Computational Sentiment Analysis

Abstract

Abstract This chapter reviews the state of the field of automated sentiment analysis and draws recommendations for available methods and best practices for communication scholars. Introducing and explaining novel methods, from simple dictionary-based approaches to more sophisticated machine learning algorithms, the authors examine the past, present, and potential future of automated sentiment analysis and critically discuss its benefits and limitations, with a particular focus on issues of validity. Acknowledging these important caveats, they discuss the vast potential of big data and computational methods to advance knowledge and theories of emotional communications and demonstrate its applicability using real-world studies of sentiment analysis in communication research.