In this paper, we propose a novel neural network based architecture which incorporates character, word and lexicon level information to predict the degree of intensity for sentiment and emotion. At first we develop two deep learning models based on Long Short Term Memory (LSTM) & Convolutional Neural Network (CNN), and a feature based model. Each of these models takes as input a fusion of various representations obtained from the characters, words and lexicons. A Multi-Layer Perceptron (MLP) network based ensemble model is then constructed by combining the outputs of these three models. Evaluation on the benchmark datasets related to sentiment and emotion shows that our proposed model attains the state-of-the-art performance.
CITATION STYLE
Ghosal, D., Akhtar, M. S., Ekbal, A., & Bhattacharyya, P. (2018). Deep ensemble model with the fusion of character, word and lexicon level information for emotion and sentiment prediction. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11305 LNCS, pp. 162–174). Springer Verlag. https://doi.org/10.1007/978-3-030-04221-9_15
Mendeley helps you to discover research relevant for your work.