Penguin Rider Optimization Algorithm-Based Deep Recurrent Neural Network for Sentiment Classification of Political Twitter Data

Abstract

This paper proposes an effective and optimal sentiment classification method named Penguin Rider optimization algorithm-based Deep Recurrent Neural Network (PeROA-based Deep RNN) to perform sentiment classification using political reviews. However, the proposed PeROA is developed by incorporating the Penguins Search Optimization Algorithm (PeSOA) with the Rider Optimization Algorithm (ROA). The sentiment classification process is progressed using the Deep RNN classifier, which in turn generate the optimal solution based on the fitness measure. Accordingly, the function with the minimal error value is accepted as the best solution. The sentiment-based features enable the classifier to perform better classification result with respect to the sentiment tweets. However, the proposed PeROA-based Deep RNN obtained better performance using the metrics, like accuracy, sensitivity, specificity, recall, F-measure, thread score, NPV, FPR,FNR and FDR with the values of 92.030%, 92.030%, 92.235%, 92.030%, 92.030%, 92.030%, 92.030%, 3.105%, 3.11%, and 3.105%, respectively.