Predicting Community Smells’ Occurrence on Individual Developers by Sentiments

Abstract

Community smells appear in sub-optimal software development community structures, causing unforeseen additional project costs, e.g., lower productivity and more technical debt. Previous studies analyzed and predicted community smells in the granularity of community sub-groups using socio-technical factors. However, refactoring such smells requires the effort of developers individually. To eliminate them, supportive measures for every developer should be constructed according to their motifs and working states. Recent work revealed developers' personalities could influence community smells' variation, and their sentiments could impact productivity. Thus, sentiments could be evaluated to predict community smells' occurrence on them. To this aim, this paper builds a developer-oriented and sentiment-aware community smell prediction model considering 3 smells such as Organizational Silo, Lone Wolf, and Bottleneck. Furthermore, it also predicts if a developer quitted the community after being affected by any smell. The proposed model achieves cross-and within-project prediction F-Measure ranging from 76% to 93%. Research also reveals 6 sentimental features having stronger predictive power compared with activeness metrics. Imperative and indicative expressions, politeness, and several emotions are the most powerful predictors. Finally, we test statistically the mean and distribution of sentimental features. Based on our findings, we suggest developers should communicate in a straightforward and polite way.

Community smells—The sources of social debt: A systematic literature review

“…We categorized 24 studies as empirical and one as theoretical according to Appendix A. Empirical studies conducted by Tamburri [5,11,20,42] and Palomba [10] offer information about the identification of the causes and effects of community smells in practice, information about the impact on work settings, and management approaches. The remaining included studies provide further information on community smells in relationship with development communities and processes [14,32,33,34,43,44,45,46] and research contributions such as models [12,13,16,47,48,49,50,51,52,53], organizational strategies [15], and tools [54]. We organized our results around the research questions described in Section 3.…”

“…Huang et al developed and enhanced the performance of a prediction model that detects community smells based on sentiment analysis in three scenarios, i.e., cross-project, within-project, and time-wise validation [51,52]. The study included ten process metrics to capture developers' activities.…”

Bug report priority prediction using social and technical features

“…The trend of some features is still unclear in Table 15, and the difficulty to interpret sentiments in line with other socio-technical research, indicating sentiment-related problems are multifaceted. 15,38,39,91 A notable example is the distribution of R_SEN_VAL. A related study 39 indicated that contributors are in higher productivity dealing with challenging and important tasks.…”

“…42 Tamburri et al 42 implemented a community smell detection tool called CODEFACE4SMELLS which evaluates development mailing list and software repository history information to detect various community smells. Based on their study, we 15 improved the prediction of community smell occurrence on individual developers by involving their sentiments, and we also suggested that developers should communicate in a straightforward and polite way to improve community healthiness. Driven by the suggestion of the empirical study on practitioners, 8 we intend to involve more aspects related to developers and third-party systems, and thus, the prior sociotechnical analysis outcomes can provide evidence on how to model contributors' characteristics in terms of their sentiments and OSS activities.…”

An Approach for Community Smells with Power Relationship Dynamics in Proprietary Software Ecosystems

“…Using the computational tool "CODEFACE4SMELLS", [Huang et al 2021] labeled "state of the art" and used the tool to detect some community smells using email list and information about the history of a software repository. The authors suggested a technique based on developer sentiment analysis to predict the occurrence of some community smells.…”