A New Software Quality Model for Evaluating COTS Components

Abstract

Studies show that COTS-based (Commercial off the shelf) systems that are being built recently are exceeding 40% of the total developed software systems. Therefore, a model that ensures quality characteristics of such systems becomes a necessity. Among the most critical processes in COTS-based systems are the evaluation and selection of the COTS components. There are several existing quality models used to evaluate software systems in general; however, none of them is dedicated to COTS-based systems. In this contribution, an analysis study has been carried out on several existing software quality models, namely: McCall's, Boehm, ISO 9126, FURPS, Dromey, ISO/IEC TR 15504-2 1998(E), Triangle and Quality Cube, for the purpose of evaluating them and defining a ground to build a new model specializing in evaluating and selecting COTS components. The study also outlines limitations found in the existing models, such as the tendency to ignore a certain quality feature like Functionality or the failure to describe how the quality measurement in these models has been carried out. As a result of this analysis, a new model has been built that supports a standard set of quality characteristics suitable for evaluating COTS components, along with newly defined sets of sub-characteristics associated with them. The new model avoids some of the limitations found in the existing models. The new model ignores quality characteristics that are not applicable to COTS components and is empowered with new ones that are. In addition, it matches the appropriate type of stakeholders with corresponding quality characteristics; such a feature is missing in all existing models. The objective of the new model is to guide organizations that are in the process of building COTS-based systems to evaluate and choose the appropriate products, and that is essential to the success of the entire system.