Rumor Identification with Maximum Entropy in MicroNet

Table 1

Micro-average accuracy comparison among different training sets.

Number of features	Training set 1		Training set 2
	Feature function	Improved feature function	Feature function	Improved feature function
50	70.84	73.30	72.21	72.35
100	73.23	75.01	76.27	76.56
150	76.89	78.96	78.62	80.24
200	77.11	79.64	80.34	83.43
250	76.61	77.36	78.48	79.03
300	75.20	76.04	75.42	75.09