PrePrint: Using software analytics to understand how companies interact in free software communities
Free, open source software development communities may become large and complex. They may also be a focus of interest for competing companies relying on their outcomes, with their employees joining the development and maintenance effort. In those cases it is specially important for both companies and communities to understand how this collaboration is working, and how it matches their policies and expectations. In this paper we show two cases (OpenStack and WebKit) which we have studied using analytics techniques on the data obtained from their software development repositories. From them, we conclude that analytics on this data can improve the factual knowledge about how development communities are performing in aspects which are of interest to companies, and fundamental to ensure transparency and fairness.
PrePrint: Software Analytics in Practice
Software analytics is to utilize a data-driven approach to enable software practitioners to perform data exploration and analysis in order to obtain insightful and actionable information for completing various tasks around software systems, software users, and software development process. Being part of the Software Analytics group of Microsoft Research Asia, we have conducted a number of research projects on software analytic with successful technology transfer and with high impact on industrial practices. In this article, we present a brief overview of software analytics and the critical success factors of conducting software analytics in practice. We then use the StackMine project, which was used by and transferred to Microsoft product teams, as an example to illustrate these success factors and to share our experiences and lessons learned with respect to these success factors.
PrePrint: A Retrospective Study of Software Analytics Projects: In-Depth Interviews with Practitioners
Software analytics has become important to guide practitioners in decision making throughout the software development process. In this context, prediction models are used to help managers to efficiently organize their resources and identify problems by analyzing patterns on the existing project data in an intelligent and meaningful manner. In the past decade, we have worked with software organizations to build metrics repositories, and predictive models to address process, product and people related issues in practice. Some of our models were deployed, some were prototypes, and some are work-in-process. In this paper, we would like to share our experience over the years reflecting the expectations and outcomes both from a practitioner and researcher points of view. We aim that our conclusions from in-depth interviews with practitioners in three case studies and our future roadmap as researchers would shed light to some issues for industry and academia.
PrePrint: Are software analytics efforts worthwhile for small and medium companies? The case of Amisoft
Amisoft, a Chilean software company with 43 employees, succesfully uses software analytics in its projects; they support a variety of strategic and tactical decisions, notably resulting in a reduction in overwork of employees. However, the analytics done at Amisoft are very different from the ones used in larger companies.
PrePrint: A Comment on C. Symons' "Exploring Software Project Effort versus Duration Trade-offs"
A recently proposed process for exploring the trade-off between project duration and effort does not correctly estimate of the strength of the relationship. This is argued by pointing out a problem with the derivation, as well as by demonstrating that the proposed process yields an incorrect result in a Monte Carlo simulation.