Hybrid Machine Learning Method for Sentiment Analysis

Sentiment Analysis classifies the sentiments that are represented in the text data and the sentiments categorized as negative or positive. This research work introduces a hybrid machine learning method to improve the accuracy of prediction of this sentiment analysis process. This hybrid method used Support Vector Machines (SVM) with Recurrent Neural Networks (RNN), to categorize sentiments into their linear component using the SVM and RNN components used to capture the text’s intricate temporal linkages efficiently. The hybrid model may provide both a detailed sentiment assessment based on salient characteristics and a nuanced sentiment analysis that considers contextual information by combining these two approaches. The findings indicate that the SVM-RNN hybrid model regularly performs better than its separate components and other traditional approaches. This is evident in its ability to achieve higher levels of accuracy 93.6% in sentiment classification.