Clicks, Retweets, and Replies: A Data-Driven Exploration of what Makes News Go Viral

The ever-growing volume of online news articles necessitates efficient methods to understand audience preferences. In this study Within pre-existing news article categories, the application of machine learning algorithms for forecasting news item popularity is analysed. K-Nearest Neighbors, Random Forest, and Support Vector Machine were used to categorize articles according to their level of popularity. The most pertinent features for prediction within each cluster were found using feature selection techniques. Based on the findings, all three algorithms produced results with encouraging accuracy, with K Nearest Neighbor and Support Vector Machine showing the best overall performance. For some clusters, popularity prediction is still difficult, indicating the need for more research into those groups’ characteristics. This research demonstrates how machine learning could be used to forecast news popularity within article clusters. News organizations can better engage their audience by customizing content creation and dissemination strategies based on their understanding of what makes articles popular within particular categories.