Back to LanguageTool Homepage - Privacy - Imprint

Gebeurd/gebeurt

One of the most common mistakes in Dutch is the d t mistake.
The mistake is unfotunately easy to see, but hard to detect technically using just postags or words.

To thorougly detect these mistakes, full senteces have to be used, so an enormous amount of rules. Or we would need chunks for Dutch, at least some, like for noun parts, adjective blocks and adverb blocks.

A totally different approach would be using AI.
This AI could easily be trained using text from a good text corpus as correct input, and generated bad input by replacing the most common correct verb forms to the wrong one.

It would really help Dutch LT to a much higher level.