*4.2.5 Rule generation versus time instant and time interval semantic networks*

Before, is getting started a set of appropriate data should be collected. Data, which create content of {**[Outbpf (i, (nvyrgood(i, j3)), (nvyrrepair(i, j3)), (nvyrwaste (i, j3) ]}** seem to be most significant, however the data contained within **{[Vyk\_zam (i, a21a (i) , a21b (i), a31 (i), a41(i))]}** linguistic sets plays a role of principle importance as well.

### *4.2.5.1 Investigated BP performance – determination of measured unit number*

When providing evaluation of investigated BP performance, a number of measured units is represented by an appropriate integer value **nvyrrepair (i, j3)), (nvyrwaste (i, j3))]}.** However, that type of representation usually does not correspond to representation needs within semantic network. As a result of that, we introduce so called ratio value xvyrgood(i), xvyrrepair(i) a xvyrwaste(i), which might be calculated with respect to formulas (28), (29) and (30)

$$\begin{array}{c} \texttt{xvypood}(\mathbf{i}) = \texttt{nvypood}(\mathbf{i}), \mathbf{j3}) / (\texttt{nvypood}(\mathbf{i}, \mathbf{j3})) + \texttt{nvyrrepair}(\mathbf{i}, \mathbf{j3}) \\ \quad + \texttt{nvyrreate} \ (\mathbf{i}, \mathbf{j3})) \end{array} \tag{28}$$
 
$$\begin{array}{c} \texttt{xvyrrepair}(\mathbf{i}) = \texttt{nvyrrepair}(\mathbf{i}, \mathbf{j3}) / (\texttt{nvyrpool}(\mathbf{i}, \mathbf{j3})) + \texttt{nvyrrepair}(\mathbf{i}, \mathbf{j3}) \\ \quad + \texttt{nvyraste} \ (\mathbf{i}, \mathbf{j3}) \end{array} \tag{29}$$

*Business Process versus Human Resources Performance DOI: http://dx.doi.org/10.5772/intechopen.98944*

$$\begin{array}{l} \texttt{xvyrwaste(i)} = \texttt{nvyrwaste(i,j3)} / (\texttt{nvyrgood(i,j3)}) + \texttt{nvyrrepair}(\texttt{i,j3})\\ + \texttt{nvyrwaste(i,j3)} \end{array} \tag{30}$$

The results are represented by decimal numbers, as a rule.

#### *4.2.6 Equations, which enable rule generation*

The relationship between investigated BP performance and performance of employee interested in BP functionality might be quantified via formulas (4.20a and 4.20b)

$$\{ [\mathbf{a15(i, j2)}] \} \approx \{ [\mathbf{a14 (i, j2)}] \} \tag{31}$$

$$\{ [\mathbf{a15}(\mathbf{i}, \mathbf{j2})] \} = \{ [\mathbf{k54} \ (\mathbf{i}, \mathbf{j2})] \} \otimes \{ [\mathbf{a14} \ (\mathbf{i}, \mathbf{j2})] \} \tag{32}$$

where

**{**[a15(i, j2)]**}** is a linguistic set, which contains data concerned with investigated BP performance

**{**[a14 (i, j2)]**}** is a linguistic set, which contains data concerned with performance of employee interested in BP functionality

**{**[k54 (i, j2)]**}** is a linguistic set, which contains data concerned with the abovementioned relation dynamics

Furthermore, we assume that the performance of employee interested in investigated BP performance is given by measure of employee content and motivation, while formula (33) might be postulated

$$\{ [\mathbf{a}\mathbf{1}\mathbf{1}(\mathbf{i}, \mathbf{1})] \} = \{ [\mathbf{a}\mathbf{1}\mathbf{1}(\mathbf{i}, \mathbf{j}\mathbf{1})], [\mathbf{a}\mathbf{1}\mathbf{1}(\mathbf{i}, \mathbf{j}\mathbf{4})] \}\tag{33}$$

Where

[a11(i, j1)] – is a measure of employee content and

[a11(i, j4)] –is a measure of employee motivation

An outgoing point for rule generation is the **time interval semantic network,** which contains appropriate attributes together with their adequate statistic values.

With respect to inputs postulated within time instant and time interval semantic network two types of rules might be generated: (a) the basic rule concerned with time instant semantic network (see also Rule no.1) and (b) the basic rule concerned with time interval semantic network (see also Rule no. 2).

**Rule no. 1** - the basic rule concerned with time instant semantic network

IF Date1 i, k ð Þ and Time1 i, k ð Þ¼ Date2 i, k ð Þ and Time2 i, k ð Þ

$$\begin{array}{l} \text{and } [\mathtt{a}\mathtt{21}(\mathtt{i},\mathtt{k},\mathtt{Date}\mathtt{2},\mathtt{Time}\mathtt{2})] \in [\mathtt{a}\mathtt{21a}(\mathtt{i})] \text{ THEN } (\mathtt{Typ\\_vystupu}(\mathtt{i},\mathtt{k})) \\\ = (\mathtt{Poci\\_m.j}(\mathtt{i},\mathtt{k})) \in \mathtt{Vyk\\_skum\\_BP} \end{array}$$

Where

k = 1....m1 is the ordinal number

i=1 … n is the ordinal number of investigated BP within set of other business processes

Date1 (i, k) – date of made measurement Time1 (i, k) - time of made measurement Output type (i, k) – output type – good, repaired, waste M. u. number (i, k) – number of units related to actual output [a21a (i)] - linguistic set.

**Rule no.2** - the basic rule concerned with time interval semantic network IF Date1 (i, k1a) and Time1 (i, k1a) = Date2 (i, k1a) and Time2 Avg(a21(i, k1a)) ∈ [a21a (i)] THEN (Typ\_vystupu (i, k1a)) = Avg ((Počet\_m.j (i, k1a))] ∈ Vyk\_skum\_BP

IF Date1 (i, k1a) and Time1 (i, k1a) = Date2 (i, k1a) and Time2 Var\_rozp(a21(i, k1a)) ∈ [a21a (i)] THEN (Typ\_vystupu (i, k1a)) = Var\_rozp ((Počet\_m.j (i, k1a))]∈ Vyk\_skum\_BP

k1a = k2-k1 - k1 – the beginning of measurement time interval t1, t2 k2 – the end of measurement time interval t1, t2 [a21(i, k1a, Date2, Time2,Avg ((M.u. number (i, k1a))]

Vyk\_skum\_BP – the semantic set, which contains data closely related to performance of employee interested in investigated BP functionality

However, further rules might be derived based on Rule no.1 and Rule no. 2 as well.

## **4.3 Investigated BP performance versus performance of employee interested in that BP functionality: The objective oriented expert system - ES\_BP\_HRP**

The previous sub-section concerned that chapter deal with theoretical aspects related to investigated BP performance and performance of employee interested in that BP functionality and the result is a conceptual model related to the abovementioned objectives. However, this section deals with implementation of the conceptual model and a result should have a shape of the objective oriented expert system, which should enable to find a response to the question "How individual psychological factors affect the investigated BP performance and a performance of employees interested in that BP functionality" as well. On the other hand, that expert system plays a role of knowledge-based support tool for the abovementioned investigated business process too.

#### *4.3.1 The ES\_BP\_HRP\_02 system: structure and functionality*

The ES\_BP\_HRP\_02 system structure consists of two subsystems typical for any knowledge-based or expert systems denoted as ES\_BP\_HRP\_02\_01Knowledge Base and ES\_BP\_HRP\_02\_02 Inference Engine. When considering the ES\_BP\_HRP\_02\_01Knowledge Base, the knowledge stored there are represented based on appropriate semantic networks (SNWs) and reference databases (RDBs), while the RDBs contain data and information, based on which adequate SNWs are being generated.

### *4.3.2 The ES\_BP\_HRP\_02 expert system knowledge base: RDBs component*

The data stored within reference databases have the {[ES\_BP\_HRP\_02\_01\_02\_02 Výstupy\_Outputbpf (i, j)] linguistic set nature, which consist of three subordinated linguistic sets [ES\_BP\_HRP\_02\_01\_02\_01\_01 Output\_OK (i, j)]<sup>8</sup> , [ES\_BP\_HRP\_02\_01\_02\_01\_02 Output\_REPAIR (i, j)]9 and

<sup>8</sup> Those subset elements are concerned with those output products, which meet the pre-defined criteria in a full range and are denoted as – good output products.

<sup>9</sup> Those subset elements are concerned with those output products, which meet the pre-defined criteria in a full range, however they might be repaired so that, they can be good output products and they are denoted as the repaired output products.

[ES\_BP\_HRP\_02\_01\_02\_01\_03 Output\_WASTE (i, j)],10 while a content of those linguistic sets is described in section denoted as Generation of rules – data -acquisition - time instant and time interval semantic network.
