ss

Keynote Speakers

Prof. Cristiano Castelfranchi, CNR, Italy

Prof. William Gaver, Goldsmiths College, UK

Prof. Ursula Hess, University of Quebec - Montreal, Canada

Prof. Klaus Scherer, University of Geneva, Swiss

 

Opening Session:

Prof. Rosalind Picard, MIT Media Lab, USA

 

Keynote Speeches Details

 

Prof. Cristiano Castelfranchi

Title: The Mental Ground of (Complex) Emotions and its Impact on Affective Interaction

Abstract: Reacting to stimuli, perceiving our bodily reaction to events, to feel something, is not enough for human complex emotions (perceiving that I'm escaping, is not enough even for 'fear'). Complex human emotions are based on specific mental states: typical configurations of beliefs, goals, motives, expectations, ...  Our body does not respond just to 'external' stimuli (events) based on some pattern matching; it reacts to our 'interpretation' of the stimulus, to the 'meaning' of the event; that is to its mental representation. It also reacts to merely endogenous representations, to mental events (like a counterfactual imagination). We feel our bodily response, and we ascribe it to that event or idea; this combination gives an emotional nature to both sides. We 'interpret' perceptual stimuli about our body and its states and reactions too; not only perceptual stimuli about the external world.
I will analyze the typical mental configurations needed for:

  1. rather simple 'anticipation-based' emotions ('hope', 'fear', 'disappointment', 'relief', 'exultance');
  2. complex social emotions like 'shame', 'envy', 'guilt', 'pity': their ingredients and their coherent structure.

Only for 'shame' I will also analyze the expressive component: body, face, blushing; and its meaning, very coherent with its mental stuff. I will also discuss the social function of shame signals.
On such a base I will claim:

  1. That in order to really model affective architectures we have to model the 'body', that is, its signals about internal states and reactions, and the two way link with cognition: interpreting and eliciting and interpreting bodily reactions.
  2. That although atomically decomposable those complex mental states have their own unitary nature, their emergent, specific, non-reducible properties and functions.
  3. That emotional interaction (Ag-Ag; H-Ag; H-robot; etc.) cannot be based only on the recognition of the expressive signals. Appropriate emotional interactions are based on the recognition of the mental stuff of the other agent: of her beliefs, suppositions, motives, expectations, ... We react to this, not just to an expressive 'face'.

Biography: Cristiano Castelfranchi is a Professor of General Psichology(Cognitive Science) at the University of Siena (Department of Communication Sciences) and Director of the Institute of Cognitive Sciences and Technologies of Italian National Research Council (ISTC-CNR) in Roma. Castelfranchi is author of several books in both italian and english (among them "Emotions and social image" and "Cognitive and Social Action") and has published more than 200 papers.
He is member of the editorial board of the international journal "Autonomous Agents and Multi-Agent Systems" and member of the editorial board of "Cognitive Science Quarterly"

 

Prof. William Gaver

Title: Avoiding Affective Automation

Abstract: The idea of computers that know about and react to people's emotions is as disturbing as it is seductive. It raises discomfiting visions of insincere electronic camaraderie, privacy invasions, and the automation of affect - and that's if the systems work. But we may gain the benefits of computational assessments of emotion, even those that are only approximate, if we maintain people's sense-making at the core of system design. In this talk, I describe the Home Health Horoscope, a system that reports sensor-based judgements of wellbeing in a form that purposely undermines their authority. This approach allows the system to raise emotions as an issue without imposing a normative model on people's interpretations, providing many of the potential advantages of affective interaction while avoiding the perils.

Biography: William Gaver is a Professor of Design at Goldsmiths, University of London, where he heads the Interaction Research Studio. His work focuses on designing technology for everyday life, and investigates the entire design cycle from inspirational research to new forms of assessment.

 

Prof. Ursula Hess

Title: Faces and Emotions: Why the same expression may not mean the same when shown on different faces

Abstract: In general, we assume that a certain expressive behavior will be perceived as signaling the same internal state regardless of who shows it.  The present talk will explore how the same expression when shown by different encoders or seen by different decoder can acquire different meaning. Two series of studies will be presented, one focusing on the effect of expresser gender and one on decoder culture.

Biography: Ursula Hess is a Professor of Psichology at University of Quebec, MontReal. Her work focuses mainly on Nonverbal communication, Emotion communication and Psychophysiology. She is currently Associate Editor of the Psychophysiology Journal.

 

Prof. Klaus Scherer

Title: Emotion in Social Interaction: Functions and Dysfunctions

Abstract: The computational modeling of emotion has made major progress, partly due to the adoption of cognitive theories of emotion that suggest appraisal as the major mechanism for the elicitation and differentiation of emotion. Given the focus of these theories on mechanisms internal to an individual, the social factors in emotional processes and the role of emotion in social interaction are rarely highlighted in this tradition. This contribution will attempt to remedy this oversight. From the point of view of componential process models, three pertinent social facets will be explored: 1) Social and cultural determinants of appraisal, 2) the role of emotional expression in the negotiation of interpersonal reactions and behavioral intentions in social interaction, and 3) emotional resonance (empathy) and emotional contagion. In each case, functions (e.g., smoothing interaction) and dysfunctions (e.g., escalation) will be discussed. In passing, the relevance of these approaches to human-agent and agent-agent interactions will be commented upon.

Biography: Klaus Scherer is a full professor of psychology at the University of Ge­neva, Switzerland, and director of the Human Assessment Centre (Laboratoire d´Evaluation Psychologique). His teaching and research activities focus on the areas of emotion, stress, motivation, personality, and organisational behavior. Klaus Scherer is a member of several in­ternational scientific societies and a fellow of the American Psychological Association and the Acoustical Society of America. Most recently, Klaus Scherer has become the Director of the Swiss National Center of Competence in Research for the Affective Sciences, based on long-term funding by the Swiss government and the National Science Foundation.