The popularity of open source has witnessed significant growth in the past two decades. The sustainability and success of open source projects rely heavily on the active community of contributors, who have the freedom to join or depart voluntarily. The churn of core developers can have detrimental effects on the continuous development and progress of these communities. Therefore, previous studies have aimed to identify the factors that influence developer churn and reveal its impact on open source communities. However, there is currently a lack of research on predicting whether a developer will leave or retain within a given timeframe. This article aims to bridge this gap by introducing a novel model called context-aware dynamic graph neural network (ConDyGNN), which is designed specifically for predicting developer churn in open source communities. By leveraging historical data, this model takes into consideration various factors such as developers’ activities, collaborative networks, and community context information. To capture both temporal dependencies and spatial correlations among different developers, temporal convolutional network (TCN) and graph convolutional network (GCN) modules are employed inConDyGNN. Furthermore, a context-aware module based on a multihead attention mechanism is proposed to effectively integrate macrolevel community information and enhance the predictive performance. The superiority of the proposed model is empirically demonstrated through extensive comparative experiments and an ablation study.

The role of focal packages in packaging ecosystems is crucial for the development of the entire ecosystem, as they are the packages on which other packages depend. However, the evolution of dependency groups in packaging ecosystems has not been systematically investigated. In this study, we examine the downstream dependency package groups (DDGs) in three typical packaging ecosystems—Cargo for Rust, Comprehensive Perl Archive Network for Perl, and RubyGems for Ruby—to identify their features and evolution. We also identify and analyze a special type of DDG, the collaborative downstream dependency package group (CDDG), which requires shared contributors. Our findings show that the overall development of DDGs, particularly CDDGs, is consistent with the status of the whole ecosystem, and the size of DDGs and CDDGs follows a power law distribution. Furthermore, the interaction mechanisms between focal packages and downstream packages differ between ecosystems, but focal packages always play a leading role in the development of DDGs and CDDGs. Finally, we investigate predictive models for the development of CDDGs in the next stage based on their features, and our results show that random forest and Gradient Boosting Regression Tree achieve acceptable prediction accuracy. We provide the raw data and scripts used for our analysis at https://github.com/onion616/DDG .

The active participation of contributors is key to the success of any open source software (OSS) projects.In the open source community, it can be found that some developers have participated together in multiple projects and also built social connections so that they form an independent team.Independent teams can be high-quality developer candidates for OSS projects.By modeling the developers and independent teams, we design approaches to recommend independent teams to OSS projects.We used the interaction data between developers to build the developers' social connection network, on which independent teams can be discovered by utilizing the community discovery algorithm.Based on the graph attention network, we designed a team recommendation model GAT-team.The performance of the GAT-team is evaluated on a real-world dataset.GAT-team achieves much better results compared with the other algorithms in the experiments.We provide the data and the code at https://

Open source software (OSS) development is a highly collaborative process where individuals, groups and organizations interact to develop, operate and maintain software and related artifacts. The developers' sentiment in this process can have an impact on their working willingness and efficiency. Monitoring sentiment factors can help to improve OSS development and management. However, no method has been proposed to dynamically monitor the sentiment phenomena during the OSS development process. In this paper, an approach to detect Negative Sentiment-related Events (NSE) is proposed. It consists of two steps. The first step is to identify the burst interval of negative comments from open source projects, which corresponds to a NSE. The second step is to annotate this NSE with its event type. To support this approach, the types of NSEs in OSS projects are defined through an empirical study and classifiers are trained to annotate event types automatically. Moreover, conversation disentanglement techniques are employed to make the comments extracted more complete. Finally, the factors that have an influence on NSEs in the OSS project are studied.

Open source software development (OSSD) projects is a collaborative process among developers and volunteers with common interests. OSSD is increasingly becoming a trend and many successful OSSD projects have contributed software packages which have been widely adopted. Unfortunately, there is Pareto principle in OSSD projects, and a large number of OSSD projects have little influence, many projects a re unable to attract user interest. Therefore, it is important to understand the factors affecting OSSD project success as measured by project popularity. While technical factors such as license and compatibility with certain operating systems have a direct influence on the popularity of OSSD projects, development-related factors also have latent impacts on the popularity of OSSD projects. Therefore, we collect data on 445 projects and successfully build a structural equation model (SEM) which depicts the inherent relationships between development-related factors and OSSD project popularity. The steps to build the SEM and the implications of this model are discussed in detail in this paper.

In the mashup ecosystem, Web APIs often form cooperative communities which evolve with time. Understanding how these communities form and evolve with time is very important to help develop strategies for improving ecosystems of Web APIs. This paper empirically studies the evolution of Web API cooperative communities based on the data of Web APIs and mashups from Programmable Web. It is found that 20% of Web APIs account for half of the connections in their communities. The Web API communities identified in terms of mashups seem to be stable. Every two years, there is a 27% growth in the number of new communities, 10% dissolve, 30% tend to expand, while 5% become smaller. It is a key stage when a Web API community becomes medium sized because in this stage, there is a high probability that a Web API community reduces its size or even dissolves.

In recent years, the organization of open source software development has evolved rapidly. Analyzing the behaviors of contributors helps us understand the development process of open source software and explore the general and special rules of it. Agent based model is a type of computing model that can simulate the behaviors and interactions between autonomous entities. With agent based models, we can simulate the self-organization process of open source software development by simulating contributors' behaviors. Therefore, we design an agent-based simulation system for open source software development, which are implemented with the Java Agent Development framework. In experiments, we obtain the simulation results by inputting historical behavior information of open source repositories on GitHub. Then we compare the results of various models. Finally, by adjusting input parameters for issue resolution process, we analyze the impact of these parameters, which shows this system also helps understand how to control the open source software development process.

Teamwork is very important to software development. There are many studies focusing on different aspects of teamwork in open source software projects, but neglecting the fact that most teams of open source projects are temporary and dependent on the context of one specific project. Whether the collaboration of such teams can extend to different projects is highly doubted. In contrast, we are interested in long-lasting socially connected teams, whose members have steady social connections and have collaborated with each other on multiple projects. Therefore, we mine Cross-Repository Socially Connected (CRSC) teams on GitHub, the largest open-source project hosting platform. Community detection methods are used to mine CRSC teams from the developer network and more than 20,000 CRSC teams are discovered on GitHub. The productivity of such teams and how the hosting repository may influence them are studied. Their preferences for repositories are investigated. Moreover, we study the structures of these teams using complex network analysis methods. Our results indicate that CRSC teams are stable, highly productive and mature. Therefore, open-source project owners and recruiters can pay more attention to such teams.

In order to facilitate collaboration and improve service quality, progress information together with predicted information (Predictive Monitoring Information, PMI) should be published to partners or customers in a timely manner. Currently, the publishing points of PMI are assigned to certain positions empirically before the execution of a service process. However, because of the dynamic nature of service processes, a decision to publish PMI should be made based on the actual progress and predictions of the future. Therefore, an adaptive approach to select the publishing points of PMI for service processes is proposed. We design a method to quantify the benefits of notifying the delay risk at a point in the service process using long short-term memory (LSTM) neural networks. Furthermore, the deep reinforcement learning model is applied to select the publishing points for PMI dynamically. Experiments on a real-world dataset prove that our approach can select the optimal publishing points.

With the extensive applications of Web APIs, to monitor and analyze personal Web API invocation behaviors is becoming more and more important. However, due to the fact that the users' Web API invocation behaviors are highly heterogeneous, it is extremely challenging to provide a unified framework. In the paper, we introduce a predictive Web API Invocation Behavior Monitoring System (AIMS). AIMS automatically analyzes the predictability of users' invocation behaviors and for the user whose behaviors can be predicted, an adaptive strategy is applied to generate the predictions in an adaptive way. In addition, a context-aware K-nearest neighbor classifier is applied to detect the anomaly of user's invocation behaviors. Experiments on both a real-world dataset and synthetic dataset show AIMS is efficient in analyzing personal API invocation behaviors.

Recently, text images deblurring has achieved advanced development. Unlike previous methods based on hand‐crafted priors or assume specific kernel, the authors recognise the text deblurring problem as a semantic generation task, which can be achieved by a generative adversarial network. The structure is an essential property of text images; thus, they propose a structural loss function and a detailed loss function to regularise the recovery of text images. Furthermore, they learn from the coarse‐to‐fine strategy and present a multi‐scale generator, which is utilised for sharpening the generated text images. The model has a robust capability of generating realistic latent images with photo‐quality effect. Extensive experiments on the synthetic and real‐world blurry images have shown that the proposed network is comparable to the state‐of‐the‐art methods.