Mining Valence, Arousal, and Dominance: Possibilities for Detecting Burnout and Productivity?
Individual Sustainability
Authors: Mika Mäntylä, Bram Adams, Giuseppe Destefanis, Daniel Graziotin, Marco Ortu
Year: 2016
Published in: MSR '16: Proceedings of the 13th International Conference on Mining Software Repositories.
Read me: DOI: https://doi.org/10.1145/2901739.2901752.
Abstract: Similar to other industries, the software engineering domain is plagued by psychological diseases such as burnout, which lead developers to lose interest, exhibit lower activity and/or feel powerless. Prevention is essential for such diseases, which in turn requires early identification of symptoms. The emotional dimensions of Valence, Arousal and Dominance (VAD) are able to derive a person’s interest (attraction), level of activation and perceived level of control for a particular situation from textual communication, such as emails. As an initial step towards identifying symptoms of productivity loss in software engineering, this paper explores the VAD metrics and their properties on 700,000 Jira issue reports containing over 2,000,000 comments, since issue reports keep track of a developer’s progress on addressing bugs or new features. Using a general-purpose lexicon of 14,000 English words with known VAD scores, our results show that issue reports of different type (e.g., Feature Request vs. Bug) have a fair variation of Valence, while increase in issue priority (e.g., from Minor to Critical) typically increases Arousal. Furthermore, we show that as an issue’s resolution time increases, so does the arousal of the individual the issue is assigned to. Finally, the resolution of an issue increases valence, especially for the issue Reporter and for quickly addressed issues. The existence of such relations between VAD and issue report activities shows promise that text mining in the future could offer an alternative way for work health assessment surveys.
Bibtex (copy):@inproceedings{10.1145/2901739.2901752,
author = {M\"{a}ntyl\"{a}, Mika and Adams, Bram and Destefanis, Giuseppe and Graziotin, Daniel and Ortu, Marco},
title = {Mining Valence, Arousal, and Dominance: Possibilities for Detecting Burnout and Productivity?},
year = {2016},
isbn = {9781450341868},
publisher = {Association for Computing Machinery},
address = {New York, NY, USA},
url = {https://doi.org/10.1145/2901739.2901752},
doi = {10.1145/2901739.2901752},
abstract = {Similar to other industries, the software engineering domain is plagued by psychological diseases such as burnout, which lead developers to lose interest, exhibit lower activity and/or feel powerless. Prevention is essential for such diseases, which in turn requires early identification of symptoms. The emotional dimensions of Valence, Arousal and Dominance (VAD) are able to derive a person's interest (attraction), level of activation and perceived level of control for a particular situation from textual communication, such as emails. As an initial step towards identifying symptoms of productivity loss in software engineering, this paper explores the VAD metrics and their properties on 700,000 Jira issue reports containing over 2,000,000 comments, since issue reports keep track of a developer's progress on addressing bugs or new features. Using a general-purpose lexicon of 14,000 English words with known VAD scores, our results show that issue reports of different type (e.g., Feature Request vs. Bug) have a fair variation of Valence, while increase in issue priority (e.g., from Minor to Critical) typically increases Arousal. Furthermore, we show that as an issue's resolution time increases, so does the arousal of the individual the issue is assigned to. Finally, the resolution of an issue increases valence, especially for the issue Reporter and for quickly addressed issues. The existence of such relations between VAD and issue report activities shows promise that text mining in the future could offer an alternative way for work health assessment surveys.},
booktitle = {Proceedings of the 13th International Conference on Mining Software Repositories},
pages = {247–258},
numpages = {12},
location = {Austin, Texas},
series = {MSR '16}
}Annotation
By Ole Peder Brandtzæg, Aaron van Diepen, Rolf Piepenbrink & Jasper Teunissen. 🪧Slides.
Burnout, among other psychological diseases, is prevalent in the software engineering domain and is signified by a loss of interest, lower activity and a sense of powerlessness. Since these are undesirable situations, prevention can prove to be useful. In order to make a first attempt at identifying the symptoms, the paper explores a way to do this with the Valence, Arousal and Dominance (VAD) metrics. Here Valence relates to attractiveness or adverseness, Arousal represents the emotional activation levels and Dominance concerns the sense of control on a stimulus. The authors have extracted the characteristics, title, description and comments of issue reports of the Apache Foundation. To calculate the VAD scores of a piece of text, you use a VAD corpus to find the VAD scores of a subset of words, like the 13,915 word general-purpose lexicon the paper used. For each piece of text, the VAD scores are calculated by taking the difference between the maximum and minimum VAD scores in a piece of text. First, using this approach, the paper evaluates the extent to which VAD relates to issue report characteristics. The authors found that Blocker issues have higher Arousal than Trivial issues. While the differences are statistically significant, the effect sizes between priorities are negligible, so the difference observed in practice is unremarkable. Furthermore, Valence is lowest for Bugs. This supports the idea that developers experience more pleasure from developing new features compared to fixing bugs. Finally, high Dominance is associated with high issue resolution time. Second, a closer look is taken at the VAD change when issues are resolved. It was shown that from the moment an issue is reported to the time when the issue is closed, Valence and Dominance tend to increase, while Arousal has a small decrease. This paper concludes that VAD scores can give relevant information on the emotions associated with the development process. Additionally, this paper finds a correlation between the VAD scores and various aspects of the issue resolving process, like resolution time. Other research on VAD scores indicates that they can indicate an increased risk for various mental issues like burnout. As this paper only mined historical data, the paper does not look at how VAD scores could be used to predict a high risk of burnout or resolution time. Future research is needed to find out if that is possible, and if so, how such a system would be implemented. The method of calculating VAD scores could be improved as well. Right now, a generic English language VAD corpus is used. Future research could look into VAD corpora specialized for software engineering, handling of ‘booster words’ and negations (‘I really like’ or ‘I don’t like’), or finding ways to detect and handle different meanings of the same words. The paper shows that VAD scores give promising results for emotional analysis of the development process, and it provides a solid basis for future research in detecting and preventing burnout in software engineering.