Page 14 - ISCI’2017

In Fig. 3 (also as in the next figures), the general distribution (distribution graph) is shown

with the help of a solid bold line and a trend line, which is shown with a dotted line. Based on the
2
approximation function and its accuracy of the approximation (R , this accuracy is enough high) we
can say that nowadays, the popularity of cybercrime is decreasing, but this is in line with the fact that
we have some reduction of some offenses in this area. For each point of all set of data in this Fig. 3,

some limits of standard accuracy (of error) are shown, which are determining an area of precision of
th
our measuring (Kavun, 2009). In addition, you can see, in fact, this model is a model of 5 degree,
-3
because a coefficient at the variable X6 is small enough being about 10 . Based on the approximation
2
function and its accuracy of the approximation (R ), as is shown in Fig. 4, we can say that nowadays,
the popularity of computer crime also is decreasing, and besides, this distribution has suffered a

reduction since 2007. Therefore, it is a good tendency acknowledging certain success in the fight
against computer crimes by some governmental and private organizations.

However, we cannot trust to this distribution graph and its forecasting in a full measure
because the accuracy of the approximation is quite low. So, this forecasting can be considered only

in an approximate sense. The distribution graph of the definition “Internet crime” is shown in Fig. 5.

2
Based on the approximation function and its accuracy of the approximation (R ), as is shown
in Fig. 5, we can say the time-frequency distribution for the definition “Internet crime”, in general,

has a common positive linear tendency (based on a common linear trend, as is shown with the long-

dotted line in this figure) during all examined time period, but also with the small precision
approximation (ε < 0.5). This fact doesn’t support fully our forecasting for this area, so we must use

this forecasting with a precaution.
In addition, we could reduce the degree of this polynomial trend line to degree 4 in contrast

to degree 6 without significant reducing of the accuracy of approximation, as it shown below.
4
3
2
6
5
y = -0,0001x + 1,5418x - 7732,4x + 2E+07x - 3E+10x + 2E+13x - 8E+15
R² = 0,4893.

14