Text Analytics For Prediction Of News Quality
This research addresses a problem that has perplexed researchers for many years: how can we predict the perceived news quality? Understanding the root factors that may influence the quality of news is critical for both business researchers, and linguistics and educators. Extant literature has studied various aspects of news, from its readability, rhetorical structure theory to vocabulary burden. Empirical evidence has been limited though. A few recent research attempts to empirically analyze the complex relationship between news quality and the underlying determinants. As a start, however, existing studies relied on the linear family of regression models. This leads to, we believe, significant controversial results among them. We advance this line of research by combining readability and rhetorical features to produce a highly predictive model of human readers’ judgments of news quality. This model is based on a supervised machine learning algorithm named C4.5 decision tree, thus allowing a non-linear functional form of determining news quality.
Keywords - Perceived News Quality, Digital Journalism, Machine Learning, Decision Tree.