Parameter Tuned Bi-Directional Long Short Term Memory Based Emotion With Intensity Sentiment Classification Model Using Twitter Data

With the advanced growth in the Internet and social networking technologies, a massive amount of comments are being generated on the Web at each and every second. In the age of big data, mining the sentiments and emotion classification using artificial intelligence (AI) techniques gained significant interest to properly understand the opinion of the public. This paper presents a new hyperparameter tuned bi-directional long short term memory (Bi-LSTM) using differential evolution (DE) model called DE-BiLSTM for the emotion with intensity based sentiment classification in Twitter data. The hyperparameters of Bi-LSTM namely batch size and number of hidden layers are determined by means of differential evolution algorithm (DE). The proposed method initially undergoes preprocessing is several stages like word to feature vector conversion and feature extraction process takes place. Finally, a softmax based classification process is carried out to identify the different intensities with diverse classes that exist in the tweets. The validation of the DE-BiLSTM model is carried out against SEMEVAL2018 Task-1Emotion Intensity Ordinal Classification dataset. The simulation results indicated that the DE-BiLSTM model has outperformed the other methods with an average precision of 93.83%, recall of 90.41%, F-measure of 91.76% and accuracy of 96.42%.