ESSEM 2013 Emotion and Sentiment in Social and Expressive Media Proceedings of the First International Workshop on Emotion and Sentiment in Social and Expressive Media: approaches and perspectives from AI (ESSEM 2013)

Social and expressive media can represent a challenge and a push forward for research on emotion and sentiment in AI. Although sentiment analysis and emotion detection have been trending topics since a while, not enough emphasis has been placed so far on social and expressive media. The latter, in particular, play a key role in applicative fields related to creativity, its expressions and outcomes, such as figurative arts, music or drama. In such fields, the advent of digital social media has brought about new paradigms of interactions that foster first-person engagement and crowdsourcing content creation: the subjective and expressive dimensions move to the foreground, opening the way to the emergence of an affective component within a dynamic corpus of contents - created or enriched by users. This calls for delving into the evolution of approaches, techniques and tools for modeling and analyzing emotion and sentiment. The workshop aims at bridging between the communities of AI researchers working in the field of affective computing under different perspectives. Such perspectives include, on the one hand, research on models and techniques for sentiment analysis and opinion mining on linguistic corpora and unstructured data from social web; on the other hand, research on formal and cognitive models in intelligent agents and multi-agent systems. The latter, in particular, is concerned with the integration of emotional states into agents and with the role of emotions in agent communication, with the possible goal of defining sophisticated emotion-aware coordination and negotiation strategies. Cross-fertilization between different but related communities will be precious in order to face the challenges raised by the social and expressive media, such as: - investigating advanced social aspects of emotions, i.e. regulative or ethic issues related to emotions in virtual agents; - extracting concept-level sentiment conveyed by social media texts by relying on structured knowledge of affective information, i.e. affective categorization models expressed by ontologies, better still if psychologically motivated and encoded in the semantic web standards; - cross-validation between sentiment-based approaches and cognitive models; - fostering the interoperability and integration of tools by encouraging compliance with emerging standards (e.g., Emotion Markup Language).