Date of Award

2017-01-01

Degree Name

Master of Science

Department

Computer Science

Advisor(s)

David G. Novick

Abstract

Embodied conversational agents are changing the way humans interact with technology. Agents are expected to behave like people, both verbally and non-verbally. Researchers within the Human-Computer Interaction community have found many attributes of an agentâ??s gesture contribute to an agentâ??s perceived personality and believability.

Gesture amplitude and frequency are two of the attributes that contribute to the naturalness of a gesture. Currently, no previous studies have defined an empirical baseline for generating natural gestures for these two attributes. I seek to discover whether gesture amplitude and frequency affect how users perceive an agentâ??s naturalness, and which gestures would be preferred for an agent based on its personality.

In particular, I seek to quantify gesture amplitude and to compare it to a previous study on the perception of an agentâ??s naturalness of its gestures. In this Thesis, non-verbal gesturing consists of movement of the arms and their location in space. I evaluate whether agents should use specific gestures more frequently than others depending on the personality type they have been designed with. My study is organized into three experiments and evaluations. My results show some indication that introverts and extraverts judge the agentâ??s naturalness similarly. The larger the amplitude the agent used, the more natural its gestures were perceived. The frequency of gestures between extraverts and introverts seem to show hardly any difference, even in terms of types of gesture used.

Language

en

Provenance

Received from ProQuest

File Size

50 pages

File Format

application/pdf

Rights Holder

Alex Michael Rayon

Share

COinS