ReDA (http://reda.naist.jp/) is a web-based visualization tool for analyzing Modern Code Review (MCR) datasets for large Open Source Software (OSS) projects. MCR is a commonly practiced and lightweight inspection of source code using a support tool such as Gerrit system. Recently, mining code review history of such systems has received attention as a potentially effective method of ensuring software quality. However, due to increasing size and complexity of softwares being developed, these datasets are becoming unmanageable. ReDA aims to assist researchers of mining code review data by enabling better understand of dataset context and identifying abnormalities. Through real-time data interaction, users can quickly gain insight into the data and hone in on interesting areas to investigate. A video highlighting the main features can be found at: http://youtu.be/ fEoTRRas0U.

Open source software (OSS) or Free/Libre OSS (FLOSS) has become an interesting source of research in software engineering. However, it has been criticized that FLOSS development is often considered as a homogeneous phenomenon grounded by assumptions rather than empirical evidence. Proper empirical methods that can shed light into FLOSS development are desirable. In this paper, we propose an empirical method to validate a software development model for FLOSS, the Adapted Staged Model for FLOSS. We mined some selected metrics from Apache Ivy and study their evolution using S-shaped curves. Our results indicate that S-shaped curves can model software evolution well for Ivy. Moreover, we demonstrated that our method can be used to identify successfully different stages of its development, validating part of the Adapted Staged Model for FLOSS.

We present four datasets that are focused on the general roles of OSS peer review members. With data mined from both an integrated peer review system and code source repositories, our rich datasets comprise of peer review data that was automatically recorded. Using the Android project as a case study, we describe our extraction methodology, the datasets and their application used for three separate studies. Our datasets are available online at http://sdlab.naist.jp/reviewmining/.

Due to the distributed collaborations and the volunteering nature of Open Source Software (OSS), OSS peer review processes differs from traditional approaches. Despite the latest research efforts to understand OSS peer review processes, very little is known. Unlike related work, this study investigates OSS peer review processes from a different perspective. We investigate the importance of OSS peer review contributor roles and their review activities by using social network analysis (SNA), proposed as PeRSoN (Peer Review Social Network). As a case study, we extracted and analyzed the review process of Android Open Source Project (AOSP). To the best of our knowledge, this is the first research constructing social networks from mining a peer review repository. Our preliminary results provided hints on relationships among the OSS peer review contributor roles, their activities, and the network structure. The results raised issues that will be used to refine our approach in the future.

This paper describes the use of a UML metric, an approximation of the CK-RFC metric, for predicting faulty classes before their implementation. We built a code-based prediction model of faulty classes using Logistic Regression. Then, we tested it in different projects, using on the one hand their UML metrics, and on the other hand their code metrics. To decrease the difference of values between UML and code measures, we normalized them using Linear Scaling to Unit Variance. Our results indicate that the proposed UML RFC metric can predict faulty code as well as its corresponding code metric does. Moreover, the normalization procedure used was of great utility, not just for enabling our UML metric to predict faulty code, using a code-based prediction model, but also for improving the prediction results across different packages and projects, using the same model.

In this paper, we discuss the challenge of making logistic regression models able to predict fault-prone object-oriented classes across software projects. Several studies have obtained successful results in using design-complexity metrics for such a purpose. However, our data exploration indicates that the distribution of these metrics varies from project to project, making the task of predicting across projects difficult to achieve. As a first attempt to solve this problem, we employed simple log transformations for making design-complexity measures more comparable among projects. We found these transformations useful in projects which data is not as spread as the data used for building the prediction model.