PoSTagger on other languages than English

[tencnivel]

Has anyone manage to use TextModels PoSTagger on another language than English?

The following does not work, everything gets tagged as ‘NNP’:

using TextModels
using TextAnalysis
using Languages

str = "André trouve cette maison très belle" 
sd = StringDocument(str)
pos = PoSTagger()
language!(sd,Languages.French())
pos(sd)

The doc is here
https://juliatext.github.io/TextAnalysis.jl/dev/features/#Neural-Model-for-Part-of-Speech-tagging-using-LSTMs,-CNN-and-CRF

[avik]

I think that model was only trained on English, unfortunately. Would be great to get models for other languages. Maybe we should document this better.

[tencnivel]

In .julia/datadeps/POS Model Dicts/word_to_embed_idx.json there are French words as well.

I thought that this meant that a pretrained model was available for French as well.

@avik, what would you recommend between these following options:

• Run an external command to PoS tagger that supports French (eg. The Stanford Natural Language Processing Group)
• Build a native Julia PoS tagging model for French

In the second case, some guidance or links to some resources would be very welcome.