Design Choices for Embodied Conversational Agents to Preserve Testimonies by Contemporary Witnesses

Abstract

Conversations with contemporary witnesses, e.g., Holocaust survivors, are valuable opportunities to learn about impacts of historical events. They lead to a more thorough conveyance of history by complementing aggregated facts and numbers with emotional personal experiences and detailed accounts. Encounters and interviews with first-hand witnesses to a number of historical periods soon are a thing of the past due to the respective survivors' advanced age. Embodied Conversational Agents using recordings of contemporary witnesses are an attempt at preserving interactive and personal testimonies for future generations of students and educators. Since the implementation of such Interactive Digital Testimonies (IDT) contains irreversible design decisions, an overview of available choices and their respective consequences are needed. While a number of IDTs have been created over the past years, there is little empirical research on available design choices and the resulting effects on users. In my dissertation I will identify the required HCI features of IDTs, as well as investigate and evaluate how these features can be implemented. This scientific basis will allow for informed decisions during the development of future IDTs.