Feature Reduction Based on Genetic Algorithm and Hybrid Model for Opinion Mining

Table 7

Classification performance of kitchen domain review.

Class/method	NB			LR			SVM			BSVM			BNB
	Actual neg.	Actual pos.	Acc. (%)	Actual neg.	Actual pos.	Acc. (%)	Actual neg.	Actual pos.	Acc. (%)	Actual neg.	Actual pos.	Acc. (%)	Actual neg.	Actual pos.	Acc. (%)
Unigram
Predicted neg.	65	33	71.42	75	49	82.41	59	16	64.84	68	12	74.73	69	23	75.82
Predicted pos.	26	66	66.66	16	50	50.00	32	83	83.83	23	87	87.87	22	76	76.76
Avg. accuracy	73.03%			65.79%			74.74%			81.58% (↑)			76.32%
Error rate	36.31%			36.84%			24.73%			18.42% (↓)			23.68%
Unigram + bigram
Predicted neg.	67	30	73.62	72	51	79.12	63	19	69.23	64	10	70.32	74	24	81.31
Predicted pos.	24	69	69.69	19	48	48.48	28	80	80.80	27	89	89.89	17	75	75.75
Avg. accuracy	75.26%			63.16%			75.26%			80.53% (↑)			78.42%
Error rate	28.42%			36.84%			24.73%			19.47% (↓)			21.57%