2011 International Symposium on Empirical Software Engineering and Measurement (2011)
Sept. 22, 2011 to Sept. 23, 2011
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/ESEM.2011.56
General software cost parameters such as size, effort distribution, and productivity are necessarily imprecise due to variations by domain. To improve this situation, empirical software cost analysis using the primary US DoD cost database has been segmented by domain. This analysis supports a software cost estimation metrics manual for improvements in acquisition policies, procedures and tools. We have addressed the challenges of consistent data definitions and taxonomies across diverse stakeholder communities, data integrity, data formats, and others. We highlight example analysis results from an application domain demonstrating cost estimating relationships, benchmarks on reuse parameters and effort distributions for estimators to use.
software cost estimation, software metrics, software cost models, software productivity, Department of Defense
T. Tan, R. Madachy, B. Boehm, B. Clark and W. Rosa, "US DoD Application Domain Empirical Software Cost Analysis," 2011 International Symposium on Empirical Software Engineering and Measurement(ESEM), Alberta, Canada, 2011, pp. 392-395.