Mansi NLP Grammar

Finite state and Constraint Grammar based analysers, proofing tools and other resources

View the project on GitHub giellalt/lang-mns

Work plan

This document contains short-term goals for the work at hand.

Spring 2023

Grammatical work on the language model

Testing

  1. Go through yaml files
    1. Proofread them
    2. Carefully extend on a need-to-test basis
    3. Improve the analyser accordingly, with the goal of achieveing 0 errors
  2. Keep lemma errors down at 0

Language model (fst) work

  1. Follow-up “Csilla’s homework”: Correct all verb entries in the stems/verb.lexc file
  2. Morphophonology (errors via yaml)

Vocabulary work on the language model

This shoiuld come after the inflectional part of the fst is in place, and the textbook is closer to being finished.

  1. Analyse the textbook with the fst and achieve 100% coverage
  2. Analyse a large corpus and improve coverage to above 90+ %

Practical issues

Git and Tower

Speller installation

Fall 2023