“…Meanwhile, for detecting hate speech and abusive language in text using the deep-learning approach, researchers have been trying to use several deep-learning models including Deep Neural Network (DNN) [41] and Long Short Term Memory (LSTM) [46] , [50] , [26] . These deep-learning models are combined with several features like POS [41] , TFIDF [41] , and word embedding (word2vec and FastText) [46] , [50] , [26] .…”