Helping testers by fault-prone functionality prediction
Keiichi Tabata, Haruto Tanno, Morihide Oinuma · 2015
In this paper, we propose a technique to predict fault-prone software functionality, instead of fault-prone module. The granularity of prediction is not a line number nor a function name in a source code, but a software functionality from point of view of testers who are dedicated to software testing. Our approach makes it possible for testers to find faults efficiently and effectively on test case creation and test case execution. We applied the proposed technique to an open source project on github. The result graphically suggests that we can predict faultprone software functionality using source code repository mining.