Gender-based multi-aspect sentiment detection using multilabel learning

J. Ashok Kumar, Tina Esther Trueman, Erik Cambria

Abstract: Highlights•Identifies multi-aspect sentiment based on genders in online drug reviews.•Explores different feature representations such as BoW, TFIDF, and GloVe.•Compares the problem transformation approach, adapted algorithms, and attention-based recurrent neural networks (RNN).•Evaluates all model based on samples, labels, and ranks measures.