Page 21 - ISCI’2017

The coefficient of importance of the criterion takes into account and gives the normal selection

the risk zone Z. According to the results of the first stage filtering, each employee can be traced within
the matrix PDKD (Table 3), including the employees belonging to a risk zone (Z = 55% it’s the

threshold identified by the authors with the help of expert approach). This is an additional rule for
MCRIM-method.

The parameter of dynamics d1 has determined some kind of accounting (by weeks, by months,
by quarter, etc.) with a possibility of detailing d2 (every day, every week, every month, etc.), then dr

= {1, 7, 14, 31, 52, 93, 186, 356}.

It should also be noted that the job categories DK is recognized on the basis of simultaneity,
then the matrix distribution of the parameter dynamics dr can be built (or created) based on a formula.

The choice of the decision maker (DM) from the table of the parameter dr is a very important decision
that will allow one to use a parameter in a decision support system (DSS).

Table 3 – Matrix of dynamic accounting of criteria’s, PDKD

Month January December

(day),
dr

№ ΣUj
Job 1 2 … 31 … 1 2 … 31
catego-

ries,
dkj

1 dk1 2 1 … 4 0 5 … 1 17

2 dk2 0 1 … 0 1 2 … 0 5
… ……. …. .. … .. .. .. … .. ..

n dkn 0 0 … 0 1 1 … 0 2

For example, the following systems can be considered as DSS: Emperor, ConceptDraw

MINDMAP to identify insiders in the company. If the parameter dr = (1÷365), then DM will get the
most accurate solutions that will allow one at this time to identify insiders in an organization.

However, it also will lead to the increase of used resources. Therefore, one need to choose a kind of

“gold selection”.
In the third stage of filtration (method MCRIM) taking into account dynamics, estimates can be

obtained that are more accurate. Therefore, the authors suggest the use of a cartographic analysis by which

21