Skip to Main Content
Table 5: 
The pivoting language, performance (and their n-gram overlap % with the LRL) selected by different criteria. δ column shows the top-30 candidate recall improvement (%) using n-gram overlap. Language with a hat use grapheme representations while the remaining ones use phoneme representations.
LRLLinguisticsn-gram Overlapδ
ti am^, 63.9 (60.8) am, 74.2 (70.9) 10.3 
om so^, 28.0 (63.7) id^, 40.9 (75.8) 12.9 
rw rn^, 46.4 (62.9) tl, 64.6 (79.0) 18.2 
si hi, 50.4 (63.1) hi, 50.4 (63.1) 
lo th, 51.4 (78.8) th, 51.4 (78.8) 
mr hi^, 72.8 (83.3) hi^, 72.8 (83.3) 
te ta, 12.6 (32.3) hi, 32.6 (45.1) 20.0 
LRLLinguisticsn-gram Overlapδ
ti am^, 63.9 (60.8) am, 74.2 (70.9) 10.3 
om so^, 28.0 (63.7) id^, 40.9 (75.8) 12.9 
rw rn^, 46.4 (62.9) tl, 64.6 (79.0) 18.2 
si hi, 50.4 (63.1) hi, 50.4 (63.1) 
lo th, 51.4 (78.8) th, 51.4 (78.8) 
mr hi^, 72.8 (83.3) hi^, 72.8 (83.3) 
te ta, 12.6 (32.3) hi, 32.6 (45.1) 20.0 
Close Modal

or Create an Account

Close Modal
Close Modal