Adapting the Interaction Style to Affective Factors

in conjunction with User Modeling 2005

Edinburgh, UK
July 25, 2005

Call For Papers

List of Contributions


Sandra Carberry
University of Delaware, US

Fiorella de Rosis
University of Bari, Italy

Program Committee

Elisabeth André
University of Augsburg
Anton Batliner
Uiversity of Erlangen-Nuerberg
Cristina Conati
University of British Columbia
Eva Hudlicka
Antony Jameson
Kia Hook
IT-University in Kista
Diane Litman
University of Pittsburgh
Jon Oberlander
University of Edinburgh
Ana Paiva
Paolo Petta
University of Vienna
Rosalind Picard
Isabella Poggi
Università di Roma Tre

Important dates

Deadline for paper submissions:
March 7, 2005
Notification of acceptance:
April 7, 2005
Deadline for camera-ready version:
May31, 2005
Workshop date:
July 22, 2005


All contributions will be made available on this Web site and published as part of an informal Annex to the main UM05 conference proceedings 


Participants will need to register for the main conference (UM03).
Please consult the main conference website for details regarding registration and accomodation.

Background and Motivation:

Workshops on Adaptation to Affective Factors have become almost a tradition at UM Conferences, with the last three editions (in 1999, 2001 and 2003). The domain has matured considerably in the recent years: several Workshops have been organized on this topic, in Europe, in the US and in Asia, and a European Network of Excellence on Emotions, with a core Consortium of about 40 institutions (HUMAINE), became operational in 2004. The 2005 edition may add, to the audience of those who participated in the previous editions, new potential contributors from this Network.

The goal of this Workshop is to promote comparison of experiences and discussion on the specific problems which arise in extending user modeling and adaptation of interaction to affective factors. The spectrum of factors which may be considered as belonging to the category of 'affect' is wide. It spans from 'long-term' personality factors to medium and short term 'moods and emotions', to other aspects which are usually classified under generic terms such as 'attitudes' (doubt, politeness, deception and others). While psychologists theorized and investigated experimentally the domain of 'strong' emotions (fear/hope, love/hate, disgust, to make a few typical examples), it is likely that the factors which mostly influence interaction with technology are softer (again to make a few examples, frustration, concern, cognitive load, satisfaction, doubt) or of a more generic kind (positive vs negative valence, for instance). This opens the problem of reflecting on these theories and these experimental results, to discuss to what extent they apply to softer affective categories. Recognition of the user state may be based on the observation of various signs: speech, language, facial expressions, movements, physiological parameters and their combination. In the majority of cases, context is fundamental to increase the recognition accuracy. To adapt appropriately to such states, however, recognition should be integrated with interpretation: that is, with the ability to hypothesize the causes which likely produced that state. Recognition should therefore be tightly integrated with user modeling, to insure a consistent representation of affective and non affective aspects of the user. Although adaptation to affective factors may apply, in principle, the same methods which were applied successfully to other kinds of factors, some specificity exists. First of all, a decision must be taken on when interaction should aim at reinforcing or weakening this state. Then, the potential impact of interaction factors on the affective state of the user should be investigated and applied appropriately.

This Workshop aims at fostering discussion on the afore mentioned topics and on others that will emerge as of interest in paper submissions. Researchers will be encouraged to submit not only descriptions of research which produced positive results, but also frank reports about studies and systems that did not work as they were expected to: such reports should include enough detail about the reasons for failure, to help others to avoid pursuing the same dead ends. As usual in this domain, integration of competences and backgrounds will be favored, although user modeling and adaptation will be the core on which the various positions will confront.

Suggested Topics:

  • Opportunities, advantages and disadvantages of adapting interaction to affective factors
  • Recognition of the affective state of the user: through speech, language analysis, facial expression, bio-physiological data or through a combination of media
  • Integration of recognition with triggers and causes of emotion
  • Interpretation of recognized affective states and user model updating
  • Integration of affective and non affective aspects in user models
  • Methods for measuring the impact of affective factors on the user
  • Adaptation of interaction strategies to the presumed affective state of the user, with focus on non-extreme and more subtle emotional states
  • Manifestation of empathic reaction in natural language, graphical and other types of interfaces
  • Theories about the relationship between language and affect, between reasoning and affect, between decision-making and affect
  • Methods for evaluating the adequacy of affective user models
  • Methods for evaluating the utility of adaptation to affective factors
  • Affective factors in interactions with mobile technologies
  • Adaptation to affective factors: in intelligent tutoring systems, call centers, medical applications, advice-giving systems and other domains.

    Workshop Format:

    This full-day Workshop will be opened by an Invited Speech and will include a long paper Session and a Poster and Demo Session. It will be concluded with a presentation of the European Network of Excellence on Emotions (HUMAINE), which will provide a European perspective on important issues of this domain relevant to academic and industrial research.

    The workshop organizers welcome submissions of high-quality papers describing completed or on-going research addressing theories, methods, techniques and results in user emotion and attitude modeling. Both papers and posters may be submitted. Submission of interactive demonstrations is particularly encouraged.

    Submission instructions

    Submissions may be sent in one of the following forms:

    1. long abstracts of 1000 words maximum. Authors of accepted papers will then provide full papers for inclusion in the workshop proceedings. Long abstracts should use a font no smaller than 10 point and must include the following:
    A. A brief overview of the work being reported or the issues being addressed
    B. The current status of any experiments or implementations
    C. A detailed enumeration of what will be included in the full paper.
    D. A specification of whether the authors intend to extend the long abstract into a full paper, a short paper and poster presentation, or a short paper and demo.

    2. If authors wish to submit complete papers rather than long abstracts, they may do so. Full papers should have a maximum length of 4000 words for long papers and 1500 words for poster and demo submissions, including figures and tables. Authors should use the APA style of citations, two-column format and a font no smaller than 10 points.

    Submissions should be pdf files and should be sent to both of the following by email: and