=Paper=
{{Paper
|id=Vol-3295/paper17
|storemode=property
|title=Digitalization of HR-Management Processes of Project-Oriented Organizations in the Field of Safety
|pdfUrl=https://ceur-ws.org/Vol-3295/paper17.pdf
|volume=Vol-3295
|authors=Oleh Kovalchuk,Dmytro Kobylkin,Oleh Zachko
|dblpUrl=https://dblp.org/rec/conf/itpm/KovalchukKZ22
}}
==Digitalization of HR-Management Processes of Project-Oriented Organizations in the Field of Safety==
https://ceur-ws.org/Vol-3295/paper17.pdf
Digitalization of HR-Management Processes of Project-Oriented
Organizations in the Field of Safety
Oleh Kovalchuka, Dmytro Kobylkina and Oleh Zachkoa
a
Lviv State University of Life Safety, Kleparivska Street, Lviv, 79007, Ukraine
Abstract
The article presents the results of research on HR-processes. The necessity of automation of
HR-processes is substantiated. The main tasks to be solved by HR-automation are identified.
Particular attention is paid to the issue of automation of personnel management. The
significance of the obtained results lies in the possibility of increasing the efficiency of the
enterprise with HR automation. The main stages of building an incremental life cycle model
for the development of an intelligent decision-making information system (HRIS) with a set
of mathematical methods for decision-making. Creating an effective system for data analysis
and selection of candidates is one of the most important tasks of a modern personnel
management system (PM). Computerization of personnel records management is one of the
main conditions for the rational organization of record-keeping processes in the organization,
a means of improving the efficiency of personnel services, a factor in increasing productivity
and efficiency of managers. Tasks that are solved in the process of staffing belong to the
class of tasks of complex decisions in conditions of uncertainty. From this point of view,
scientific and methodological tools were analyzed, acceptable and selected expert method.
An information-analytical system has been developed for the formation of project teams of
higher education seekers in higher education institutions of the civil protection system and to
improve the quality of personnel decision-making through an organized relational database
and knowledge.
Keywords 1
Information system, project team, database, digitalization, HR process
1. Introduction
Projects of digitalization and digitalization of personnel management and formation of project
teams are developing rapidly. The transformation of HR-technologies under the influence of
digitalization of business processes has led to the development of innovative human resource
management systems (HRM systems). An important factor for higher education institutions in the
civil protection system at the stage of initiating the development of a new information analytical
system of personnel management (HRIS) are resource constraints, which determine the choice of
flexible Lean project management methodology (careful use and allocation of resources).
2. Analysis of recent research and publications
Domestic and foreign scientists, such as Bushuyev S.D. [1-3], Chumachenko I.V. [5],
Kononenko I.V. [4] and others, deal with the issues of effective human resources management.
The development of modern science and computer technology poses a number of new challenges
___________________________
Proceedings of the 3rd International Workshop IT Project Management (ITPM 2022), August 26, 2022, Kyiv, Ukraine
EMAIL: Justdoitolejka@gmail.com (Oleh Kovalchuk); dmytrokobylkin@gmail.com (Dmytro Kobylkin); zachko@ukr.net (Oleh Zachko)
ORCID: 0000-0001-6584-0746 (Oleh Kovalchuk); 0000-0002-2848-3572 (Dmytro Kobylkin); 00000-0002-3208-9826 (Oleh Zachko)
2022 Copyright for this paper by its authors.
Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR Workshop Proceedings (CEUR-WS.org)
for scientists and researchers, such as optimizing the recruitment process and reducing the processing
time of relevant data of project team members.
According to research by Bersin & Associates, organizations with a formalized strategy in the field
of HR are 26% more effective than their competitors. HR managers of innovation organizations use
an approach based on data analysis by statistical methods and analytical complex. Due to this, the
manager chooses the optimal solution. According to a survey conducted by KPMG International
Cooperative, only 17% of managers believe that HR analysis in their organizations is focused on
solving rational personnel tasks. The key trends in HR transformation are HR analytics, optimization
of the use of resources for personnel and strategic planning of human resources. HR analytics allows
you to use different types of data to predict effective models and improve processes. Cost
optimization based on data analysis helps to increase the efficiency of human resource management
and identify new opportunities to achieve the goals of the organization. Through analytical processing
of information, managers can effectively plan the need for human resources, anticipate the number of
staff in the future, assess the staffing or overstaffing, as well as recommendations for specific
proposals to support management decisions.
Based on analytics data to manage project team members, you can optimize talent selection and
increase their involvement. And given the percentage of voluntary redundancies, the task of retaining
employees becomes key for HR teams. The greatest investment in HR analytics and data-based
decision support systems requires integration with databases to provide an integrated approach to
analyzing team members' personal data. According to Gartner, by 2025, an average of 60% of
organizations in the world will invest in automation of HR processes.
A simple and fast method of consolidating and comparing BIO HR data (benchmarking, insights,
opportunities) allows to analyze eNPS (staff loyalty), assess competencies, performance of team
members on "reference" parameters and identify "hidden" knowledge and patterns that correlate with
the results of the organization.
Deloitte found that the income of companies that use HR analytics is 82% higher than those who
do not use this approach.
To support decision-making, effective information support HR managers need to use databases
and knowledge bases, the optimal application of methods and rules of forming a system of criteria,
ranking candidates, selection and analysis of candidates will intellectualize the personnel decision-
making process.
In [18] the scientific and methodological apparatus of personnel selection for the internal troops of
the Ministry of Internal Affairs of Ukraine is considered, which uses the theory of fuzzy sets to
describe four-point scales used in expert assessment of candidates. This has significantly simplified
the procedure for processing evaluation results, formalized them and developed a software product
that implements information and analytical technology of professional selection.
Fuzzy cognitive maps (FCMs) were proposed by B. Cosco [16] and are used to model the causal
relationships found between the concepts of a particular field. Unlike simple cognitive maps, FCMs
are a fuzzy oriented feedback graph whose nodes are fuzzy sets. Thus, FCM combines the properties
of fuzzy systems and neural networks. The active use of fuzzy cognitive maps as a means of modeling
systems is due to the possibility of visual representation of the analyzed system and the ease of
interpretation of causal relationships between concepts.
In [15] as a prototype was selected developed at the department software product "Selection", used
to solve problems of decision support for the selection of personnel to replace certain categories of
servicemen of the Ministry of Internal Affairs of Ukraine. To create an intelligent system, it is
advisable to use the concept of "rapid prototype", the essence of which is that at the initial stage a
version of the IS is developed, which must meet two conflicting requirements: solve typical problems
and labor costs should be minimal.
In [14] the possibility of applying the concordance coefficient W, Spearman's rank correlation
coefficients WS and Fechner sign correlation coefficients WF and modified concordance coefficient
(MCC) Wm is considered to solve the problems of rating evaluation of some objects of comparison
(OP) based on the results of examination with the use of four-point order scale. When solving
professional selection problems, it is advisable to use the modified concordance coefficient in
algorithms for constructing ranked lists of any comparison objects using a four-point order scale. This
ratio is also suitable for establishing the consistency of the results of expert evaluation, the level of
preparedness of an individual specialist or production unit to perform tasks within the so-called team.
In [18] the analytical system for the admission committee of the university is investigated. The
system (IAS) provides organization and support of the admissions committee at all stages, from the
analysis of the recruitment plan to the formation of the student body. To implement the functions, an
automated workplace was created with the help of WEB-technologies, the implementation of an
interactive user mode with the implementation of the necessary checks directly in the process of
entering data into the database. IS is a single WEB-interface for the operator, which allows you to
serve the applicant. The operator carries out the registration of the entrant by filling in the personal
data of the entrant. The recommendation for enrollment is made by creating a list of applicants using
existing forms through a web-browser. The information is entered into the database.
Willingness to use HR analytics remains a serious problem. After several years of discussing this
issue, only 8% of respondents said they had useful data; only 9% believe that they have a good
understanding of what characteristics of employees lead to success in their organizations; and only
15% of all used HR systems and a set of talent indicators for line managers.
Analysis of information systems has shown that not one of them does not allow to effectively
solving problems due to their low functionality. At the same time, the cost of these IAS was
unacceptably high. As a result, it was decided to develop its own automated IAS, which would meet
the requirements of higher education institutions of the civil protection system, would solve all
current problems, and would be able to expand their functionality when changing the regulatory
framework.
Despite a number of studies in this area, many issues related to the automation of the information
system of the human resources department remain unresolved. Existing information systems in HR do
not fully use analytical data processing and forecasting based on data. This reduces efficiency in
human resource planning. The main difficulties arise when combining the organization's HR
operations with personnel data, underdeveloped analytics and integration between different data
sources, HR systems.
3. The bulk of research
In the current market of IT products, the available automated personnel management systems can
be divided into those based on the concept of ERP, CRM-systems (customer relationship
management), financial and analytical systems, reference systems, information security systems,
design systems CASE means.
ERP (Enterprise Resource Planning) - an organizational strategy for integrating production and
operations, human resource management, financial management and asset management, focused on
continuous balancing and optimization of enterprise resources with a specialized integrated
application software package that provides a common data model and processes for all areas of
activity. ERP is primarily an information system that allows you to store and process most of the
critical data for the company's work - the role of analytics.
HRIS software often contains a number of interconnected databases. HRMS (Human Resource
Management Software) is a more complete human resource management tool that offers several
functions of human resource management, such as payroll, payroll administration, performance
analysis and review, and recruitment and training. Personnel management system is a multifunctional
software. Therefore, you should be extremely careful before choosing one for your company; HRIS
should be mobile and user-friendly. In the Ukrainian markets, mainly such products of Western
companies as «SAP», «Oracle», «BAAN», «PeopleSoft» and «Platinum» represent this type of
system.
To automate the activities of the personnel department, as for any other department of the
enterprise, a number of software products have been developed. The choice of automated information
system (IAS) depends on its functional features, the scale of the enterprise and so on. The list of
software products developed for human resources is quite large and constantly updated. However, not
all companies can afford quite expensive software products, and some organizations still keep all
records in paper form.
For successful management of organizations it is necessary to be able to properly make various
management decisions and choose methods of making them. The article considers the peculiarities of
the application of the method of expert assessments for decision-making in the functioning of project
institutions. The current objective methods of determining the optimal development of the
organization in conditions of uncertainty are not able to accurately reflect in quantitative terms the
qualitative content of HR processes and do not allow to determine a comprehensive assessment.
Therefore, one of the alternatives is to use the method of expert assessments.
When solving expert evaluation tasks in the fields of qualimetry and professional selection using
different scales, there is a need to identify the relationship between quantitative and qualitative
indicators of some objects of comparison (OP), if they need or can be ranked. To do this, use the
Pearson correlation coefficient for scales of relations, intervals and quantitative scales, Spearman or
Kendall rank correlation and others - for the order scale.
In the practice of solving the problems of expert evaluation, the concordance coefficient is used as
the opinions of experts agree [15].
, (1)
where n is the number of factors; m - number of experts; dj - deviation of the amount from the
average amount; Ti - the results of intermediate calculations.
Sj- , (2)
where Sj is the sum of the ranks. The concordance coefficient takes values from 0 to 1. The greater
the value of the concordance coefficient, the greater the degree of agreement of experts. At
W = 1 there is a complete agreement of experts; if W = 0, then there is almost no consistency.
When evaluating a comparison object by several parameters, the total evaluation of the object is as
follows: experts make judgments about the weight of parameters (eg, criteria weights) and evaluate
the object according to all parameters (eg, evaluation of alternatives by criteria). Analysts process the
received estimates. Calculate the normalized weight of parameters (eg, criteria) according to the
formulas of arithmetic mean, geometric mean or weighted average. Then comprehensive assessments
are normalized. For the analysis of candidates, it is advisable to use the index method, which is based
on the concept of "reference" candidate - a talent who has the necessary skills for the project. The
suitability of the applicant is determined on the basis of the ratio of the deviation from the "ideal
candidate" to the maximum deviation. If the deviation of the candidate is equal to the maximum, the
coefficient of conformity will be equal to 0. If the candidate has all the necessary skills, it coincides
with the "portrait of the ideal candidate" and the coefficient of conformity is 1. Deviation from the
reference candidate is calculated as the distance between points. Since the number of requirements for
inclusion in the team of a higher education institution of the civil protection system may vary due to
the regulatory component, the following is proposed (formula 3):
, (3 )
where – is the distance from the skill set of the candidate to the skill point ,
n – is the number of skill requirements,
xi – is the coordinate of the point x on the i-th axis.
The calculation of the deviation fr
om the "reference" object of comparison for inclusion in the project with two requirements can be
visualized in the form of a coordinate system, where the X and Y axes reflect the level of skills for
training in ZVO TsZ (Figure 1).
Skills level 2
max deviation from The area of ideal
the "reference candidates
value"
Differential
(deviation) from the
"reference value"
Candidate
Skills level 1
Figure 1: Area of the "ideal" candidate for inclusion in the team of higher education institutions of
the civil protection system
If the candidate's level of all skills is equal to or greater than required, then he falls into the area of
"ideal" applicant and the coefficient of compliance will be equal to 1 (Figure 1). The input index
method receives candidate profile data and information on the requirements for inclusion in the
project team. As a result, a list of profiles of candidates is generated, sorted in descending order of the
applicant's compliance with the project for training in a higher education institution of the civil
defense system. The most suitable candidates will be presented to decision makers from the first
records of the received rating.
Information and analytical system of professional selection involves the implementation of
such a sequence of procedures.
1. Expert assessment of candidates' compliance with a certain model.
2. Preliminary processing of evaluation results. Construction of visualized personograms of
candidates. Calculation of the coefficients of conformity of each of the components of the
professional profile for each of the candidates.
3. Calculation of generalized indicators and ratings of each of the candidates and compiling a
ranking list.
A graphical representation of the incremental life cycle model is presented in Figure 2.
Incremental development is the process of partial implementation of the entire system and
functionality. It operates on the principle of a cascading model with floors, so that the functional
capabilities of the product, which are suitable for use, are formed earlier. Requires a complete pre-
formed set of requirements, can begin with the formation of common goals, which are then clarified
and implemented. The incremental model is an advanced cascade model. As a result of each
increment, a functional product is obtained. The use of successive increments allows you to combine
the results into a complex product, ie the ability to divide the problem into parts that can be
effectively managed.
The composition of software and hardware depends on the specific conditions of the enterprise,
namely the scale of production, staff, organizational structure of the management staff, the scale of
document flow, the need for operational and retrospective information, the degree of centralization of
documents and more.
HRIS is a set of intelligent information applications and tools used to manipulate data, analyze it
and provide the results of such analysis to the end user. Modern DSS makes it possible to predict the
degree of influence of decisions on the further development of the organization. By multidimensional
analysis we mean the technique of presenting data from different points of view, or "measurements".
Data is uploaded to the repository as facts, and "measurements" are indexes that provide easy and
quick access to these facts from different directions. The implementation of multidimensional analysis
requires the support of a specialized multidimensional database.
initiation
Increment 1 analysis of HR Increment 4 logical level Increment 7 cognitive model of
operations knowledge
f
(HR) E1 E2
f1
(HR)
f2
(HR)
... fn(HR) A1
R
ER diagram An
Increment 5 physical level of Increment 8 output of logical
Increment 2 IDFEO simulation judgments based on the neural
relational database
network
RelationR1 x 1
xn Xn 1
HRIS RelationR2 x
2
2 X y
A0 x
Relation Rn n
n
Increment 3 Decomposition of Increment 9 Recommendation
Increment 6 block diagram of the
the HRIS unit for the manager
database
A1 Information and
A2 Database analytical HR
... manage support
An ment
system
Iteration n
Iteration 1 testing
Iteration 2 ...
Figure 2: Incremental HRIS development lifecycle model for higher education institutions of the civil
protection system
Multidimensional processing tools can be implemented within relational technology. In DSS
operating on aggregate data, the traditional technology of preparing integrated information based on
queries and reports has become ineffective due to a sharp increase in the amount and variety of source
data. The solution was found and formulated in the form of the concept of data warehouse (Data
Warehouse, DW) [18].
Using the client-server architecture maintains the maximum level of storage reliability, relevance
and reliability of programs designed for many users of IS with a centralized database, independent of
the hardware of the database server.
Features of the software solution are:
its web-orientation;
adaptability to the specifics of the activity;
the ability to select candidates from the list, after entering the requirements for the
candidate;
support for an unlimited number of users working with databases;
the ability to create reports and various documents of templates.
Therefore, the client-server concept was chosen for HRIS, the architecture of the system is shown
in the Figure 3.
The use of the universal modeling language UML allows you to define, visualize and document
object models of system software. In turn, this makes it possible to simplify the understanding of the
structural organization of the software product and determined the specification of the tasks of
implementation of the software complex of the HRIS information system.
User interface
HR
Client desk ОПР managers
experts analysts candidates
Data input, output and cosolidation unit
HRIS components
Relational database management system
Data
Data warehouse Database
models
Server part
Knowledge base management system, models and rules
Block of
OLAP operational Data
forecasting and
Analytical subsystem data analysis unit mining
logical judgments
Figure 3: HRIS Architecture
For the design, construction and software implementation of the system requires its orientation to
support decision-making within the subject area. The simulation phase builds regression and
optimization of a subset of variables, decision-making based on neural network techniques,
construction of classification trees for optimal set of variables and optimal partitioning of many
objects, clustering and optimal grouping of objects. At the stage of data preparation access to any
relational databases, text files is provided. Based on the prepared data, special procedures
automatically build different models for further forecasting, classification of new situations,
identification of analogies. The application data supports the construction of five different types of
models - neural networks, classification and regression decision trees, Bayesian analysis and
clustering (Figure 4).
Requirements for
Methods of building relational database
data models
Logical data model
HRIS Data model
development Physical data model
Database design
Change management
Software for
Characteristics of automated UML
functional modeling
subsystems and tasks Formation and filling
of the organization of the database of
Infological and
candidates in the
datalogical levels
Central Election Testing
CASE systems Commission systems
Information support
Figure 4: Contextual diagram of the decomposition of the process of creating a human resources
management database for higher education institutions of the civil protection system
Structural and functional modeling of IDEF0 for the HRIS project allows graphically describing
processes and comprehensively studying the information system. Due to the methodology of
functional modeling, the system can be seen as a set of interconnected functions (functional blocks).
Therefore, the process of designing a web-system must begin with the development of a context
diagram IDEF0.
The following entities have been identified for the database:
users (decision makers, experts and other users) store information about users;
candidates (applicants for inclusion in the project teams of higher education institutions
of the civil protection system) store information about the candidates;
resumes - stores information about personal data on the basis of which a personal file is
formed;
role - stores information about roles in projects.
The figure shows information about the attributes to the corresponding entities. Defining the
essence and attributes of the information system was built ER-diagram, which is shown in the
Figure 5. ER-diagram is a graphical representation (Figure 5) of entities and their relationships.
HR manager
PKid_user
name
last name
Candidates Personal data
PKid_candidate PKid_resume
Name id_candidate Role in the project
Last name id_role PKid_role
Email adress resume description
name
status
Skills Require skills
PKid_скіли PKid_requireskills
id_candidate id_role
name name
value value
Additional Information
PKid_info
id_candidate
name
value
Figure 5: ER diagram of entities and their relationships
With the help of this chart you can see how the entities are interconnected in the information
system. Candidates and resumes have connections "to each other", one candidate can send one
resume, and one resume should refer to one candidate; roles and resumes have a one-to-many
relationship, one resume can refer to one vacancy, one vacancy can have many resumes; role and
required skills have a one-to-many relationship, one skill can relate to one role, one role can have
many skills candidates and skills have a many-to-many relationship, one skill can relate to many
candidates, one the candidate must have at least one skill; candidates and additional info have a
"many to many" link, one item can apply to many candidates, one candidate must have at least one
parameter. The advantage of relational databases is obvious. Virtually all database management
systems allow you to add new data to the table, modify, view and print them.
During the interaction of different functional units of higher education institutions, the civil
protection system accumulates a lot of data on its activities, but these data still need to be structured
into information for effective management decisions. Analytical methods allow decision makers to
complete the entire cycle of work with larger volumes and unexplained statistical structure through
Data Mining. It includes the following stages: sampling, research, modification, modeling, evaluation
of results. The analytics and forecasting subsystem contains methods of statistical data processing,
which can be divided into four interrelated sections:
preliminary analysis of personnel statistics;
identification of connections and regularities (linear and nonlinear regression analysis,
correlation analysis);
multidimensional statistical analysis (linear and nonlinear cluster analysis, component
analysis, factor analysis);
dynamic models and forecast based on time series [theory of decision support systems].
The analytical software package (ASP) includes a special set of software and tools for rating
analysis, which allows you to study the data of comparison objects and form on the basis of their
indicators different ratings for decision makers. Rating analysis allows you to assess both the current
state of the set of objects, and their state in the past on the basis of time series. The comparison of the
obtained result with the state of other similar objects of comparison is carried out.
The analytical software package implements a wide range of opportunities to view various charts
and compile ratings-reports. Another tool for analyzing and presenting data is Web Intelligence,
which has the means to build reports through a web browser.
To model the base of rules, it is advisable to use fuzzy cognitive maps, tasks of rules and functions
of belonging to the thermal baths and derivation of the dynamics of system development under
different input influences. Analysis of the developed cognitive map allows you to quickly obtain
information about the behavior of the system and conduct experiments. Institution of higher education
in the system of civil protection - a complex organizational and technical system, which consists of a
train (formula 4):
D(t), S(t), Y(t), E(t), t , (4)
which takes into account the following parameters: D - actions of top management, resource
allocation; S - environmental factors; Y - initial indicators of organization development; E - is a set of
concepts connecting input and output variables; t is the time. The task of optimal management of this
system and the study of its behavior in the process of human resource management on the basis of
higher education institution in the system of civil protection and the environment is initiated. The
system is characterized by a fuzzy logic of human factors. One of the approaches to the construction
of a generalized fuzzy cognitive map is proposed, in which input and output variables are
distinguished, and connections are described by fuzzy rules. In the set of concepts C of the fuzzy
causal network G = (C, W) there are many input effects X = {x1, x2, ..., xn}, many output effects
Y={y1, y2, ..., ym} and intermediate concepts E = {e1, e2, ..., ep}, set of connections between
concepts, W [0; 1]. Each concept ei, i=1,…P is characterized by a term-plural of linguistic variables
(formula 5)
Ti={T1,T2,...Tm}, (5)
where mj - is the number of typical states of the concept. To describe each term Ti, a term set with
membership function μ (x) is constructed. We present the model of a higher education institution in
the civil protection system in the form of a generalized fuzzy cognitive map (Figure 6).
Figure 6: Model of the personnel department of the institution of higher education of the civil
protection system in the form of a fuzzy cognitive map
Let's highlight the concepts:
E1 - Innovative projects of the HEICPS;
E2 - resource constraints of the HEICPS;
E3 - competence of team members;
E4 - advanced training;
E5 - management of selection and placement processes;
E6 - personnel needs planning;
E7 - development of project teams of he HEICPS;
E8 - human resource management;
E9 - scale and type of organization;
Developed fuzzy cognitive map, which simulates the behavior of ATS ZVO TsZ, covers its
main operating elements:
strategic management of the project management office (PMO) of the HEICPS;
allocation of resources for the development of new projects (concepts X1,…, Xn);
functioning of the HEICPS (concepts Ei, i = 1);
factors that determine the activities of the organization (concepts E1, E2);
characteristics of candidates (concepts E3, E7, E8, E9);
Fuzzy-cognitive approach to building simulation models of complex systems allows for optimal
control of such systems without building an accurate mathematical model. .
As a function of belonging to the rules was chosen Gaussian type function, which has become
widespread in fuzzy networks. It is described by the formula:
(6)
and operates with two parameters: δ and c. The parameter c means the center of the fuzzy set, and
the parameter δ is responsible for the function. FCMapper software is used to calculate the parameters
of fuzzy cognitive maps. The general sequence of steps for building scenarios based on the analysis of
fuzzy cognitive maps is presented in Figure 7.
Figure 7: Generalized process of building rule base scenarios based on fuzzy cognitive maps
The function of belonging of an element to the set takes values in the interval [0, 1], and not only 0
or 1 (a characteristic feature of fuzzy logic). Thus, Cosco's cognitive maps allow us to indicate the
"intensity" of the influence between factors. Such a mathematical structure allows to formalize the
purely subjective opinion of the decision-maker, formed in the context of incomplete information
about the membership of an element in a group. Figure 8 shows a neural network model of personnel
selection for the team of a higher education institution in the civil protection system using the “black
box” model (Figure 8).
q1
1
d1=f ( qi1ki)
1
q2 d1' d1
2 2
q3 d2=f ( qi2ki) K32' d2' d1
... ...
... ... n
dn dn
m
d3=f ( qimki)
qn
Figure 8: The scheme of the neural network of the formation of the project team of HEICPS
The basic component of the neural network is the data processing node. Each processing node
sums the values of its inputs. Next, this amount passes through an arbitrary activation function to
obtain the original value of the node. The state of the original neuron is determined according to
formula 7:
(7)
where q𝑖 - is the value of the i-th input of the neuron (initial data of the candidate);
k𝑖 - weight of the i-th synapse (candidate);
d - the value of the state of the neuron (inference of logical judgment-management decision).
Depending on the positive or negative answer, decisions will be made for decision makers. A
genetic algorithm was chosen as the neural network self-learning algorithm. Genetic algorithm is an
adaptive heuristic search method, which is a probabilistic search algorithm based on the mechanism
of optimal selection and natural genetics. It is used to add hidden scales and source layers of the
neural network. This algorithm contains the following component procedures: formation of the initial
population, crossover operator, mutations, assessment of the fitness of individuals, selection. The
population contains many alternative solutions, presented in the form of population persons. The
algorithm completes its work if the value of the recognition error of the best person of the population
does not change n populations. The larger n, the fewer recognition errors and the more accurate the
neural network. Recognition error is calculated by formula 8:
𝜀𝑖 = 1 − (8)
where 𝑦0 – is the reference value of the output signal (portrait of the "ideal candidate");
𝑦 – the estimated value of the original when recognizing the defect of the printed circuit board
from a self-learning sample with this set of weights.
The artificial neural network shown in Figure (7) is reformatted into a person's chromosome,
filling it with weights (from top to bottom, left to right). Thus, the chromosome is a set of genes -
weights of candidates for inclusion in the project team of HEICPS is described by formula 9.
𝐶 = (k11, k12, … k1𝑚, k21, k22, … k2𝑚, … k𝑛1, k𝑛2 , … k𝑛𝑚, k11, k12 , … k1m, k21, k22, … ,
k𝑚 , k𝑚2 , … k𝑚n) (9)
The only direct calculations of machine learning genetic algorithm is the movement of the neural
network. Because of this, the system requirements are very flexible compared to in-depth neural
network training); adaptability (various tests and ways of manipulating the flexible nature of genetic
algorithms could be adapted and integrated).
4. Conclusions
Thus, information system has been developed to support personnel decision-making for higher
education institutions as an expert system. The analytical subsystem will ensure the organization and
support of the Admissions Committee at all stages, starting from the analysis of the recruitment plan
for higher education institutions of the civil protection system. The practical significance of the
obtained results is that the improved model and methods should be implemented in the form of a
software module. This improves the quality of candidate selection. The created system will allow
sorting, selecting the necessary information from the lists, performing arithmetic operations and
performing many other functions that will automate the routine work of the HR specialist.
The developed models consider the use of databases and knowledge bases required for storage,
monitoring and analysis of large amounts of information for the operation of an intelligent system
focused on decision support for different classes of tasks. The availability of comprehensive decision
support methods used at each stage of project team formation will significantly increase the number
of functional tasks in human resource management.
5. References
[1] S. Bushuyev, D. Bushuiev, A. Zaprivoda, J. Babayev, Ç Elmas, Emotional infection of
management infrastructure projects based on the agile transformation. CEUR Workshop
Proceedings 2565 (2020) 1-12.
[2] N. Bushuyeva, I. Achkasov, V. Bushuieva, B. Kozyr, Ç. Elmas, Managing infrastructure projects
driving by global trends. CEUR Workshop Proceedings 2565 (2020) 13-23.
[3] N. Bushuyeva, D. Bushuiev, V. Bushuieva, I. Achkasov, IT Projects Management Driving by
Competence, in: 2018 IEEE 13th International Scientific and Technical Conference on Computer
Sciences and Information Technologies, CSIT 2018, pp. 226-229, doi: 10.1109/STC-
CSIT.2018.8526680.
[4] І. Kononenko, G. Sushko, Forming a project team to develop information and communication
technologies. Information Technologies and Learning Tools 73 (2019) 307-322.
[5] I. Chumachenko, Methods of human resources management in the formation of teams of
multiprojects and programs, National Aerospace University Zhukovsky, KhAI, Kharkov (2015).
[6] D. Kobylkin, O. Zachko, Structural models of safety-oriented management of infrastructure
projects decomposition, in: IEEE 15th International Scientific and Technical Conference on
Computer Sciences and Information Technologies, vol. 1, Lviv-Zbarazh, Ukraine, 2020, pp.
131–134.
[7] D. Kobylkin, O. Zachko, N. Korogod, D. Tymchenko, Development of models for segregation
the elements of infrastructure projects management with the application of a mono-template
under safety-oriented management. Eastern-European Journal of Enterprise Technologies 6 (3)
(108) (2020) 42–49.
[8] D. Kobylkin, O. Zachko, R. Ratushny, A. Ivanusa, C. Wolff, Models of content management of
infrastructure projects mono-templates under the influence of project changes, ITPM, Slavsko,
Lviv region, Ukraine, 2021, pp. 106–115.
[9] O. Zachko, D. Chalyy, D. Kobylkin, Models of technical systems management for the forest fire
prevention. Naukovyi Visnyk Natsionalnoho Hirnychoho Universytetu 5 (2020) 129–135.
[10] D. Kobylkin, O. Zachko, V. Popovych, N. Burak, R. Golovatyi, C. Wolff, Models for Changes
Management in Infrastructure Projects, ITPM, Slavsko, Lviv region, Ukraine, 2020 106–115.
[11] O. Zachko, O. Kovalchuk, D. Kobylkin, V. Yashchuk, Information technologies of HR
management in safety-oriented systems, in: IEEE 16th International Scientific and Technical
Conference on Computer Sciences and Information Technologies (CSIT 2021), vol. 2, Lviv,
2021, pp. 387–390.
[12] Y. Kozlov, O. Novikova, Method of constructing ranked lists of candidates for replacement
positions for personnel decisions. Collection of scientific works of the Kharkiv National
University of the Air Force (2018) 111-115.
[13] G. Khimicheva, Y. Farfanyuk, N. Kolesina, Application of expert decision-making methods in
the management of project organizations. Eastern-European Journal of Enterprise
Technologies 3(3(51) (2011) 23–25. doi.org/10.15587/1729-4061.2011.1507
[14] V. Dubrovina, V. Kozlov, Y. Kozlov, O. Novikova, Establishing consistency of results in solving
problems of expert evaluation. Collection of scientific works of the National Academy of the
National Guard of Ukraine 2 (24) (2014) 92-94.
[15] S. Poltorak, Practical application of information-analytical technology of personnel selection,
Collection of scientific works of the Academy of Internal Troops of the Ministry of Internal
Affairs of Ukraine 1 (19) (2012) 49-51.
[16] B. Kosko, Fuzzy Cognitive Maps, International Journal of Man-Machine Studies 1. (1986) 65-
75.
[17] O. Novikova, Information technology to support personnel decision-making for higher education
institutions of Ukraine, author's ref. dis. for science. degree can. tech. Science: 05.13.22.
Kharkiv, 2019.
[18] A. Moroz, N. Pokhlebina, V. Hobin, Information-analytical system of the admission commission
of ONAHT as a basis of automated management of student contingent formation. Automation of
Technological and Business Processes 12 (4) (2020) 36-42. doi.org/10.15673/atbp.v12i4.1933.
[19] A. Ivankevich, V. Piterska, A. Shakhov, V. Shakhov, V. Yarovenko, Proactive Strategy of Ship
Maintenance Operations, in: IEEE 2019 14th International Scientific and Technical Conference
on Computer Sciences and Information Technologies, CSIT 2019, 2019, pp. 126-129, doi:
10.1109/STC-CSIT.2019.8929741.
[20] O. Danchenko, M. Palchynska, I. Azhaman, N. Telichko, M. Sadova, Psychological means of
theoretical modeling of the optimum number of project staff. International Journal of
Management 11(4) (2020) 414-426.
[21] S. Chernov, L. Chernova, S. Titov, Reduction in Discrete Optimization Problem, in: IEEE 13th
International Scientific and Technical Conference on Computer Sciences and Information
Technologies, CSIT 2018 – Proceedings, 2018, pp. 230-233. doi: 10.1109/STC-
CSIT.2018.8526718.