Table 2. The advantages of suggested FLMatch over BERT are shown by comparing performance measures and t-tests.

	QA TYPE-I		QA TYPE-II		QA TYPE-III		QA TYPE-IV
Models	MAP	MRR	MAP	MRR	MAP	MRR	MAP	MRR
BERT	0.7109	0.71829	0.8919	0.82617	0.8671	0.8719	0.8181	0.8919
Proposed F LMatch model	0.7516	0.7618	0.9102	0.8178	0.8919	0.9192	0.8367	0.9181