Artificial Error Generation with Machine Translation and Syntactic Patterns

Open paper

Benchmark	Model	Rank	Results
grammatical-error-detection-on-conll-2014-a1	Ann+PAT+MT	–	F0.5: 21.87
grammatical-error-detection-on-conll-2014-a2	Ann+PAT+MT	–	F0.5: 30.13
grammatical-error-detection-on-fce	Ann+PAT+MT	–	F0.5: 49.11