International Journal of Synthetic Emotions, 3(1), 1-17, January-June 2012 1
Copyright © 2012, IGI Global. Copying or distributing in print or electronic forms without written permission of IGI Global is prohibited.
Keywords: Emotion, Emotional State, Gtrace, Perceived Effect, Tracing Emotion
INTRODUCTION
A decade ago, our group used research on human
emotion to argue that continuous representations
were a natural tool for affective computing
(Cowie et al., 2001). The idea has worn well,
and there is growing interest in the problem of
using signs in various modalities to construct
a ‘trace’ that describes how a target person’s
emotional state appears to be rising and falling
from moment to moment. However, problems at
the human end have been perhaps unexpectedly
obstinate. Progress depends on databases that
provide traces reflecting human judgments, so
that machines can use them as models. Acquir-
ing suitable traces poses problems which are
both substantial and interesting.
The aim of this paper is to provide an
overview of the issues involved in acquiring
traces from human raters, and the techniques
available to solve them. Among those is a new
system for acquiring traces, Gtrace, designed to
capitalise on the lessons described here. Some
of the issues are well known in psychology,
and are dealt with in sources such as Rosenthal
(2005). Others are specific to the particular
technique, and have not been dealt with in a
systematic way before.
Tracing Emotion:
An Overview
Roddy Cowie, Queen’s University Belfast, UK
Gary McKeown, Queen’s University Belfast, UK
Ellen Douglas-Cowie, Queen’s University Belfast, UK
ABSTRACT
Computational research with continuous representations depends on obtaining continuous representations
from human labellers. The main method used for that purpose is tracing. Tracing raises a range of chal-
lenging issues, both psychological and statistical. Naive assumptions about these issues are easy to make,
and can lead to inappropriate requirements and uses. The natural function of traces is to capture perceived
affect, and as such they belong in long traditions of research on both perception and emotion. Experiments
on several types of material provide information about their characteristics, particularly the ratings on which
people tend to agree. Disagreement is not necessarily a problem in the technique. It may correctly show that
people’s impressions of emotion diverge more than commonly thought. A new system, Gtrace, is designed to
let rating studies capitalise on a decade of experience and address the research questions that are opened
up by the data now available.
DOI: 10.4018/jse.2012010101