Characterizing the hyperspecialists in the context of crowdsourcing software development
Tóm tắt
Companies around the world use crowdsourcing platforms to complete simple tasks, collect product ideas, and launch advertising campaigns. Recently, crowdsourcing has also been used for software development to run tests, fix small defects, or perform small coding tasks. Among the pillars upholding the crowdsourcing business model are the platform participants, as they are responsible for accomplishing the requested tasks. Since successful crowdsourcing heavily relies on attracting and retaining participants, it is essential to understand how they behave. This exploratory study aims to understand a specific contributor profile: hyperspecialists. We analyzed developers’ participation on challenges in two ways. First, we analyzed the type of challenge that 664 Topcoder platform developers participated in during the first 18 months of their participation. Second, we focused on the profile of users who had more collaborations in the development challenges. After quantitative analysis, we observed that, in general, users who do not stop participating have behavioral traits that indicate hyper-specialization, since they participate in the majority of the same types of challenge. An interesting, though troubling, finding was the high dropout rate on the platform: 66% of participants discontinued their participation during the study period. The results also showed that hyperspecialization can be observed in terms of technologies required in the development challenges. We found that 60% of the 2,086 developers analyzed participated in at least 75% of challenges that required the same technology. We found hyperspecialists and non-specialists significantly differ in behavior and characteristics, including hyperspecialists’ lower winning rate when compared to non-specialists.
Tài liệu tham khảo
Mao K, Wang Q, Jia Y, Harman M (2016) PREM: Prestige Network Enhanced Developer-Task Matching for Crowdsourced Software Development. Department of Computer Science, UCL (University College London), London.
Tunio MZ, Luo H, Cong W, Fang Z, Gilal AR, Abro A, Wenhua S (2017) Impact of personality on task selection in crowdsourcing software development: A sorting approach. IEEE Access 5:18287–18294.
Xie H, Lui JCS, Jiang JW, Chen W (2014) Incentive mechanism and protocol design for crowdsourcing systems In: 2014 52nd Annual Allerton Conference on Communication, Control, and Computing (Allerton), 140–147.. IEEE, Washington, DC. https://doi.org/10.1109/ALLERTON.2014.7028448.
Karim MR, Messinger D, Yang Y, Ruhe G (2016) Decision support for increasing the efficiency of crowdsourced software development. CoRR abs/1610.04142. 1610.04142.
Zanatta A, Steinmacher I, Machado L, de Souza C, Prikladnicki R (2017) Barriers faced by newcomers to software-crowdsourcing projects. IEEE Softw 34(2):37–43.
Dubey A, Abhinav K, Taneja S, Virdi G, Dwarakanath A, Kass A, Kuriakose MS (2016) Dynamics of software development crowdsourcing In: 2016 IEEE 11th International Conference on Global Software Engineering (ICGSE), 49–58.. IEEEE, Washington, DC. https://doi.org/10.1109/ICGSE.2016.13.
Mao K, Capra L, Harman M, Jia Y (2015) A survey of the use of crowdsourcing in software engineering. J Syst Softw 126:57–84.
Saremi RL, Yang Y (2015) Dynamic simulation of software workers and task completion In: 2015 IEEE/ACM 2nd International Workshop on CrowdSourcing in Software Engineering, 17–23.. IEEE, Washington, DC. https://doi.org/10.1109/CSI-SE.2015.11.
Surowiecki J (2005) The Wisdom of Crowds: Why the Many Are Smarter than the Few, New York.
LaToza TD, van der Hoek A (2016) Crowdsourcing in software engineering: Models, motivations, and challenges. IEEE Softw 33(1):74–80.
Yang Y, Saremi R (2015) Award vs. worker behaviors in competitive crowdsourcing tasks In: 2015 ACM/IEEE International Symposium on Empirical Software Engineering and Measurement (ESEM), 1–10, Washington, DC. https://doi.org/10.1109/ESEM.2015.7321192.
Ambreen T, Ikram N (2016) A state-of-the-art of empirical literature of crowdsourcing in computing In: 2016 IEEE 11th International Conference on Global Software Engineering (ICGSE), 189–190.. IEEE, Washington, DC.
Malone TW, Laubacher RJ, Johns T (2011) The age of hyperspecialization. Harv Bus Rev 89(7-8):56–65.
Hosseini M, Phalp K, Taylor J, Ali R (2014) The four pillars of crowdsourcing: A reference model In: RCIS Conference 2014, 1–12.. IEEE, Washington, DC.
Gadiraju U (2015) Make hay while the crowd shines: Towards efficient crowdsourcing on the web In: Proceedings of the 24th International Conference on World Wide Web, 493–497.. ACM, New York.
Gray ML, Suri S, Ali SS, Kulkarni D (2016) The crowd is a collaborative network In: Proceedings of the 19th ACM Conference on Computer-Supported Cooperative Work & Social Computing, 134–147.. ACM, New York.
Mao K, Yang Y, Wang Q, Jia Y, Harman M (2015) Developer recommendation for crowdsourced software development tasks In: 2015 IEEE Symposium on Service-Oriented System Engineering, 347–356, Washington, DC. https://doi.org/10.1109/SOSE.2015.46.
Hinkle DE, Wiersma W, Jurs SG, et al (2002) Applied Statistics for the Behavioral Sciences. 5th edn. Houghton Mifflin, Boston.