Abstract:

We apply modified K-means and decision trees classification methods to classify projects in two ways: according to industry, organization size, project duration and project management type, in order to find out which risks are most common for each class, and additionally according to risk types identified in the risk identification process, in order to determine which risk types indicate that the customer expectations with respect to project outcome may turn out to be substantially different at the end of the project than they were at its beginning. Results of a questionnaire with 350 respondents are used for calculations. The conclusions may improve the project management process: they will help to prioritize risk classes and to decide when to put more weight on project scope and requirements management.