The Research And Analysis Of The Eigenvalues Of Trustworthy Measurement In Large-scale Software

Measure is the reflection from the reality or experimental world to the mathematical world, through which people can more easily understand the features of entities and the relationship between them. With the increasing of software scale, software becomes more and more complex. All classes or modules have to inter-cooperate with each other to complete functions expected by users and create the new features as a whole which hide in the structure of software. But the traditional software measurement methods are unable to measure large-scale software effectively. Therefore, trustworthy measurement gives an accurate measurement to these emerging features, providing valuable perspectives and different research dimensions to understand software systems.More and more large-scale software systems appear the characteristics of "small world" and "scale-free" of complex networks. The paper introduces the complex network theory to software measurement methods and proposes a statistical measurement methodology. First, the paper studies the basic parameters of the complex network, and introduces three new measurement parameters:structural holes, closeness; mixing coefficient. Based on software network measurement parameters, the paper design and implement a large software network measurement parameters computing platform. The platform can resolve the software source code and calculate the value of the structure network measurement parameters. To study the characteristics of software, the paper counts the laws of the measurement parameters about200different scale object-oriented open software. In addition, the paper summarizes the evolution trend of these properties with the evolution of software.The paper proposes that the measurement system is consistent with actual system. Measurement is correct and effective and can be used as a supplementary measure method. In addition, trustworthy measurement parameters describe the structure and behavior characteristics of software and provide a new method to measure and design software systems.