Hybrid Machine Learning for Analyzing Social Media Marketing Influence on Tourists’ Revisit Intentions

Every aspect of human contact, including the way tourists behave, has been profoundly affected by the rise of social media. The proliferation of travel-related content on social media has changed the way people learn about, enjoy, and return to popular tourist spots. With its interactive visual and textual content, social media marketing (SMM) is a powerful tool for influencing tourists' desire to return. In this study, we use sentiment analysis on social media data to look at how SMM affects tourists' tendencies to return. They used feature extraction techniques such POS tagging, BOW, and HASS. DAE and CNN were combined in a new hybrid model called DAECN. To improve feature learning, the model also included fully linked and pooling layers. Outperforming five benchmark models, the model showed remarkable adaptability in recognising behavioural patterns from social media inputs and achieved an amazing training accuracy of 97.42%. The dataset included issues with emoticons, URLs, and HTML tags, yet the model still performed well. These results demonstrate that SMM is useful for forecasting and shaping tourists' intentions to return, and they highlight the importance of deep learning for comprehending digitally driven tourism trends.