JCTICOL at SemEval-2019 Task 6: Classifying Offensive Language in Social Media using Deep Learning Methods, Word/Character N-gram Features, and Preprocessing Methods

Abstract

In this paper, we describe our submissions to SemEval-2019 task 6 contest. We tackled all three sub-tasks in this task "OffensEval-Identifying and Categorizing Offensive Language in Social Media". In our system called JCTICOL (Jerusalem College of Technology Identifies and Categorizes Offensive Language), we applied various supervised ML methods. We applied various combinations of word/character ngram features using the TF-IDF scheme. In addition, we applied various combinations of seven basic preprocessing methods. Our best submission, an RNN model was ranked at the 25 th position out of 65 submissions for the most complex sub-task (C).

Classification of Imbalanced Offensive Dataset – Sentence Generation for Minority Class with LSTM

“…Graff et al applied B4MSA, FastText, and EvoMSA for classification and obtained a 0.774 F-score with EvoMSA [35]. In [36], 6 different architectures were trained and the most successful among them was RNN with an Fscore of 0.74. In this RNN architecture, uni-gram and bi-gram features were used and in a fully connected (FC) layer, seven different machine learning models were performed.…”