TABLE 7.

TextNetTopics performance over top topics for PubMed 20k RCT dataset. #Accumulated_Topics column is the number of significant topics, #Words is the average of words on each level over the 100 iterations.

#Accumulated_Topics	#Words	Accuracy	Sensitivity	Specificity	F1 score	AUC	Precision
10	74.96	0.83	0.74	0.90	0.81	0.88	0.84
8	65.54	0.83	0.74	0.90	0.80	0.88	0.84
6	54.26	0.82	0.71	0.90	0.76	0.87	0.83
4	41.74	0.81	0.69	0.90	0.75	0.86	0.82
2	27.54	0.80	0.66	0.90	0.73	0.84	0.82
1	17.94	0.79	0.63	0.90	0.71	0.82	0.82