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

The classification of documents is one the problems studied since ancient times and still continues to be studied. With social media becoming a part daily life its misuse, importance text has started increase. This paper investigates effect data augmentation with sentence generation on performance in an imbalanced dataset. We propose LSTM based method, Term Frequency-Inverse Document Frequency (TF-IDF) Word2vec apply Logistic Regression (LR), Support Vector Machine (SVM), K Nearest Neighbour (KNN), Multilayer Perceptron (MLP), Extremly Randomized Trees (Extra tree), Random Forest, eXtreme Gradient Boosting (Xgboost), Adaptive (AdaBoost) Bagging. Our experiment results Offensive Language Identification Dataset (OLID) that machine learning significantly outperforms.