Research Repository

A graph-based approach to measuring semantic relatedness in ontologies

Hawalah, A and Fasli, M (2011) A graph-based approach to measuring semantic relatedness in ontologies. In: UNSPECIFIED, ? - ?.

Full text not available from this repository.

Abstract

The need of determining the degree of semantic similarity, relatedness or distance between two concepts within the same ontology or two different ontologies is becoming an increasingly important task in the field of Information Retrieval. Although a great attention has been paid to design semantic similarity/relatedness methods based on taxonomies, there has been little discussion about the design of semantic similarity/relatedness methods based on ontologies. In this paper we introduce a novel graph-based semantic relatedness approach to calculate semantic relatedness considering both hierarchical and non-hierarchical concepts in an ontology. In addition, our approach considers some important properties such as different relation types, concepts' depth and distance that play an essential role in measuring semantic relatedness. Three experimental studies are provided to first illustrate that our approach give a different results than other measures in the literature. Then, we compare our approach against existed methods using a benchmark dataset, and finally, evaluate our approach by using a real ontology and compare the predictions of our semantic relatedness approach against the human-subject judgment. The results in all the experiments show a considerable improvement against traditional taxonomy-based measures. © 2011 ACM.

Item Type: Conference or Workshop Item (UNSPECIFIED)
Additional Information: Published proceedings: ACM International Conference Proceeding Series
Subjects: Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Divisions: Faculty of Science and Health > Computer Science and Electronic Engineering, School of
Depositing User: Users 161 not found.
Date Deposited: 14 Aug 2012 16:09
Last Modified: 17 Aug 2017 18:08
URI: http://repository.essex.ac.uk/id/eprint/3653

Actions (login required)

View Item View Item