Computational Methods for Affect Detection from Natural Language

Research in human affect has a long established tradition in social sciences - Philosophy, Psychology, Socio-psychology, Cognitive Science, Pragmatics, Marketing, Communication. The study of affect from a computational point of view is a recent field in Artificial Intelligence, denominated Affective Computing. Despite the novelty of the subject, the volume and importance of research in automatic human affect recognition, classification and simulation has been constantly growing in the past decades, leading to the development of further sub-areas of research. One of these directions deals with the study of automatic affect treatment from text, in the Artificial Intelligence area of Natural Language Processing. In this context, different tasks have been developed, from emotion detection, subjectivity analysis, opinion mining to sentiment analysis and appraisal analysis. The aim of the present book is to give a theoretical and practical overview of the most important topics of research dealing with the computational treatment of affect from text. Its goal is to present the reader with the most relevant concepts related to this field (affect, sentiment, subjectivity and others), the existing research areas and familiarize them with the present approaches, with theoretical explanations and practical examples.