Ensemble Based Emotion Detection Model for Multi-Social Platforms

Abstract

In recent years, there is an exponential growth in public generated data such as image, video and text, this is due to the rapid emanation of diverse social media users. This available textual data is frequently adopted and significantly important for extracting information such as user’s sentiments, and emotions. Considering the complexity and large amount of textual data, the adoption of various machine learning (statistical models), and deep learning model (neural network) for the analysis of emotion has not yet attained optimum accuracy. Recently, Transformer based Architecture (BERT) are achieving state of art accuracy. Hence, this study adopts an ensemble based model using BERT -Large, LSTM and SVM for detecting user’s emotion. The experimental evaluation carried out resulted in an optimum accuracy of 93%.