Innovation and Knowledge Management: A Global Competitive Advantage 1112 Towards a Comprehensive Ontology-Driven Software Development Approach Nehemiah Mavetera, North West University, Mmabatho, South Africa, Nehemiah.Mavetera@nwu.ac.za Abstract The field of software development has struggled for a long time to get a development approach that captures semantics, pragmatics and organizational context in addition to the unavoidable use of syntax in the software product. This has led to several research studies, detailing some varied software development frameworks and approaches, both in structure and content that use ontologies. However, there seems to be a persistent gap in what the software development field requires to address the missing human aspects of organizational systems in the software product. This problem is not necessarily by design, but a big short sight in the eyes of the method engineers. In short, these method engineers do not address the social construction nature of the process of developing software development approaches. This has left them focusing on the design aspects and not the constructivist nature of the process of formulating software development approaches and methodologies. Using Grounded theory method, this paper will take the constructivist nature of both ontologies and the method engineering process and present a software development approach that can improve the process of capturing softer, human elements of organizational systems in software products. At the end, the methodological design implications will be discussed. Keywords: Ontology, Software development approach, Framework, constructivist research Introduction Software development or software engineering can be regarded as an example of a model-driven approach to developing artifacts (Henderson-Sellers, 2011).In the same vein, ontologies are considered to be software models (Alonso, 2006; Henderson-Sellers, 2011). There has also been a concerted effort from the software development researchers to incorporate ontologies at development and runtime stages of the software products. This has resulted in many ontology-based information systems artifacts being developed and many have termed this process an ontology-driven approach to software development. Also, many researchers such as Holten et al (n.d.), Liska and Navrat (2010), Alonso (2006), Mavetera (2007), Mavetera and Kroeze (2010b) and Shama and Ingle (2001) to mention but a few have also looked into the feasibility of this approach. The only difference is that, others have concentrated on the design of ontologies as IS artifacts, while very few have been concerned with the role ontologies play in the ISDM method engineering process, a constructivist approach to research in this field. As models, ontologies will be regarded as abstractions of reality that look at the mapping of the world view to some representation such as a picture, drawing, or text using some specific language. This is also supported by Henderson-Sellers (2011) who require such a model to conform to a certain conceptualization. The field of software development has relied heavily on the formulation of these models to capture structure and represent reality with the final product, the software product itself being a good example of such a model. On the other hand, ontologies have been touted as the next big thing in the development of software products as reflected from the works of Gruber (1993), Wand and Weber (1993). However, while researchers have been preoccupied by the form of representation of organizational systems as models or ontologies, they have paid very little attention to the way ontologies can be introduced in the software development process. The biggest problem that has bedeviled the development of an ontology-driven software development approach is the emphasis of researchers on the development of ontologies and ontology development platforms, issues that fall within the science of design (Hevner et al., 2004; Purao et al., 2008). As design artifacts ontologies portray concrete and observable aspects of reality (Isabella, 1990). Quite often, these do not address holistically the constructivist nature of the process of developing a methodology. In this paper, we encourage