July 10, 2016
New York
Over the past decade, work in computational linguistics has driven text mining into the realm of predicting sentiment, opinions and even attitudes and happiness; it has become, functionally, like a barometer of thought. However, the link between the statistical analysis of the text and what appear to refer to psychological states in humans has yet to be made. For example, sentiment analysis correlates units of text with dictionaries that are rarely grounded in psychological theory. Simply put, sentiment analysis and kin are models of the statistical relations between psychological states and textual features, not of the psychological states in and of themselves. This deviates significantly from history in two important ways. First, computational linguistics originally focused much more on modeling psychological processes related to language production, comprehension and development. Second, sentiment, opinions, and attitudes have been studied in extreme depth by psychologists for several decades, in fact, even before computation played a central role in science whatsoever. By bringing together researchers from different disciplines, this workshop will: 1) Address the similarities and differences in how attitudes and sentiment are defined in computational linguistics and psychology; 2) gauge the degree of cross-fertilization across these two fields; 3) explore new and emerging empirical approaches for validation of core assumptions of attitudes and sentiment; 4) discuss emerging applications. In the end, this workshop will provide a clearer understanding of how to map a problem domain to the appropriate technical approach.
Sentiment Analysis
Computational Models of Attitude
Computational Linguistics (particularly around attitudes and sentiment).
Psychology of Attitudes
Affective Computation
Cognitive Neuroscience of Emotion
We encourage submissions from a wide variety of disciplines, including:
Computational linguistics
Computer science.
Psychology
Cognitive Neuroscience
Communications
Sociology