HLTDI is a research group for human language technology and the democratization of information led by Michael Gasser. We work on natural language processing tools for under-represented languages, focusing these days on machine translation. You can pronounce "HLTDI" like "HuLaTeDi".
Here are some links to our repositories for current projects:
- L3, our machine translation system based on dependency grammars (code for morphological analyzers is also here)
- Cross-lingual word sense disambiguation -- tools for lexical choice in hybrid machine translation systems
- webapps for collecting corpora for under-resourced languages: Guampa and Tahekami. We're building these out for Guarani first. Watch for a public launch pretty soon.
- Other projects might collect on our github organization page...
- ... and also on Michael Gasser's research software page.