Framework for Maintainability Measurement of Web Application for Efficient Knowledge-Sharing on Campus Intranet

Abstract

Web Application is placed and accessed over a network, which could be an intranet, extranet or the internet. An intranet is identified as an ideal platform for knowledge- sharing and collaboration in organizations, or institutions. But it is at times hampered by maintainability issue which indeed is a key quality attribute of web applications. This paper presents an explicit description of a process for prediction of maintainability of web application based on design metrics and statistical approach. The work investigates whether a set of measures identified for UML class diagram structural properties (size, complexity, coupling, cohesion) could be good predictors of class diagram maintainability based on the sub-characteristics; understandability, analyzability, and modifiability. Results indicate that useful prediction models can be built from the measures and identified the strongest predictors from the proposed metrics. This framework can be applied to construct maintainability prediction models to control the maintenance tasks of the system and promote efficient collaboration in university campus.