Skip to Main Content
Table 4: 

Test vs. dev: Results with fastText for all Models. morph-gold presents an ideal upper-bound.

EvalModeldevtest
Morph-Level morph gold 80.03± 0.4 79.10± 0.6 
morph hybrid 78.51 ± 0.5 77.11 ± 0.7 
morph standard 72.79 ± 0.5 69.52 ± 0.6 
token-multi hybrid 75.70 ± 0.5 74.64 ± 0.3 
 
Token-Level morph gold 80.30± 0.5 79.28± 0.6 
morph hybrid 79.04 ± 0.5 77.64 ± 0.7 
morph standard 74.52 ± 0.7 73.53 ± 0.8 
token-multi 77.59 ± 0.4 77.75 ± 0.3 
token-single 78.15 ± 0.3 77.15 ± 0.6 
EvalModeldevtest
Morph-Level morph gold 80.03± 0.4 79.10± 0.6 
morph hybrid 78.51 ± 0.5 77.11 ± 0.7 
morph standard 72.79 ± 0.5 69.52 ± 0.6 
token-multi hybrid 75.70 ± 0.5 74.64 ± 0.3 
 
Token-Level morph gold 80.30± 0.5 79.28± 0.6 
morph hybrid 79.04 ± 0.5 77.64 ± 0.7 
morph standard 74.52 ± 0.7 73.53 ± 0.8 
token-multi 77.59 ± 0.4 77.75 ± 0.3 
token-single 78.15 ± 0.3 77.15 ± 0.6 
Close Modal

or Create an Account

Close Modal
Close Modal