Design and development of risk assessment model for software projects using hybrid fuzzy mechanism

Abstract

In many organizations around the world, the failure of software development projects is considered to be a common phenomenon. A mechanism minimizing the project risk is software project risk management. Removal or elimination of software project risks completely doesn t mean the risk mitigation process; but, it indicates the risk reduction achieved to an acceptable level. The main aim of the research work is to design an efficient risk management framework using integrating the soft computing based approaches for defect prediction and prioritizing the most significant unavoidable software project risks to attain the complete project success in the software development process. To achieve this goal, a risk management framework is designed with two phases. A defect prediction system is developed in the first phase to predict the defected software modules during project development. At first, a supervised machine learning approach is developed to classify the predictable software risks based on naïve bayesian classification. Initially, risk factors are monitored accurately by integrating multi-source information. Some of the critical factors such as time, resources and cost budgets can be largely affected in the presence of risk factors. Risk management includes the threats developed in the project environment such as controlling events, planning, analyzing, and identifying and so on. It is essential to monitor the risk attributes such as spatial distribution, temporal and intensity. Then, using this attribute, the risk in the software project is predicted by applying the supervised learning methods. In the first step, appropriate input is selected. This input selection process is done by browsing the data from the dataset. The single statistical data matrix defines the dataset used in this work. newline