Neural Network

19.01.2021
14:30-15:00
Chiara, Tommi, Linda

which analysis do we use to get the part-of-speech
hfst-tokenise --print-all --giella-cg --no-weights --unique tokeniser-disamb-gt-desc.pmhfst | vislcg3 --grammar mwe-dis.bin | cg-mwesplit | vislcg3 --grammar disambiguator.bin | vislcg3 --grammar korp.bin | vislcg3 --grammar dependency.bin

lang-sme/src/cg3/disambiguator.cg3

TODO
* splitting script writes to file correct and splitted, and a second script analyse those and add pos to json
* if more analyses, take the first one