In recent years increasing attention has been given to making interactions with Intelligent Virtual Agents (IVAs) more human-like, for example by incorporating social competencies such as expressive speech and body gestures, incorporating backchannels and more realistic turn-taking into dialogue systems. Although such competencies have been seen as important in improving the human-like nature of IVA interaction, little is known about how these aspects contribute to the level of engagement between human and virtual agent or how such a concept can be evaluated in these contexts. This workshop explores the role of these social competencies in engaging interlocutors in successful social communication and how to evaluate user engagement in such interactions.
Engagement refers to the quality of the user experience that emphasizes the positive aspects of the interaction. Numerous methods have been proposed to measure engagement, but further work is needed to validate and relate these measures and so provide a firm basis for assessing the quality of the user experience in a virtual agent context. By exploring the potential of subjective (questionnaire, user perception) and objective (behavioral and multimodal feature extraction for engagement classification) engagement measures, the ultimate aim of this workshop is to contribute to the discussion on how user engagement can be studied, measured, and explained, to facilitate the development of user interfaces able to engage users in social communication.
Particular topics of interest include but not limited to:
- Analysis of social aspects of speech communication relevant for engagement.
- Development of technology for Social Competent Intelligent Virtual Agents.
- Evaluation of applications of IVAs which expect to engage in the interaction and provide a more immersive communication similar to that found in communication between humans, such as companion agents for health care, animated characters in video games, etc.
- Experimental methodologies that are useful for evaluation of engagement.
- Inference of engagement from semantic analysis of the human speech, biometric data captured using sensors and other multimodal data such as gestures.
- Social signal processing and computer vision techniques to capture and interpret social behaviour such as posture, gesture and facial signals.
- Machine learning methods for classification of engagement.
Papers should be submitted by email to email@example.com and should not be longer than 10 pages. It is also possible to submit short-papers for extended abstracts or demos that should not be longer than 4 pages. For submitting short and long papers please use the Springer templates available at IVA 2015.
Please note that papers will be published in the proceedings of the ESIVA workshop which is not planned to be included in the IVA proceedings. The organisers of ESIVA are also responsible for the review process of this workshop which runs independently of the reviewing process of the IVA conference.
Acceptance notification: 1st July 2015
Workshop: 25th August 2015Contact:
For any questions regarding the ESIVA workshop please contact the organizers through firstname.lastname@example.org
New: Download the program and proceedings here
| Technical Committee:
Emer Gilmartin (Trinity College Dublin, Ireland)
Fabio Tesser (Institute of Cognitive Sciences and Technologies, Italy)
Hiroshi Shimodaira (University of Edinburgh, UK)
Ingmar Steiner (Saarland University/DFKI, Germany)
Isabel Trancoso (INESC-ID/Instituto Superior Técnico, Portugal)
Jens Edlund (Royal Institute of Technology, Sweden)
Korin Richmond (University of Edinburgh, UK)
Päivi Kristiina Jokinen (University of Tartu, Estonia)
Patrizia Paggio (University of Copenhagen, Denmark)
Piero Cosi (Institute of Cognitive Sciences and Technologies, Italy)Chairs:
Loredana Cerrato (Trinity College Dublin, Ireland)
João Cabral (Trinity College Dublin, Ireland)
Nick Campbell (Trinity College Dublin, Ireland)
Anna Esposito (Salerno University and IIASS, Italy)
Benjamin Cowan (University College Dublin, Ireland)
Kristiina Jokinen (University of Helsinki, Finland)