Methodology for ontology design and construction

Resumen

In this paper we present a methodology for ontology design and construction which incorporates the most outstanding design principles and a thorough evaluation process. An ontology provides logical formulation of complex problems of decision sciences like risk management, decision making under uncertainty, statistics and forecasting, negotiation and financial analysis. The main stages of this methodology are: requirements specification, formal design, construction, and evaluation. At each stage of the methodology a series of tasks are defined together with methods and techniques to build the ontology considering quality characteristics. Description logics is used as the formal language during design, a set of informal competency questions is used to support ontology conceptualization; at the evaluation stage, the set of questions are translated to a formal reasoning language and are used for evaluation purposes. Many methodologies and tools have been reported in literature, but little attention has been paid in the creation of consistent, modular, coherent, usable and reusable ontologies as an objective from the beginning of the design process. A comparative analysis with other methodologies is discussed and an Ontological Model for Medical Diagnosis is presented.