support more languages to learn in

[StephanAkkerman]

If we want to support more languages to learn in, we need to add more models that support languages other than English. This is an overview of what needs to be updated:

• IPA dataset for languages (used to generate phonetic sim feature vectors)
  • https://github.com/CUNY-CL/wikipron/tree/master/data/scrape
  • https://kaikki.org/dictionary/rawdata.html
  • https://github.com/open-dict-data/ipa-dict/tree/master/data (used by G2P model)
• FastText (or other semantic sim) support for the language (157 language support)
• Imageability model (or just translate to English and use the same model)
• Update the support languages page

The datasets that we might be dependent on are mainly used in:

• 10

[StephanAkkerman]

We are already using the wikipron data for English IPA (US Broad)

[StephanAkkerman]

Maybe we also need a dataset for the narrow English so that we can use that for foreign languages that use narrow transcriptions?

[StephanAkkerman]

Voor nu alleen ondersteuning in: Engels, Nederlands, Spaans, Frans, Chinees?