Erzya-Moksha shallow-transfer machine translation for measurement of language diversity

  • Rueter, Jack (Project manager)
  • Erina, Olga (Project manager)
  • Kabaeva, Nadezhda (Project manager)

Project: Research project

Project Details

Description (abstract)

The Mordvin language forms, Erzya and Moksha, with their separate dialects are closely related. They both provide, however, ample lexical, morphological, syntactic and phraseological evidence for language diversity. By setting up a shallow-transfer MT system the diversity of these two language forms can be calculated, and the resulting measurements can be compared to analogical measures in other language pairs.

Layman's description

This project sets up a machine translation system where increments (lexical, morphological and syntactic structures) are transferred at face value from one language to another (shallow transfer). The system is then used specifically for measuring the linguistic distance between the two Mordvin standard languages, Erzya and Moksha.
Short titleMordvin shallow-transfer MT
StatusActive
Effective start/end date26/06/2019 → …

Fields of Science

  • 6121 Languages
  • open-source
  • shallow-transfer machine translation
  • finite-state analysis
  • Mordvin languages
  • Erzya language
  • Moksha language
  • Apertium
  • Giella
  • finite-state morphology
  • multilingual dictionaries
  • XML