See T144121: Add language support for Meta-Wiki
This task is done when there is a model train for metawiki that implements a revert predictor with a /reasonable/ level of fitness (>= 0.85 ROC-AUC).
See T144121: Add language support for Meta-Wiki
This task is done when there is a model train for metawiki that implements a revert predictor with a /reasonable/ level of fitness (>= 0.85 ROC-AUC).
I think we should do some comparison in unicode character ranges to detect language type issues generically.
E.g.
>>> ord("a") # Latin (English) 97 >>> ord("ジ") # Japanese 12472 >>> ord("ز") # Arabic (Persian) 1586