Skip to Main Content
Table 7 

Tagging accuracies (%) relative to sizes of training data. Size = number of sentences in the labeled training corpus. Bold identifies best performance at the given size.

SizeSupervised+c100+c500+c1000
100 66.94 76.78 71.91 68.59 
500 77.89 84.19 82.17 81.14 
1000 83.70 87.68 87.74 86.49 
5000 90.57 91.84 91.93 91.95 
10000 93.17 94.00 93.91 94.02 
15000 93.76 94.42 94.47 94.39 
20000 94.11 94.76 94.58 94.73 
SizeSupervised+c100+c500+c1000
100 66.94 76.78 71.91 68.59 
500 77.89 84.19 82.17 81.14 
1000 83.70 87.68 87.74 86.49 
5000 90.57 91.84 91.93 91.95 
10000 93.17 94.00 93.91 94.02 
15000 93.76 94.42 94.47 94.39 
20000 94.11 94.76 94.58 94.73 
Close Modal

or Create an Account

Close Modal
Close Modal