Exploring Composition-based Group User Modeling in an Embedded Interaction System

Candidate: Ying Jiang
Type: Master of Science (MSc), School of Interactive Arts and Technology
Date: July 18, 2008
Senior Supervisor: Ron Wakkary
Thesis: Download Thesis Document

Abstract

Embodied interaction system design faces increasing challenges of design for groups within social context; therefore, group user modeling within an embodied interaction environment becomes a critical area warranting exploration. Most existing group models are derived using quantitative-based approaches, which pose limitations when being applied to groups and embodied interaction. An alternative approach based on compositions and personality classifiers may address these issues, however it remains underexplored. This study discusses the rationale for choosing Keirsey Temperament as a composition classifier, presented design and implementation of our research in which we incorporated both qualitative and quantitative methods with an ecological approach. Results suggest excluding Keirsey Temperament as an effective basis for constructing composition-based group models in an embodied interaction system yet shed light on a future research focus of investigating psychological types for a wider range of dimensions as a possible basis for constructing composition-based group models in embodied interaction system design.

Keywords: group model; embodied interaction; personality type

Subject Terms: HCI; user model; interaction design; artificial intelligence