On the Time-Based Conclusion Stability of Cross-Project Defect Prediction Models

TL;DR

This paper investigates whether defect prediction models' conclusions remain stable over time, revealing that their performance varies significantly across different time periods, which challenges the generalization of empirical findings.

Contribution

It introduces a time-aware evaluation approach for defect prediction models, highlighting the temporal instability of their performance and emphasizing the need for context-specific claims.

Findings

Model performance varies across different time periods.

Product updates can drastically change defect prediction accuracy.

Empirical claims should be limited to specific evaluation contexts.

Abstract

Researchers in empirical software engineering often make claims based on observable data such as defect reports. Unfortunately, in many cases, these claims are generalized beyond the data sets that have been evaluated. Will the researcher's conclusions hold a year from now for the same software projects? Perhaps not. Recent studies show that in the area of Software Analytics, conclusions over different data sets are usually inconsistent. In this article, we empirically investigate whether conclusions in the area of defect prediction truly exhibit stability throughout time or not. Our investigation applies a time-aware evaluation approach where models are trained only on the past, and evaluations are executed only on the future. Through this time-aware evaluation, we show that depending on which time period we evaluate defect predictors, their performance, in terms of F-Score, the area under the curve (AUC), and Mathews Correlation Coefficient (MCC), varies and their results are not consistent. The next release of a product, which is significantly different from its prior release, may drastically change defect prediction performance. Therefore, without knowing about the conclusion stability, empirical software engineering researchers should limit their claims of performance within the contexts of evaluation, because broad claims about defect prediction performance might be contradicted by the next upcoming release of a product under analysis.