Computational Models of Emotions

 

 

 

Eva Hudlicka

 

Psychometrix Associates

Blacksburg, VA, USA

 

 

 

Emotion plays a critical role in adaptive behavior. Neuroscience and psychology research has demonstrated a close connection between cognition and emotion. Cognition plays a key role in emotion generation via the processes involved in cognitive appraisal.  Emotions in turn exert a broad range of effects on attention, perception and the cognitive processes involved in decision-making, problem-solving and learning.  Computational models of emotion and emotion-cognition interactions provide an important tool for understanding the mechanisms of these processes.  The resulting models also have a broad range of applications in human-computer interaction.

 

This course discusses the motivation and alternatives for developing computational models of emotion and emotion-cognition interactions, within the context of cognitive-affective architectures. The primary focus will be on symbolic models of the core affective processes: cognitive appraisal and emotion effects on cognition. The course will provide background information about emotion research in psychology and neuroscience, that provide the empirical and theoretical basis for model development. An analytical framework will be presented to help organize the variety of existing models.  Approaches to model development, and examples of existing models and architectures, will then be presented, within the context of this framework. The course will conclude with a discussion of validation approaches, and the challenges and future trends in emotion modeling.

 

The afternoon seminar sessions will provide opportunities for discussion of the material covered in the morning lectures, and the broader relevance of emotion models for affective human-computer interaction and affective agents.  Participants will also be able to explore the use of belief nets in modeling cognitive appraisal and emotion effects on cognition.

 

 


 

 

Course outline

 

 

Day 1:  Overview of Emotion Research in Psychology and Neuroscience [download presentation]

 

o       Definitions

 

o       Multi-modal nature of emotions

 

o       Taxonomy of affective states & traits

 

o       Roles of emotion: interpersonal & intrapsychic

 

o       Core processes in emotion: cognitive appraisal and emotion effects on cognition

 

o       Neuroscience theories & data

 

 

 

Required readings

 

Ellsworth, P.C. and Scherer, K. R. (2003). Appraisal Processes in Emotion.
In R. J. Davidson, K.R. Scherer & H.H. Goldsmith (Eds.), Handbook of Affective Sciences.NY: Oxford University Press.

 

Ellsworth, P.C. and Scherer, K. R. (2003). Introduction: Cognitive components of emotion.
In R. J. Davidson, K.R. Scherer & H.H. Goldsmith (Eds.), Handbook of Affective Sciences.NY: Oxford University Press.

 

Frijda, N. H. (1993). Moods, Emotions Episodes, and Emotions. In M. Lewis & J.M. Haviland (Eds.), Handbook of Emotions. New York: The Guilford Press.

 

 

Optional readings

 

Mineka, S., Rafael, E. & Yovel, I. (2003). Cognitive Biases in Emotional Disorders: Information Processing and Social-Cognitive Perspectives. In R. J. Davidson, K.R. Scherer & H.H. Goldsmith (Eds.), Handbook of Affective Sciences.  NY: Oxford University Press.

 

Phelps, E. A. (2006). Emotion and Cognition: Insights from Studies of the Human Amygdala. Annual Review of Psychology, 57:27–53.

 

Rolls, E. T. (2002). A Theory of Emotion, Its Functions, an Its Adaptive Value.

In R. Trappl, P. Petta & S. Payr (Eds.), Emotions in Humans and Artifacts. Cambridge, MA: The MIT Press.

 

Scherer, K. (2003). Introduction: Cognitive Components of Emotion.
In R. J. Davidson, K.R. Scherer, H.H. Goldsmith (Eds.), Handbook of Affective Sciences.  NY: Oxford University Press.

 

 

 

Day 2: Modeling Methods & Data Requirements [download presentation]

 

o       Computational models of cognition: research & applied objectives

 

o       Categories of modeling approaches & model resolution levels

 

o       Representational and reasoning formalisms: symbolic & connectionist

 

o       Cognitive architectures, agent architectures & user models

 

o       Data & knowledge requirements

 

 

Required readings

 

Cooper, R.P.  (2002). Modelling High-Level Cognitive Processes. Mahwah, NJ: Lawrence Erlbaum Associates.  (Chapter 1)

 

Hudlicka, E. (forthcoming).Cognitive Models and Architectures. In Organizational Simulation and Modeling. S. Van Hemel, G. Zacharias, J.McMillan (eds.). Washington, DC: National Academies Press.

 

Optional readings

 

Pew, R. &  Mavor, A. (1998). Modeling Human and Organizational Behavior. DC: National Academies Press.  (Chapter 3) (Although oriented toward military modeling, this chapter provides a comprehensive overview of cognitive architectures developed in the US up to 1998).

 

Ritter, F.E., Shadbolt, N.R., Elliman, D., Young, R., Gobet, F. and Baxter, G.D. (1999).  Techniques for modelling human performance in  synthetic environments: A supplementary review. Technical Report # 62. University of Nottingham, UK: Department of Psychology.


 

 

Day 3:  Modeling emotion I:  Appraisal and Emotion Effects [download presentation]

 

o       Models of cognitive appraisal

 

o       Models of effects of emotions on attention, perception & cognition

 

o       Theories & methods (representational & reasoning formalisms)

 

o       Multiple levels of resolution: black-box vs. process models

 

o       Data & knowledge requirements

 

 

Required readings

 

Gratch, J. & Marsella, S.. (2004). A Domain-independent Framework for Modeling Emotion. Journal of Cognitive Systems Research, 5(4), 269-306.

 

Hudlicka, E. (in press). Reasons for Emotions. In W. Gray (Ed.), Advances in Cognitive Models and Cognitive Architectures. NY: Oxford University Press.

 

Optional readings

 

Busemeyer, J., Dimperio, E., & Jessup, R.K. (in press). Integrating Emotional Processes into Decision Making Models. In W. Gray (Ed.), Advances in Cognitive Models and Cognitive Architectures. NY: Oxford University Press.

 

Fellous, J-M. (2004). From Human Emotions to Robot Emotions. In Proceedings of the AAAI Spring Symposium: Architecture for Modeling Emotion: Cross-Disciplinary Foundations. AAAI Technical Report SS-04-02. Menlo Park, CA: AAAI Press.

 

Roseman, I.J. & Smith, C.A. (2001). Appraisal Theory: Overview, Assumptions, Varieties, Controversies. In K.R. Scherer, A. Schorr, & T. Johnstone (Eds.), Appraisal Processes in Emotion: Theory, Methods, Research. NY: Oxford University Press.

 


 

 

Day 4:  Modeling emotion II:  Cognitive - Affective Architectures & Framework for Model Requirements and Analysis [download presentation]

 

o       Research & applied objectives for cognitive-affective architectures & models

 

o       Framework for model requirements & analysis

 

o       Model development guidelines

 

o       Examples of cognitive-affective architectures

 

o       Emergence of generic architectures

 

 

Required readings

 

Canamero, L. (2002). Designing Emotions for Activity Selection in Autonomous Agents. In R. Trappl, P. Petta & S. Payr (Eds.), Emotions in Humans and Artifacts. Cambridge, MA: The MIT Press.

 

Ortony, A., Norman, D., & Revelle, W. (2005). Affect and Proto-Affect in Effective Functioning. In J-M. Fellous & M. A. Arbib (Eds.),  Who Need Emotions? NY: Oxford University Press.

 

Optional readings

 

Breazeal, C. & Brooks, R. (2005). Robot Emotion: A Functional Perspective. In J-M. Fellous and M. A. Arbib (Eds.),  Who Need Emotions? NY: Oxford University Press.

 

Fellous, J-M. & Arbib, J. (2005). Toward Basic Principles for Emotional Processing: What the Fearful Brain Tells the Robot. In J-M. Fellous and M. A. Arbib (Eds.), Who Needs Emotions? NY: Oxford University Press.

 

Hudlicka, E. (2006). Framework for Modeling Emotion Within Cognitive Architectures. Technical Report 06-08. Blacksburg, VA: Psychometrix Associates.

 

Sloman, A., Chrisley, R., and Scheutz, M. (2005). The Architectural Basis of Affective States and Processes. In J-M. Fellous and M. A. Arbib (Eds.), Who Need Emotions? NY: Oxford University Press.


 

 

Day 5:  Model validation, Challenges, Future Developments

 

o       Approaches to validation of cognitive architecture: shared data sets and benchmark problems

 

o       Extending ‘cognitive’ validation methods to models of emotion

 

o       Pragmatic considerations in model development:

 

o       Data & theory requirements

 

o       Brittleness & model knowledge-base development

 

o       Facilitating model development (standards, shared components & ontologies)

 

 

Required readings

 

Campbell, G. E. a. B., A.E. (2005). HBR Validation: Integrating Lessons Learned From Multiple Academic Disciplines, Applied Communities, and the AMBR Project. In R. W. P. K.A.Gluck (Ed.), Modeling Human Behavior with Integrated Cognitive Architectures: Comparison, Evaluation, and Validation (pp. 365-395). Mahwah, NJ: Lawrence Erlbaum Associates.


Gratch, J., Marsella, S. (2004). Evaluating a Computational Model of Emotion. Journal of Autonomous Agents and Multiagent Systems, Special Issue on the best of AAMAS 2004.


Prendinger, H., Ishizuka, M. (2005). Human Physiology as a Basis for Designing and Evaluating Affective Communication with Life-Like Characters. IEICE Transactions on Information and Systems, E88-D(11), 2453-2460.

 

Optional readings

 

Tenney, Y. J., Diller, D.E., Deutsch, S., Godfrey, K. (2005). The AMBR Experiments: Methodology and Human Benchmark Results. In R. W. P. K.A. Gluck (Ed.), Modeling Human Behavior with Integrated Cognitive Architectures: Comparison, Evaluation, and Validation (pp. 13-44). Mahwah, NJ: Lawrence Erlbaum Associates.


Richard W. Pew, R. W., Gluck, K.A., Deutsch, S. . (2005). Accomplishments, Challenges, and Future Directions for Human Behavior Representation. In R. W. P. K.A. Gluck (Ed.), Modeling human behavior with integrated cognitive architectures: Comparison, evaluation, and validation. (pp. 397-414). Mahwah, NJ: Lawrence Erlbaum Associates.


Kieras, D. E. (2003). Model-based evaluation. In J. Jacko & A. Sears (Eds.), Handbook for human-computer interaction (pp. 1139-1151). Mahwah, NJ: Lawrence Erlbaum.

 

 

 

Assignments

 

Participants taking the course for credit will have the option of either writing a short paper elaborating a selected topic discussed in class, or describing a high-level design of an emotion model for a particular application. The paper or design description should be between 5 – 8 pages.

 

 

 

 

Eva Hudlicka

Psychometrix Associates, USA

 

 

Eva Hudlicka is a Principal Scientist and President of Psychometrix Associates, in Blacksburg, VA, US. Her primary research focus is the development of computational models of emotion; both the cognitive processes involved in appraisal, and the effects of emotions on cognition. Her prior research includes affect-adaptive user interfaces, UI design, decision-support system design, and knowledge elicitation. Dr. Hudlicka received her BSc in Biochemistry from Virginia Tech (1977), her MSc from The Ohio State University in Computer Science (1979), and her PhD in Computer Science from the University of Massachusetts-Amherst (1986). Prior to founding Psychometrix Associates in 1995, she was a Senior Scientist at Bolt Beranek & Newman in Cambridge, MA.