Ranking News Articles Based on Popularity Prediction

News articles are a captivating type of online content that capture a significant amount of Internet users´ interest. They are particularly consumed by mobile users and extremely diffused through online social platforms. As a result, there is an increased interest in promptly identifying the articles that will receive a significant amount of user attention. This task falls under the broad scope of content popularity prediction and has direct implications in various contexts such as caching strategies or online advertisement policies. In this paper we address the problem of predicting the popularity of news articles based on user comments. We formulate the prediction task into a ranking problem where the goal is not to infer the precise attention that a content will receive but to accurately rank articles based on their predicted popularity. To this end, we analyze the ranking performance of three prediction models using a dataset of articles covering a four-year period and published by 20minutes.fr, an important French online news platform. Our results indicate that prediction methods improve the ranking performance and we observed that for our dataset a simple linear prediction method outperforms more dedicated prediction methods.