An Approach to Semantic and Structural Features Learning for Software Defect Prediction

Table 2

Details of the datasets.

Project	Releases	Avg. (#files)	Avg. (defect rate) (%)	IR (imbalance rate)
Camel	1.4, 1.6	892	18.6	5.01
Lucene	2.0, 2.2	210	55.7	1.14
Poi	2.5, 3.0	409	64.7	0.55
Synapse	1.1, 1.2	239	30.5	2.70
Xalan	2.5, 2.6	815	48.5	1.07
Xerces	1.2, 1.3	441	15.5	5.20