Emotion, Content, and Context in Sound and Music - Game Sound Technology and Player Interaction: Concepts and Developments - page 256

Table 5. Defined Set of Emotional States (E-states)

E -State Depressed Unhappy Neutral Happy Zoned

Numeric Range 0-3 3-4 4-6 5-8 7-9

results obtained from the Wiimote). Similarly,
for each of the other input parameters we work
with, such as weather conditions, amount of light
and so on, a range of states was also defined. As
previously mentioned, due to our current limita-
tions of time and resources, we focused on only
the locomotive state of the user. It is hoped that
in future, further ratification will be achieved by
using other user measurements such as heart rate
and galvanic skin response.
To avoid additional complication, and allow
greater control over the testing procedure, this set
of parameter ranges and values was loosely defined
according to the empirical and historic knowledge
of the individual. However, this is too fixed and
logical to fit the way things tend to work in the
real world: therefore, to make these values more
representative of real scenarios, they are fuzzified ,
when defined in the fuzzy logic system. In simple
terms, this means the boundaries and degree of
accuracy of each point on a scale is related within
the range of all available values.
The implemented fuzzy logic system provides
a single output value that represents the predicted
E -state of the listener. For simplification, we began
by defining five emotional states and assigned a
numeric range to each state to allow the repre-
sentation of varying degrees of this state and the
overlap where states merge into one another. This
is appropriate, since it is very difficult to place
an absolute, quantitative metric onto the complex
emotions felt by humans. A table representing
these allocations is shown in Table 5.
To verify the effectiveness of using the Wii
controller as a device to measure movement, and
particularly locomotive state, we performed a
number of experiments benchmarked against a

Qualisys motion detection system. By determin-
ing the rate of acceleration from the accelerom-
eters in the Wii controller and asking a subject to
provide three locomotive states (walking, jogging,
and running) we can see that the Wii controller
allows rapid identification of each of these states,
as the graph in Figure 9 shows.
We defined a number of different scenarios,
that combined a range of contextual parameters
which a user might typically find himself or her-
self in. Each of these scenarios is shown in Table
6.

These parameters included those that the user
has control over, in this case the locomotive state,
and external, environmental factors, beyond the
control of the user. We carried out a quantitative
investigation, using 10 subjects, where each sub-
ject was asked to map one of the emotional states
from Table 5 against one of the scenarios from
Table 6. From this investigation, the average E -
state response rating for each scenario is shown
in Table 7. Although the use of an average response
is not ideal, it provides a sufficient insight into
the common perception of each scenario and when
we performed analysis of each response, there
was strong correlation across all of the subjects.
Each of the songs in the mini-database was
allocated a range of values from the E -state table
by the research team. These values reflected the
researchers' perceptions of the content and emo-
tional indicators present in the music. Naturally,
in future research, we will explore the perceived
emotional state attached to each song, by employ-
ing a more detailed sample of a suitable population.
However, for now, these allocations were de-
cided to be a controlled factor in this particular
investigation. These allocations were then mapped