Ontology based Bayesian software process improvenent

This paper presents an ontology based approach that can support small and medium-sized software enterprises (SMEs) to achieve their software process improvement goals. The approach consists of four steps: i) assessment of the software process and identification of areas under improvement, ii) development of a process knowledge base, iii) conceptualization and analysis of an ontology that represents the process domain, iv) Bayesian analysis on the ontology, experimentation and suggestions for process improvement. The main concept of the approach is presented through a generic software process ontology model. To validate the approach parts of this model was instantiated using company specific process data from a telecommunication SME. The resulted process models are further analysed through applying Bayesian analysis.