On the relative value of cross-company and within-company data for defect prediction

By Burak Turhan, Tim Menzies, Ayse Basar Bener, Justin S. Di Stefano. Empirical Software Engineering, 2009

696 citations recorded by the source corpus.

Subfields: Information Systems, Software

View Paper: https://www.semanticscholar.org/paper/cfe0a2bf6fb69e97b2d8173df35518e67efbf537

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Systems

In Software

Citations in the corpus

Defect prediction from static code features: current results, limitations, new approaches (2010, background): “In order to make a claim that other researchers can verify, we designed a controlled experiment to assess Predictions 1 and 2 in a reproducible manner (Turhan et al. 2009).”
Heterogeneous Defect Prediction (2015, methodology): “The basic idea of the NN filter is that prediction models are built by source instances that are nearest-neighbours of target instances [12].”
Transfer defect learning (2013, methodology): “[43] analyzed Cross-Company (CC) and Within-Company (WC) data for defect prediction, and they confirmed the difficulty of reusing CC data to predict defects in the software of other companies.”
Towards identifying software project clusters with regard to defect prediction (2010, background): “[21] obtained probability of false alarm greater than 50% in most of the cross project predictions.”
A Systematic Literature Review on Fault Prediction Performance in Software Engineering (2012, positive): “[[15]] report that Regression via Classification (RvC) works well.”
Evaluating defect prediction approaches: a benchmark and an extensive comparison (2011, background): “The categories are: All bugs, non trivial bugs (severity>trivial), major bugs (severity>major), critical bugs (critical or blocker severity) and high priority bugs (priority>default).”
An empirical study on software defect prediction with a simplified metric set (2014, methodology): “Hall et al. [22] investigated how the context of models, the independent variables used, and the modeling techniques applied affected the performance of defect prediction models according to 208 defect prediction studies.”
How, and why, process metrics are better (2013, background): “Portability (also called “cross-project prediction”) recently attracted quite a bit of attention [14, 22, 24, 25, 28]; but we have not found any that compare the portability of different types of metrics.”
Automatically Learning Semantic Features for Defect Prediction (2016, methodology): “Turhan et al. [61] proposed to use a nearest-neighbor ﬁlter to improve CPDP. Nam et al. [42] proposed TCA+, which adopted a state-of-the-art technique called Transfer Component Analysis (TCA) and optimized TCA’s normalization process to improve CPDP.”
A systematic review of machine learning techniques for software fault prediction (2015, background): “Hence, total number of 122 studies [1, 2, 20-139] were identified for further processing and analysis.”
Deep Learning for Just-in-Time Defect Prediction (2015, methodology): “Prior studies mainly focus on predicting defects at coarse granularity level, such as file, package, or module [4], [5], [6].”
Software fault prediction metrics: A systematic literature review (2013, result): “[143] did not find cross-company models to be useful when compared to internal company models.”