Abstract | Many web applications have evolved from simple HTML pages to complex service-oriented applications that have high maintenance cost. This high maintenance cost is due to the heterogeneity of web applications, to the fast Internet evolution and to the fast-moving market which imposes short development cycles and frequent modifications. In order to control the maintenance cost, quantitative metrics for predicting web applications' maintainability must be used. This study introduces class diagram design metrics based on Conallen Web Application Extension for UML. The study will use data from an industrial web application to show the correlation between the class diagram metrics and maintenance effort measured by the number of lines of code changed. |
---|