A Rule-Based Metric for Calculating Semantic Relatedness Score for the Motion Picture Industry
Sharmin Choudhury, Kerry Raymond, Peter L. Higgs · 2008
Semantic relatedness is an important part of developing ontology based systems, as it provides the ability to measure the relatedness of two concepts. This is important not only for computational linguistics but also for intelligent querying and data classification. In this paper we present a rule-based calculation for a semantic relatedness score. This metric was specifically developed for Loculus, an ontology we have developed for the motion picture industry. The Metric is evaluated using three axes: the temporal axis, the inheritance axis and the linkage axis. The metric has been tested on pairs of concept; the resulting scores are consistent with human expectations.