Software project profitability analysis using temporal probabilistic reasoning; an empirical study with the CASSE framework

Abstract

Undertaking adequate risk management by understanding project requirements and ensuring that viable estimates are made on software projects require extensive application and sophisticated techniques of analysis and interpretation. Informative techniques and feedback mechanisms that help to assess how well and efficiently a specific development methodology is performing are still scanty. Analyzing project tasks would enhance how well individual tasks are estimated, how well they are defined, and whether items are completed ontime and on-budget. In this paper, we propose a temporal probabilistic model that addresses feedback control mechanisms in project planning using the Complex Adaptive Systems Software Engineering framework (CASSE). We have tested our approach in industry with a software development company in South Africa on two commercial project evaluations. Our preliminary results show that the temporal probabilistic model of the framework demonstrably enhances practitioners’ understanding in managing software projects profitably - hence increasing business sustainability and management