Data and Code for "From fluent to legal: knowledge constraints for Mongolian machine translation"

Description

This dataset contains the data, code, and evaluation artifacts associated with the study “From fluent to legal: knowledge constraints for Mongolian machine translation”. It includes the expanded n=340 morphology evaluation suite, the prior n=120 suite, stem and affix tables, morphology legality checking code, evaluation scripts, UniMorph integration hooks, and experimental summaries used in the reported experiments. The resources support reproduction and analysis of the knowledge-constrained Mongolian machine translation experiments. API keys and private execution traces are not included.

Categories

Licence