Studying the association between Gitcoin's issues and resolving outcomes
Issued Date
2023-12-01
Resource Type
ISSN
01641212
Scopus ID
2-s2.0-85171843513
Journal Title
Journal of Systems and Software
Volume
206
Rights Holder(s)
SCOPUS
Bibliographic Citation
Journal of Systems and Software Vol.206 (2023)
Suggested Citation
Choetkiertikul M., Puengmongkolchaikit A., Chandra P., Ragkhitwetsagul C., Maipradit R., Hata H., Sunetnanta T., Matsumoto K. Studying the association between Gitcoin's issues and resolving outcomes. Journal of Systems and Software Vol.206 (2023). doi:10.1016/j.jss.2023.111835 Retrieved from: https://repository.li.mahidol.ac.th/handle/20.500.14594/90258
Title
Studying the association between Gitcoin's issues and resolving outcomes
Author's Affiliation
Other Contributor(s)
Abstract
The development of open-source software (OSS) projects usually have been driven through collaborations among contributors and strongly relies on volunteering. Thus, allocating software practitioners (e.g., contributors) to a particular task is non-trivial and draws attention away from the development. Therefore, a number of bug bounty platforms have emerged to address this problem through bounty rewards. Especially, Gitcoin, a new bounty platform, introduces a bounty reward mechanism that allows individual issue owners (backers) to define a reward value using cryptocurrencies rather than using crowdfunding mechanisms. Although a number of studies have investigated the phenomenon on bounty platforms, those rely on different bounty reward systems. Our study thus investigates the association between the Gitcoin bounties and their outcomes (i.e., success and non-success). We empirically study over 4,000 issues with Gitcoin bounties using statistical analysis and machine learning techniques. We also conducted a comparative study with the Bountysource platform to gain insights into the usage of both platforms. Our study highlights the importance of factors such as the length of the project, issue description, type of bounty issue, and the bounty value, which are found to be highly correlated with the outcome of bounty issues. These findings can provide useful guidance to practitioners. Editor's note: Open Science material was validated by the Journal of Systems and Software Open Science Board.