In natural language processing, sentiment analysis of reviews is a big task. The deep learning techniques promising high accuracy compared with traditional machine learning algorithms like SVM, Naïve Bayes, KNN, etc. In this paper, we implemented a multi-channel Convolutional Neural Network-Long Short Term Memory (CNN-LSTM) model, which consist of multi-channel CNN and LSTM. The CNN extracts active local features using different filters as n-grams. This information is passed sequentially to the LSTM, which captures the long term dependencies across the review in the process of classification. The model is evaluated using accuracies, and the results show that the proposed model outperforms the baseline algorithms.
Volume 11 | 12-Special Issue