Emotions Inference Through Content and Sentimental Analysis in COVID- 19 Context

Abstract

Emotions are part of a person's behavior, and some feelings can affect their actions, they can even prevent a person from producing an intelligent result. Interest in researching all aspect of the relationship between humans and computers has been increasing during the last years. Emotion is a fundamental component of being human; enjoying, hating, disgusting, among a lot of other emotions, help to value the entire human experience. Nowadays, it is recognized that the emotions play a critical and important role in all relationships with the computer, from playing games, shopping online to reading some news in a digital newspaper. Human computing interaction is an area that involves elements from different disciplines of psychology, anthropology, computer science, artificial intelligence, among others, for the study of emotions. In this paper, we took different digital Chilean newspaper to analyze if the designs of the texts influence the emotions contained in the comments of the news about COVID-19. We combine content and sentiment analysis to help an effective recognition of feelings, including ironies and hybrid language. Interesting results were obtained to identify and understand the models associated with the design of texts and the reactions of readers.