Using Sentence-Level Neural Network Models for Multiple-Choice Reading Comprehension Tasks

Table 3

Comparison of different reader model on three testing datasets.

Method	BCEETest	SBCEETest1	SBCEETest2
HABCNN-TE	0.428	0.442	0.438
Match Reader	0.461	0.452	0.455
SM Reader	0.495	0.491	0.499
CAS Reader	0.513	0.503	0.516
Sentence-Level Attention Reader (mode:max+avg)	0.581	0.535	0.578