Error Leakage and Wasted Time: Sensitivity Analysis of a Requirements Consistency Checking Process. Hayes, J. H., Gu�h�neuc, Y., Antoniol, G., Li, W., & Truszczynski, M. In Proceedings of the 1<sup>st</sup> North American Search Based Software Engineering Symposium (NasBASE), pages 1061–1080, February, 2015. Elsevier. 15 pages.
Error Leakage and Wasted Time: Sensitivity Analysis of a Requirements Consistency Checking Process [pdf]Paper  abstract   bibtex   
A myriad of techniques are used by requirements engineering researchers and practitioners to address difficult problems, such as consistency checking of temporal requirements. Often, complex problems are addressed by building processes/tools that combine multiple techniques where the output from one technique becomes the input to the next technique, e.g., feature location that uses information retrieval and dynamic analysis techniques in sequence to perform the three step process of preparing a corpus, generating queries, and retrieving results. While powerful, these techniques are not without flaw. Inherent errors in each technique may leak into the subsequent step of the process. Errors then can be viewed as variations in the overall process. Errors of omission, or failure to retrieve elements, are viewed as error leakage because the "lost" elements will not be processed in subsequent steps. Errors of commission, or retrieval of irrelevant elements, amount to wasted time as human analysts will review/analyze these extraneous elements. As software quality professionals, developers, and researchers depend on these processes to verify and validate software and attendant artifacts, it is important to understand the impact of these errors on the quality of the output of the final step of the processes, e.g., the accuracy of the list of features retrieved using feature location. Therefore, we model and study one such process, for checking the consistency of temporal requirements. We study the process and assess error leakage and wasted time considering this process as fully automated. We perform an exploratory sensitivity analysis using Monte Carlo simulations of the input factors of our model to determine the effect that these sources of uncertainty, i.e., the errors of omission and commission, may have on the final accuracy of the consistency checking process. The sensitivity analysis uses published data on accuracy of previous techniques and data collected by applying the process on a real-world system. We share insights gained and discuss its applicability to other processes built of piped techniques.
Downloads: 0
About
Documentation
Keywords
Search
Users
Terms and Conditions of Use
Privacy Policy
Sitemap
© BibBase.org 2021