Agile Ontology Development: A Comprehensive Framework from Preliminary Investigations to Evaluation

Abstract

This research proposes a comprehensive framework for ontology development within the database domain, integrating agile methodologies from initial investigations to evaluation. The framework comprises three main phases: Preliminary Investigations, Ontology Design and Development, and Ontology Evaluation, each encompassing specific sub-stages. In the Preliminary Investigations phase, entity recognition is conducted through interviews, document analysis, and keyword extraction using tools like Monkey Learn. Relationships between entities are identified to establish hierarchies and associations, facilitating semantic representation. The Ontology Design and Development phase involves domain modeling and ontology implementation, including the transformation of entity relations into OWL classes and properties. Guidelines for mapping entity relations to OWL are provided, ensuring a seamless transition from Entity Relation Diagrams (ERD) to ontology representation. Finally, the Ontology Evaluation phase employs two methods: Domain Experts Evaluation and Pellet Reasoner. Domain experts assess the ontology's credibility, consistency, completeness, and conciseness, while the Pellet Reasoner ensures logical consistency and provides essential inference services. The proposed framework offers a systematic and agile approach to ontology development, particularly beneficial in dynamic domains like database management.