Language Technology at UiT

The Divvun and Giellatekno teams build language technology aimed at minority and indigenous languages

View GiellaLT on GitHub divvungiellatekno/giellalt.uit.no

Page Content

The test results below were gathered before the test bench was available, and were thus collected manually.

Text Words tp fp tn fn precision recall accuracy Version
Njihtan_400 327 4 34 289 0 0.11 1.00 0.90 Alpha
Njihtan_400 327 4 24 299 0 0.14 1.00 0.93 Beta 0.2
Njihtan_400 327 4 19 303 0 0.17 1.00 0.94 Beta 0.3
anemarja 2317 178 65 2071 3 0.73 0.98 0.97 Beta 0.3
Áigot 438 25 43 370 0 0.37 1.00 0.90 Beta 0.3
fredag 401 13 31 357 0 0.30 1.00 0.92 Beta 0.3
UAKK 780 94 13 671 2 0.88 0.98 0.98 Beta 0.4
Markansluska 752 116 28 606 2 0.81 0.98 0.96 Beta 0.4
Iver_Jåks 407 15 33 359 0 0.31 1.00 0.92 Beta 0.5
Mearrasámiide 372 14 9 348 1 0.61 0.93 0.97 Beta 0.5
Finnmárkkus 854 55 32 761 6 0.63 0.90 0.96 Beta 0.5
Eanet_sámi 383 15 12 355 1 0.56 0.94 0.97 Beta 0.5

Sitemap