Engagement Patterns: Exploring the Effect of Article Content on Likes in Social Media

In this study, we explore how articles rich in knowledge affect engagement levels of users on social media. We utilize a unique text mining technique to simplify the analysis of content and to detail interactions such as "likes" on social media platforms. Our approach integrates direct text mining of raw social media data with the semantic capabilities of Word2Vec, and combines this with the analytical strengths of regression analysis and machine learning methods. This combination not only extends the analytical depth of regression analysis but also leverages the predictive abilities of machine learning. Crucially, our results show that our approach effec-tively identifies words that positively or negatively influence user engagement, thereby im-proving the impact of social media posts. By adopting this method, we aim to refine the way so-cial media content is delivered, creating a more direct link between creators and their audience and ensuring that relevant, high-quality content effectively reaches its target.