Applying Deep Learning Technologies to Evaluate the Patent Quality with the Collaborative Training

Table 9

Data transfer learning results of 2208 Chinese patent English text data after active learning expansion.

Category	Training set				Test set
	Precision	Recall	-score	Support	Precision	Recall	-score	Support
1	0.93	0.96	0.95	676	0.93	0.83	0.88	35
2	0.85	0.71	0.77	404	0.88	0.78	0.82	18
3	0.79	0.85	0.81	506	0.74	0.81	0.77	21
4	0.59	0.77	0.67	258	0.39	0.70	0.50	10
5	0.61	0.64	0.62	154	0.50	0.38	0.43	8
6	0.41	0.54	0.47	69	0.33	0.33	0.33	3
7	0.63	0.54	0.58	87	0.33	0.67	0.45	3
8	0.78	0.37	0.50	54	1.00	0.50	0.67	2
Microavg	0.79	0.80	0.79	2208	0.74	0.74	0.74	100
Macroavg	0.70	0.67	0.68	2208	0.64	0.63	0.63	100