This paper describes my efforts in evaluating how editing news headlines can make them funnier within the frames of SemEval 2020 Task 7. I participated in both of the sub-tasks: Sub-Task 1 “Regression” and Sub-task 2 “Predict the funnier of the two edited versions of an original headline”. I experimented with a number of different models, but ended up using DeepPavlov logistic regression (LR) with BERT English cased embeddings for the first sub-task and support vector regression model (SVR) for the second. RMSE score obtained for the first task was 0.65099 and accuracy for the second – 0.32915.

This paper describes the submissions of our team, HAD-Tübingen, for the SemEval 2019 - Task 6: “OffensEval: Identifying and Categorizing Offensive Language in Social Media”. We participated in all the three sub-tasks: Sub-task A - “Offensive language identification”, sub-task B - “Automatic categorization of offense types” and sub-task C - “Offense target identification”. As a baseline model we used a Long short-term memory recurrent neural network (LSTM) to identify and categorize offensive tweets. For all the tasks we experimented with external databases in a postprocessing step to enhance the results made by our model. The best macro-average F1 scores obtained for the sub-tasks A, B and C are 0.73, 0.52, and 0.37, respectively.