In this section, we report the experimental results in detail.
In section IV-A, we report the results from comparing the
effort-aware ranking performance for the text mining based
and the metrics based models (RQ1). In section IV-B, we
report the results from comparing the effort-aware
classification performance for these two kinds of models
(RQ2). In section IV-C, we report the results from statistically
significant test and effect sizes of the differences between text
mining based and metrics based models to examine whether
text mining based model is more cost-effective from the
viewpoint of practical application (RQ3).