[Retracted] Analysis and Application of the Ideological and Political Evaluation System of College Students Based on Text Mining

Table 4

The performance comparison of the SVM classifier before and after the data preprocessing in the ideological and political evaluation topic of the data set.

Special topic of ideological and political evaluation	Traditional SVM model					The SVM model is based on the word frequency statistics rule
	Accuracy (% )	Precision (%)	Recall (%)	F1 (%)	Time (s)	Accuracy (%)	Recall (%)	Precision (%)	F1 (%)	Time (s)
Acq	87.89	76.36	70.99	73.58	2279	88.63	78.03	72.53	75.18	611
Crude	94.28	50.61	47.98	49.25	2037	94.55	53.09	49.71	51.34	574
Earn	84.65	84.63	75.02	79.54	2285	86.29	86.99	77.04	81.71	637
Grain	92.94	70.49	61.70	65.80	2196	93.38	72.98	63.22	67.75	605
Interest	94.55	42.75	41.26	41.99	1986	94.28	45.59	43.36	44.45	632
Money	93.92	58.12	54.88	56.45	1959	94.15	59.80	56.74	58.23	553
Ship	96.59	39.74	36.05	37.81	1891	96.72	42.68	40.70	41.67	548
Trade	94.68	45.04	40.41	42.60	2025	94.92	47.73	43.15	45.32	571
Macro average	92.44	58.47	53.66	55.96	2082	92.93	60.86	55.81	58.23	591