PolyglotToolkit

PolyglotToolkit

A mostly-coherent set of projects all in the category of free and open-source language learning software.

They may be written in any programming language and framework - whichever one makes sense for the project at hand.

Created and maintained by yelircaasi.

Projects

parallel-text-aligner

Tools for aligning pairs of texts in different languages (or different translations in the same language).

parallel-text-finisher

Tools for creating attractive output formats from aligned parallel texts.

henry-higgins

Application for pronunciation training. Powered by neural networks, speech signal processing techniques, and speech-to-text alignment algorithms.

grecromancer

High-quality, end-to-end TTS for Classical Latin and Attic Ancient Greek.

WiktionaryTool.jl

Julia package for investigating cognates and borrowings, with a focus on their use in language learning.

Related Projects

CLTK

Roadmap

  • minimal working version of parallel-text-finisher, with nice basic CLI interface
  • Gale-Church algorithm coded and build in adapted version of original C code
  • port Gale-Church to Cython, or alternatively make bindings to the C code
  • minimal working version of parallel-text-aligner with nice basic CLI interface
  • investigate Wiktionary API to determine how much needs to be done via HTML vs API
  • find best Julia packages needed by WiktionaryTool.jl
  • get minimal version of WiktionaryTool.jl working
  • look into native compilation and distribution options for Julia (i.e. binary for people who don't want to install Julia)
  • plan subsequent tasks