h_da Submission for the Germeval Shared Task on the Identification of Offensive Language

This paper describes the Darmstadt University of Applied Sciences (h da) submission to the binary classification task of the Germeval Task 2018 - Shared Task on the Identification of Offensive Language. We submitted three runs, both a combination of lexical resources and an adapted sentiment analysis system. In run 1 (hda coarse 1.txt) and 2 (hda coarse 2.txt), we had a thresholdbased approach (with different thresholds) and in run 3 (hda coarse 3.txt), an approach based on machine learning