Ensemble Deep Learning Model to Evaluate Sentiments in Social media Posts

The exponential rise in social media users makes it more difficult to comprehend what other netizens think about a certain topic. Therefore, sentiment analysis is essential for assessing the same through the use of cutting-edge technologies such as deep learning models, machine learning, and natural language processing (NLP).Numerous deep learning models are currently in use to examine the sentiments present in various social media datasets. For a thorough examination of sentiments, natural language processing and a single neural network are insufficiently powerful. Therefore, the suggested model uses the sentiment analysis ensembling learning mechanism to combine the two neural networks, which ultimately serves as a catalyst for the process of generating appropriate decisions in sentiment analysis. The proposed design combines two neural networks—a long short term memory (LSTM) neural network and a convolution neural network (CNN)—to train on data gathered from publicly accessible repositories, including posts from Facebook, Instagram, and X (formerly Twitter). To improve the sentiment analysis process utilizing deep learning models, the training data from the ensembled model is fed into the random forest classification model.