to qualitativequeries submitted by users. In an
effort to contribute to this research area, we have recently
developed a semantic Web system where the underlying
knowledge base covers linked data about scientific research.
The objective of the system is to provide answersto qualitative
queries that representthe evolving consensus of the community
of researchers. The system is expected to answer qualitative
queries such as: “Who are the groundbreaking researchers in
The current issue and full text archive of this journal is available on
Emerald Insight at: www.emeraldinsight.com/2398-6247.htm
Information Discovery and Delivery
46/1 (2018) 45–56
© Emerald Publishing Limited [ISSN 2398-6247]
[DOI 10.1108/IDD-10-2017-0070]
Received 10 October 2017
Revised 11 November 2017
Accepted 22 November 2017
45
To continue reading
Request your trial