A quality measure for automatic ontology evaluation and improvement

Abstract

Ontologies are a useful tool for storing information, as they allow for automated reasoning. However, creating, maintaining, or evaluating ontologies are time-consuming processes. We started with automating the evaluation process, based on a framework from semiotics. Thereafter, we show a method to improve on existing NLP algorithms for the task of generating ontologies. The main contributions of this work are a multi-criteria, automated ontology evaluator, and a proof of concept for using such an evaluator for an automatic improvement of ontologies.