Using Bayesian Networks to estimate Strategic Indicators in the context of Rapid Software Development

Abstract

Background: During Rapid Software Development, a large amount of project and development data can be collected from different and heterogeneous data sources. Aims: Design a methodology to process these data and turn it into relevant strategic indicators to help companies make meaningful decisions. Method: We adapt an existing methodology to create and estimate strategic indicators using Bayesian Networks in the context of Rapid Software Development, and applied it to a use case. Results: Applying the methodology in the use case, we create a model to predict product quality based on software factors and metrics, using companies' business knowledge and collected data. Conclusions: We proved the methodology's feasibility and obtained positive feedback from the company's use case. CCS CONCEPTS• Software and its engineering → Software creation and management KEYWORDS Strategic indicator, Bayesian Network, Decision Making, Rapid software development

A Method to Estimate Software Strategic Indicators in Software Development: An Industrial Application

“…As mentioned in section 2, the SESSI method was conceived from dealing with the industrial needs of the Q-Rapids industrial partners and devising solution attempts in their contexts. Preliminary formative results of the specification and estimation model construction phases are provided in [16] and [17] respectively. The method presented here extends these preliminary results by consolidating both phases and providing not just results in the context of a single partner but consolidating the experiences from all of them, broadening the applicability of the method.…”

“…We applied our solution attempts in these partners, and we shaped the SESSI method based on the lessons learnt and insights gained from them. Preliminary formative results of the application of our solution attempts have been published elsewhere [16,17]. − Stage 3.…”

Quality-Aware Rapid Software Development Project: The Q-Rapids Project

“…SI assessment. The dashboard provides two strategies to compute SIs: a quantitative approach based on computing the average of the quality factors, and a qualitative approach involving experts and historical data to define a Bayesian Network model [15]. Fig 5 shows the BN model for the Product Quality strategic indicator, impacted by Code Quality, Stability, and Testing Status quality factors.…”

Q-Rapids: Quality-Aware Rapid Software Development – An H2020 Project

“…The key strategic indicators can be assessed qualitatively defining a Bayesian Network based on the factors impacting on them [4]. • Predicting future quality levels through multiple predictive methods [5].…”