Construct quality index system of foundational software platform using improved PCA method

As a result of the trend of platform, network, service-oriented appeared during the process of software development, the Foundational Software Platform (FSP), as a kind of software modality, has received more and more attention. However, as it has been widely used, the measurement on FSP cannot be implemented effectively, which further influence the quality of its upper applications. This, to a great extent, is because of the uniqueness of basic software platform. There is no a scientific method which can be used to determine its quality index system. This paper proposes an improved method of Principle Components Analysis to create appropriate quality index system for FSP. This method can integrate a series of indicators into a few comprehensive ones like the Principle Component Analysis (PCA). Meantime, it could handle the situation of vague or large amount of indicators, which is the flaw of the PCA. It guarantees the accuracy and authenticity of the results. Moreover, we verify the quality index system through practice. It has been proven that the quality model is helpful and the Improved PCA approach is useful when constructing index system.