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