POS tagger intrinsic evaluation - experiment 2 by matej
Try it out
By clicking the button below you will create a copy of this workflow and will be able to edit it, run it or just look at it without harming the original.
Launch workflow!Description
In this experiment we measure the influence of the size of the training set on POS tagger performance. 7 different POS taggers are trained on first 1000, 2000, 3000 and all sentences of the NLTK WSJ sample. Their performance is measured on GUM corpus.
Try it live
By clicking on the button below you will launch the workflow and be able to run it or edit it.
Launch workflow!
