I’ve been trying to follow the instructions on http://wiki.languagetool.org/finding-errors-using-n-gram-data to get a local instance of Language Tool to identify the errors in the test sentences on that page (e.g. “I can’t remember how to go their.”)
On languagetool.org this error is identified using the
CONFUSION_RULE. This doesn’t work when I use it locally configured with the ngram dataset. When looking at the source code, the
CONFUSION_RULE doesn’t seem to use or look at the ngram dataset that is configured in the
SuggestionsOrdererConfig (as far as I can tell this is the only place in the HTTP server where this configuration parameter is passed). This config appears to only be used in the
SuggestionsOrderer (in turn by the
Any thoughts as to what I might be missing?