Data mining techniques and machine learning methods are commonly used in several disciplines. It is possible that they could also provide a basis for quality assessment of software development processes and the final software product. Number of researches who employ such techniques and methods on software cost and effort estimation are increasing. This article provides a software quality perspective on data mining and machine learning applications, and explains the current research challenges that are related to incorporating these tools based on several metrics derived from the software development processes and the software itself.
Burak Turhan, F. Onur Kutlubay