Risk management is one of the most important processes in Project management that directly affects the success of the project. By determining the common characteristics of occurred risks in past projects, and by creating a smart risk model, a project's success can be estimated at the early stages of the project. Due to this objective, in recent years, project success forecasting and risk evaluation studies have a significant increase. Hence, in this thesis study, primarily, risk management in software projects and risk evaluation methods for the past studies are inspected.
In this thesis work, the data for evaluating smart forecasting model is used from software projects that were developed in a telecommunication company between 2007 and 2014. Since the changes due to the competition in telecommunication sector and the fast evolution of telecommunication technology, the risk management in software projects is becoming vital.
Using clustering and classification algorithms, the relationship between the common characteristics of the given application data in software projects and the risks are defined. Referring to this relationship, a fuzzy logic model is created to forecast just at the beginning how a similar project will end. When predefined project factors are used, the project success is assessed with two different scales such as %76 ratio - %20 deviation. |