Measuring the inconsistencies between process model and process execution

At present, using process models to simulate, validate and guide software development is an important approach to improve software development processes, and enhance the quality of software products, but in reality the actual executing process always deviates from the expected process model. How to detect and measure such inconsistencies is a challenging task. The highly dynamic and exceptional nature of software processes means that simple yes-no answers carry too little information about the significance of any given inconsistency. Managers need to understand where an inconsistency occurs and how severe that inconsistency might be before taking any corrective action. This paper uses event trees to represent the process model and process execution, puts forwards MinCost and Similarity metrics to measure the severity of inconsistencies, and designs an algorithm to assist in compare event trees and compute metrics.Engineering, Electrical & ElectronicSCI(E)EI0ARTICLE2223-2261