Multimodal Deep Learning Model for Measuring the Impact of Social Media Advertising Using Visual Linguistic Representation Learning

The research examines the application of Natural Language Processing (NLP) and Deep Convolutional Neural Networks (Deep-CNN) in forecasting social media activity. The study aims to improve Social Media Awareness (SMA) by integrating these technologies. The research utilizes 500,000 Facebook posts to develop a model that predicts user behavior based on the number of posts, post count, and sentiment. The study found that image and text data performed better than unpredictability methods, demonstrating the importance of data fusion in predicting user behavior. This could revolutionize online advertising methods and establish the basis for a Decision-Making System (DMS) that includes advertising data analytics and Artificial Intelligence (AI). The research project used a hybrid model to predict user participation in advertisements, while a random model predicted post count, share count, and post sentiment for 60% of each blog post. The models accurately predicted post sentiment, post count, and share count 61%, 62%, and 65% of the time, setting an acceptable standard for future studies.