PerAnSel:  A  Novel Deep Neural Network-Based System for Persian Question Answering

Table 10

The MRR measure of the PerAnSel method on PASD, PerCQA, and WikiFA.

System	Dataset
	PASD (%)	PerCQA (%)	WikiFA (%)
ASBERT-MBERT
CETE-MBERT		61.36	42.74
SOVWO		52.28	30.84
OWO-ParsBERT		55.68	43.46
OWO-AlbertFA		58.21	48.47
OWO-DistilmBERT		61.02	50.36
OWO-MBERT		66.47	52.96
PERSEL-ParsBERT		55.97	50.38
PERSEL-AlbertFA	77.21	59.78	47.59
PERSEL-DistilmBERT	81.55	65.44	62.66
PERSEL-MBERT	89.36	69.48	66.08