Empirical validation of human factors in predicting issue lead time in open source projects

• DOI: 10.1145/2020390.2020403
• Publication Date: 2011-09-23T15:13:10Z
• Authors: Nguyen Duc Anh; Daniela S. Cruzes; Reidar Conradi; Claudia Ayala
• Keywords: Lead (geology); Empirical Research

[Context] Software developers often spend a significant portion of their resources resolving submitted evolution issue reports. Classification or prediction lead time is useful for prioritizing issues and supporting human allocation in software maintenance. However, the predictability still research gap that calls more empirical investigation. [Aim] In this paper, we empirically assess different types models using factor measures collected from tracking systems. [Method] We conduct an investigation three active open source projects. A machine learning based classification statistical univariate multivariate analyses are performed. [Results] The accuracy ten-fold cross-validation varies 75.56% to 91%. R2 value linear regression ranges 0.29 0.60. Correlation analysis confirms effectiveness collaboration measures, such as number stakeholders comments, models. assignee past performance also effective indicator time. [Conclusions] results indicate average important variables constructing should be explored achieve better performance.