This small project is part of a semester project. The goal is to build a statistical grammarchecker. The error detection was done using ngram counts from the google ngram corpus. As a proof of concept of the statistical model german was added as a second language. For a little bit more detailed explanation of the implemented concepts see "NLP - GrammarChecker.pdf".
To start it run the docker-compose.yml or if you want to install all the necessary python modules install them according to requirements.txt and run the app.py file.