Page 175 - Cyber Terrorism and Extremism as Threat to Critical Infrastructure Protection

GRAEME BALLARD: IF THE FACE FITS: IS IT POSSIBLE FOR ARTIFICIAL INTELLIGENCE TO ACCURATELY PREDICT THREATS TO
PROTECT CRITICAL INFRASTRUCTURE?

it still goes without saying that stress, presented at unusual times, could also be an indicator
of an imminent threat in the form of terrorist or other illegal activity.

Whether or not contextual data is considered, Gjoreski et al. (2017) identified a number of
weaknesses surrounding research on arousal and stress when using only biological variables.
• Stress, itself, is highly subjective.
• Due to this subjectivity, it is difficult to define the start, duration and intensity of stress.
• Accuracy of monitoring equipment – at the moment the results appear to be equipment-
specific.
• Constrained environments – have to “calibrate” the equipment to filter “noise”.
• Issue of subjective stress labelling in non-constrained environments.
• In order to explain and, perhaps, overcome these issues, it is important to understand arousal.

3.1 Arousal

There are a couple of important things to note about the nature of arousal. Arousal is not a
simple, linear, process but is subject to two other factors: biological rhythms that are unique to
the individual; and cognitive interpretation, which is also unique to the individual. In general,
when arousal increases, performance increases as well; but this relationship does not continue
indefinitely (Hebb, 1955) as seen in Figure 2. This is because human beings are not linear
creatures.

Figure 2: Arousal

3.2 Solomon’s Opponent Process

Regardless of how and why an individual is aroused, whenever a person experiences an
increase in positive affect – e.g. an emotional state or change in level of arousal – they are likely
to experience an increase in negative affect afterwards. Conversely, when a negative affect is
initially experienced, an individual will tend to experience a positive affect afterwards. This is
because humans are designed in such a way that whenever affect departs from a baseline, an
opponent process is triggered to return the person to that baseline, as seen in Figure 3, below.

175