Hybrid Deep Learning Approach for Sentiment Classification of Malayalam Tweets

Social media content in regional languages is ex-panding from day to day. People use different social platforms express their suggestions and thoughts native languages. Sentiment Analysis (SA) the known procedure for identifying hidden sentiment present sentences categorizing it as positive, negative, or neutral. The SA of Indian challenging due unavailability benchmark datasets lexical resources. analysis has been done using lexicon, Machine Learning (ML), Deep (DL) techniques. In this work, baseline models hybrid Neural Network (DNN) architecture have used classification Malayalam tweets negative Since, sentiment-tagged dataset not readily available, on manually created translated Kaggle dataset. study combine Convolutional Networks (CNN) with variants Recurrent Net-works (RNN). RNN are Long Short-Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM) Gated Unit (GRU). All these improve performance Classification (SC) compared LSTM, Bi-LSTM GRU.