COMPARISON OF SOFTWARE METRICS FOR QUALITY ESTIMATION OF THE SOFTWARE PRODUCTS

Abstract

There is a well known principle, that you can manage only what you can measure – therefore adequate measurement and estimation of different elements in any engineering area is a very important development factor, and software engineering is not an exception. Software metrics represent numeric values of software characteristics and are used for analysis of different aspects of the product in order to predict possible problems. Software metrics can be applied to any stage of the software life cycle. For instance, on a stage of requirement analysis and the problem formulation, these metrics help to estimate the required amount of work and, respectively, the amount of financial and work resources. In addition, the metrics can be used directly in estimation of the program code complexity to assess it’s performance, effectiveness and the ability to modify it if needed. Software metrics are an important part of the whole software quality management process as they can be easily automated and applied in order to achieve some valuable technical characteristics. Moreover, they can be combined with another quality management tools and are a beneficial addition to any software product development model. In general, measuring the attributes of software during any phase of development gives a number of beneficial results to the developers as it helps to use the time, finances and effort effectively. This is achieved because of early problems identification, estimation and prevention and because of that the size, cost, quality and simplicity of the maintenance of the product is always under control. This article describes the results of research on different software metrics types and the actual representatives of these types. It assesses the usage area and the convenience of it’s usage on evaluating the quality of both a finished product and a product in the development stage.