Faul t - proneness Filtering:スパムフィルタに基づく不具合混入ソフトウェアモジュールの予測手法

摘要

Prediction of fault-prone software modules has been one of the most classical and importaantarea of software engineenring so far. Many aproaches has been carried out using softwarecomplexity metrics and mathematical models. Such approaches, however, have difficulties in collecting the metrics and construcling mathematical models based on the metrics. We proposed a novel aproach for predicting fault-prone modules using a spam fi1tering technique, named Fault-proneFiltering.In ourapproach,faultpronemodulesaredetectedin away thatthesourcecodemod ules considered text filesaand are applied to the spam filter directly. For the validation of our method, weperfo(are) rmedacross-v(as)alidation experimentusing anopensourcesoftwareproject.