A Prediction Model for Software Change using Object-oriented Metrics
Mi‐Jung Lee, Heung Seok Chae, Tae‐Yeon Kim · Jeongbo gwahaghoe nonmunji. so'peuteuweeo mich eung'yong · 2007
Software changes for various kinds of reasons and they increase maintenance cost. Software metrics, as quantitative values about attributes of software, have been adopted for predicting maintenance cost and fault-proneness. This paper proposes relationship between some typical object-oriented metrics and software changes in industrial settings. We used seven metrics which are concerned with size, complexity coupling, inheritance and polymorphism, and collected data about the number of changes during the development of an Information system on .NET platform. Based on them, this paper proposes a model for predicting the number of changes from the object-oriented metrics using multiple regression analysis technique.