Lemmatisation

Top  Previous  Next

Lemmatisation is the process of normalising morphologically complex and inflected words to their base or dictionary form, i.e. the lemma of the word.

 

Evaluation results for Lemmatisation:

(As partly reported in Du Toit, J. S., & Puttkammer, M. J. 2021. Developing Core Technologies for Resource-Scarce Nguni Languages. Information, 12(12), 520. https://doi.org/10.3390/info12120520)

 

Language

Accuracy

Afrikaans

91.64%

isiNdebele

90.35%

isiXhosa

92.99%

isiZulu

90.33%

Sesotho

82.03%

Sesotho sa Leboa

90.10%

Setswana

83.45%

Siswati

90.20%

Tshivenḓa

71.47%

Xitsonga

84.57%