<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Archiving and Interchange DTD v1.0 20120330//EN" "JATS-archivearticle1.dtd">
<article xmlns:xlink="http://www.w3.org/1999/xlink">
  <front>
    <journal-meta>
      <journal-title-group>
        <journal-title>V. A method for improving the consistency of individual expert rankings
during their aggregation. Journal of Automation and Information Sciences</journal-title>
      </journal-title-group>
    </journal-meta>
    <article-meta>
      <article-id pub-id-type="urn">nbn:de:</article-id>
      <title-group>
        <article-title>Mathematical support for automated candidate selection using a documentary approach to assessing their competence⋆</article-title>
      </title-group>
      <contrib-group>
        <contrib contrib-type="author">
          <string-name>Hryhorii Hnatiienko</string-name>
          <email>hnatiienko@knu.ua</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Oleh Ilarionov</string-name>
          <email>oleg.ilarionov@knu.ua</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Hanna Krasovska</string-name>
          <email>hanna.krasovska@knu.ua</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <contrib contrib-type="author">
          <string-name>Yevhen Patkin</string-name>
          <email>yevhen.patkin@knu.ua</email>
          <xref ref-type="aff" rid="aff0">0</xref>
        </contrib>
        <aff id="aff0">
          <label>0</label>
          <institution>Taras Shevchenko National University of Kyiv</institution>
          ,
          <addr-line>Volodymyrska 64, 01033 Kyiv</addr-line>
          ,
          <country country="UA">Ukraine</country>
        </aff>
      </contrib-group>
      <pub-date>
        <year>2025</year>
      </pub-date>
      <volume>3241</volume>
      <issue>1</issue>
      <fpage>0000</fpage>
      <lpage>0002</lpage>
      <abstract>
        <p>This paper proposes an approach to formalizing the profile of higher education programs by constructing a matrix of links between disciplines and program learning outcomes. The concept of quantitative measurement of the contribution of each discipline to the achievement of program outcomes through the use of ECTS credits is introduced. Such a model allows for increasing the transparency of the structure of educational programs, ensuring their adaptability to changing labor market conditions, and identifying possible gaps in the educational process. The formulation of the task and the mathematical model of automated candidate selection using a documentary approach, which is used to assess their competence, are also considered. The stages of candidate selection and the tools for determining the similarity between job requirement profiles and candidate documentary profiles are presented. The purpose of the work, the formulation of the task, and the mathematical model of automated candidate selection are described. The educational trajectory is also considered as an indirect justification for the use of a documentary approach in personnel selection and in the tasks of assessing the competence of experts.</p>
      </abstract>
      <kwd-group>
        <kwd>candidate selection</kwd>
        <kwd>expert competence</kwd>
        <kwd>documentary approach</kwd>
        <kwd>similarity measure</kwd>
        <kwd>model</kwd>
        <kwd>formalization</kwd>
        <kwd>individual educational trajectory 1</kwd>
      </kwd-group>
    </article-meta>
  </front>
  <body>
    <sec id="sec-1">
      <title>1. Introduction</title>
      <p>Ensuring automated ranking of candidates for positions based on a documentary approach is a
pressing issue, the resolution of which will contribute to the objectification of many personnel
selection problems, in which the human factor always plays a necessary role.</p>
      <p>In the context of rapid technological development and constant changes in the labor market,
ensuring prompt and adequate personnel assessment is a pressing issue. The reason for the constant
attention to objective and reasonable assessment is the need to take into account the subjective
aspects of decision-making in all procedures related, in particular, to the selection of candidates for
various selection and decision-making tasks.</p>
    </sec>
    <sec id="sec-2">
      <title>2. Review of research in this scientific field</title>
      <p>The development of higher education programs requires ensuring that their structure meets the
current requirements of the labor market, professional standards, and student needs. One of the key
tasks is to ensure the transparency and adaptability of these programs [1, 2]. Formalizing an
educational program through its profile allows for establishing clear links between competencies,
program outcomes, and educational components. The diversity of educational programs creates
difficulties when transferring students between programs, especially in the context of minimizing
the loss of existing competencies and program outcomes [3, 4].</p>
      <p>The purpose of this work is to describe approaches to automating the assessment of candidates'
suitability for vacant positions, as well as to describe and develop a formalized model of an
educational program profile that ensures flexibility and adaptability in the organization of the
educational process and the transfer of academic credits to higher education applicants. Based on the
above formalization, the application of a documentary approach is considered when selecting
candidates for expert groups or for certain positions that allow for such a formal comparison. At a
minimum, this technology can be applied in the early stages of candidate selection.</p>
      <p>Such approaches minimize the influence of subjectivity in the selection and recruitment of
personnel. An automated approach can be useful in decision-making situations characterized by the
following factors:
●</p>
      <p>A large number of candidates.
● The possibility of reliable self-presentation.
● The task involves a very significant subjective factor, etc.</p>
      <p>The subjective aspect of decision-making regarding personnel assessment, in turn, depends
significantly on the adequacy of the tasks facing the expert group and on the assessment tools used.
The arsenal of expert assessment methods is quite broad, but the choice of the most attractive
methods from among the available tools plays an extremely important role and significantly
influences the final results [5, 6]. The human factor becomes a strategic aspect in the development
of an enterprise in today's unstable market conditions [7, 8].</p>
    </sec>
    <sec id="sec-3">
      <title>3. Problem setting and mathematical model of automated candidate selection using a documentary approach</title>
      <p>Let us assume that in a decision-making situation related to human resource management, there is
a requirement for the level of education of candidates and the possibility and necessity to formalize
the requirements for the work experience required from the candidate. We will assume that this
requirement is characterized by a vector consisting of two elements
An Example of equation</p>
      <p>
        In vector (
        <xref ref-type="bibr" rid="ref1">1</xref>
        ), element x1 corresponds to the field or fields of knowledge in which the candidate
received their education and, in turn, consists of three elements:
An Example of equation
      </p>
      <p>
        = ( 1,  2).
 1 = ( 11,  12,  13),
 2 = ( 21,  22),
(
        <xref ref-type="bibr" rid="ref1">1</xref>
        )
(
        <xref ref-type="bibr" rid="ref2">2</xref>
        )
(
        <xref ref-type="bibr" rid="ref3">3</xref>
        )
x2 – work experience, which in turn is also a vector and consists of two elements:
x21 – candidate's field of activity;
x22 – length of service in the relevant position in years or number of expert assessments officially
carried out by the candidate.
      </p>
      <p>Thus, developing a profile of the ideal candidate means, first and foremost, formalizing a set of
skills, assessing competencies, and determining the level of experience.
selection is made.
the relevant competencies.</p>
      <p>
        Work experience in expression (
        <xref ref-type="bibr" rid="ref6">6</xref>
        ) is selected from a menu that formalizes the candidates' fields
of activity.
      </p>
      <p>
        Note 1. In all values of vectors (
        <xref ref-type="bibr" rid="ref1">1</xref>
        )-(
        <xref ref-type="bibr" rid="ref6">6</xref>
        ) and their elements, it is permissible to use logical expressions
with the disjunction &amp; and conjunction V signs.
      </p>
      <p>
        Based on the analysis of vectors (
        <xref ref-type="bibr" rid="ref1">1</xref>
        )-(
        <xref ref-type="bibr" rid="ref3">3</xref>
        ) and (
        <xref ref-type="bibr" rid="ref4">4</xref>
        )-(
        <xref ref-type="bibr" rid="ref6">6</xref>
        ), it is necessary to determine the degree of
proximity between the requirements of the employer or selection organizer and the profiles of
candidates [9]. Subsequently, after calculating these proximity measures, a rating is compiled and a
      </p>
      <p>
        Heuristic H1. The quality of the educational trajectory is sufficient to assert that the holder of a
higher education document has sufficiently mastered the declared program outcomes and acquired
(
        <xref ref-type="bibr" rid="ref4">4</xref>
        )
(
        <xref ref-type="bibr" rid="ref5">5</xref>
        )
(
        <xref ref-type="bibr" rid="ref6">6</xref>
        )
(
        <xref ref-type="bibr" rid="ref1">1</xref>
        )
where qi1, in turn, is also a vector consisting of three elements

= ( 1,  2), іє = {1, … ,  },
 1 = ( 1,  12,  13),  є .
      </p>
      <p>
        The values of the elements in formula (
        <xref ref-type="bibr" rid="ref5">5</xref>
        ) correspond to the specialties in diplomas, respectively,
of levels 1, 2, and 3 of education obtained by the candidate. If the candidate does not have a diploma
of the corresponding level, the value of the element corresponding to this level of education is set to
zero in this case.
      </p>
      <p>
        The second element of vector (
        <xref ref-type="bibr" rid="ref4">4</xref>
        ), in turn, is also a vector consisting of two elements:
 2 = ( 21,  22),  є .
      </p>
      <p>
        The methods for setting requirements for the documentary level of candidates have the same
format as the requirements for the position or the profile of the ideal candidate of type (
        <xref ref-type="bibr" rid="ref1">1</xref>
        )-(
        <xref ref-type="bibr" rid="ref3">3</xref>
        ). That
is, depending on the practical situation, the recruiter or the candidates themselves provide feedback
in the form of vectors of type
      </p>
      <p>The rationale for heuristic H1 is described to some extent in this paper below in the section
"Educational trajectory as a rationale for using a documentary approach."</p>
      <p>The quality of the educational trajectory can also be determined based on testing candidates [10].
Depending on the educational trajectory, it is possible to predict the mathematical expectation of
competence acquisition.</p>
      <p>Heuristic H2. The distances between programs adequately reflect the correlation between the
knowledge, skills, and competencies of holders of diplomas or other documents certifying different
levels of higher education: bachelor's, master's, or doctor of philosophy.</p>
      <p>
        Based on heuristics H1 and H2 and the mathematical model described by formulas (1-)(
        <xref ref-type="bibr" rid="ref6">6</xref>
        ), metrics
can be applied to measure the distances between job profiles (
        <xref ref-type="bibr" rid="ref1">1</xref>
        )-(
        <xref ref-type="bibr" rid="ref3">3</xref>
        ) and candidate profiles (
        <xref ref-type="bibr" rid="ref4">4</xref>
        )-(
        <xref ref-type="bibr" rid="ref6">6</xref>
        ). Let
us introduce a metric based on which we will determine the distances from the profile of the ideal
candidate, i.e., the "standard" given by formulas (1-)(
        <xref ref-type="bibr" rid="ref3">3</xref>
        ), to the profile of each of the candidates of type
(
        <xref ref-type="bibr" rid="ref4">4</xref>
        )-(
        <xref ref-type="bibr" rid="ref6">6</xref>
        )
      </p>
      <p>=  ( ,  ),  є .</p>
      <p>To do this, we first create an expert-based similarity matrix of specialties Ms and a similarity
matrix of fields Mg. We also introduce additional heuristics that reflect various relationships between
different specialties and different fields of knowledge.</p>
      <p>
        At the next stage of selection, based on the results of technical interviews and assessments of the
candidates' experience and its duration, the values obtained in (
        <xref ref-type="bibr" rid="ref4">4</xref>
        )-(
        <xref ref-type="bibr" rid="ref6">6</xref>
        ) are refined, depending on the
level of the position. In doing so, emphasis may be placed on creativity, diversity of interests, depth
of specialized knowledge, breadth of views on future activities, etc.
      </p>
      <p>An important element in determining the ranking of candidates for vacant positions in an
organization is the use of a comprehensive approach to candidate assessment. In particular, a
decision on the level of a candidate's suitability for the proposed position can be made based on the
results of a technical interview and assessments of the relevance and duration of their experience.</p>
    </sec>
    <sec id="sec-4">
      <title>4. Stages of decision-making regarding the selection and recruitment of candidates for positions</title>
      <p>Let us describe the structure of the personnel selection and recruitment processes.</p>
      <p>Staff selection or recruitment is a multifaceted process that includes searching, attracting,
selecting, interviewing, and hiring. In turn, staff selection is the evaluation and selection of those
who are already involved in the recruitment process. In general, it should be noted that it is only
possible to verify the reality of candidates' competencies based on the results of a probationary
period, and even such verification is often largely probabilistic in nature.</p>
      <p>Personnel selection is one of the stages of recruitment that focuses on assessing the professional
and psychological qualities of candidates. Candidate selection involves evaluating resumes,
analyzing competencies, determining the level of professionalism, conducting interviews, and testing
[9, 10].</p>
      <p>The assessment of candidates usually consists of conducting interviews, testing their knowledge
and skills, and evaluating their suitability for the corporate culture [11, 12].</p>
      <p>Determining the degree of suitability of candidates for vacant positions can be interpreted as the
distance to the ideal. Therefore, selection procedures should be preceded by the development of a
profile of the ideal candidate, which may include a set of skills, competencies, experience, etc. [13,
14].</p>
      <p>Among the most popular approaches to personnel assessment at the selection stage are:
● determining the suitability of the candidate's profile to the job profile
● testing candidates
● establishing selection rules and preliminary classification of candidates
● documentary approach, etc.</p>
      <p>Moreover, each of these approaches is multifaceted, diverse, and ambiguous in its application.
Possible areas of application for automated rapid analysis of candidates, which can be used in practice
in various decision-making situations:
● selection of experts to form an expert group in various fields of activity and in various subject
areas
● selection of reviewers for student research competitions
● selection of reviewers and opponents when forming ad hoc academic councils for the defense
of doctoral dissertations
● initial selection of members of specialized academic councils for the defense of doctoral
dissertations
● selection of candidates for a certain vacant position
● preliminary assessment of candidates in various competitions using sequential analysis and
screening out options that are unlikely to be successful in advance [15, 16]
● formation of program committees for international conferences [17]
● formation of editorial boards of scientific journals [18], etc.</p>
    </sec>
    <sec id="sec-5">
      <title>5. Features of collective expert evaluation</title>
      <p>An important element in the selection of candidates for vacant positions in any organization is the
consideration of various aspects of collective decision-making in expert evaluation, in particular, the
consistency of assessments.</p>
      <p>The concept of consistency is applied in four aspects:
● if a relation is given on a set of candidates and a certain resulting relation is selected,
consistent with the given
●
when analyzing the consistency of experts or selection methods based on the relations they
specify on a set of candidates without determining the resulting relation
● for metrized preference relations, the concept of cardinal consistency in strength of
preference or supertransitivity is introduced
● in some decision-making situations, consistency is understood as the task of determining the
ranking of candidates for vacant positions according to a metrized or qualitative matrix of
pairwise comparisons that corresponds to a non-transitive relation.</p>
      <p>The accumulated experience of expert assessment in various fields of human activity
convincingly demonstrates that any statistical operations become more useful and justified when the
number of characteristics used for analysis is reduced. Therefore, the problem of aggregating
candidate characteristics into a smaller number of constructed factors, aspects, etc. occupies a
significant place in personnel selection tasks. The analysis of the set of expert assessments of the
entire group of experts or the selection methods used consists in determining the level of overall
consistency of expert assessments and, if necessary, identifying homogeneous subgroups within the
group that bring together experts with consistent assessments. These tasks are dictated by the fact
that the transition to the aggregation of expert judgments is possible only after the structure of expert
judgments has been identified. For example, if the overall consistency of expert assessments is low
and the expert group is divided into several subgroups within which the consistency of expert
assessments is high, then the expert judgments for these subgroups should be aggregated. Kenneth
Joseph Arrow [19] formulated five axioms that correspond to the intuitive idea of fair group choice:
● axiom of universality
● the axiom of connection between group ordering and individual ordering – the condition of
monotonicity
●
●</p>
      <p>Axiom of independence of disconnected objects</p>
      <p>Axiom of sovereignty of group members
● the axiom of the absence of a dictator.</p>
      <p>Kenneth Arrow proved that when there are more than two objects, there is no rule for the
consistency of individual preferences that would satisfy all of the above conditions (Arrow's
paradox). The problems and main results in the field of group choice are discussed in detail by Herve
Moulin in [20]. He also provides a classification, description of methods, algorithms, and procedures
that are widely used in the global practice of collective expert assessment.</p>
    </sec>
    <sec id="sec-6">
      <title>6. Personnel selection tools in the context of personnel assessment approaches</title>
      <p>When assessing personnel, in particular when determining the level of relative competence of experts,
the following groups of assessment tools are the most common, popular, and effective [21, 22]:
●</p>
      <p>Documentary approach
● Self-assessment
● Peer assessment</p>
      <p>Combined methods
●
●
● Testing</p>
      <p>Objective approaches: Effectiveness of participation in previous assessments; Based on the
results of a control assessment; Research on participation in a specific assessment.</p>
      <p>The above-mentioned tools for assessing candidates for vacant positions in an organization are
presented in more detail in Table 1.</p>
      <sec id="sec-6-1">
        <title>It causes distrust due to the methods of formalization and</title>
        <p>consideration of objective data about experts, as well as the
tendency characteristic of modern times to devalue official
documents and the uncritical nature of some institutional
decisions</p>
      </sec>
      <sec id="sec-6-2">
        <title>When conducting the expert self-assessment procedure, information is often obtained about the expert's level of selfconfidence rather than their actual competence.</title>
      </sec>
      <sec id="sec-6-3">
        <title>In some tasks, confrontation may arise within the expert commission and among its members, which sometimes distorts the actual individual competence of experts.</title>
      </sec>
      <sec id="sec-6-4">
        <title>Instead of a synergistic effect, processing unreliable data obtained by different methods in some cases leads to the accumulation of cumulative error</title>
      </sec>
      <sec id="sec-6-5">
        <title>A psychodiagnostic method, which is a process of assessing</title>
        <p>the competencies, knowledge, skills, and personal qualities of
candidates to determine their professional suitability for the
position, compatibility with the corporate culture, and
development potential</p>
      </sec>
      <sec id="sec-6-6">
        <title>Relatively reliable methods for determining the level of competence of experts, the adequacy of which can be assessed and justified to a sufficient degree 6.1</title>
      </sec>
    </sec>
    <sec id="sec-7">
      <title>7. Mathematical model of alignment between competencies, program outcomes, and educational components</title>
      <p>The higher education system in Ukraine is undergoing profound reforms related to the transition to a
competency-based approach and the updating of state standards. According to Order No. 441 of the
Ministry of Education and Science of Ukraine dated April 3, 2024, the new standards no longer define
program outcomes, but only outline integral competencies and a list of general and professional
competencies. The formulation of program learning outcomes is entrusted directly to the developers
of educational programs, which, on the one hand, increases the autonomy of educational institutions,
and on the other, creates the challenge of ensuring transparent and reasonable correspondence
between program learning outcomes and the competencies defined by the standards.</p>
      <p>In such conditions, the content of educational components (disciplines) becomes crucial.
Disciplines are carriers of knowledge and skills that, through the formation of program learning
outcomes, ensure the achievement of graduate competencies. Gaps or duplications in course topics
or imbalances in their structure directly affect the quality of training, even if the program formally
covers all the necessary program learning outcomes and competencies. In this regard, it is important
to find formal tools that can assess the quality of educational content not only at the level of an
individual course, but also in the context of the entire educational program.</p>
      <p>Formal models that can not only describe the structure of educational content, but also provide a
quantitative assessment of its consistency with program outcomes and competencies. To find such
models, let us turn to contemporary research that applies mathematical, ontological, and categorical
approaches to structuring educational content.
7.1.</p>
      <sec id="sec-7-1">
        <title>The educational trajectory as an indirect justification for the possibility of using a documentary approach</title>
        <p>The structure of higher education standards by specialty provides for general and specific
competencies that are ensured by program outcomes. Based on higher education standards, higher
education institutions develop their own educational programs, which differ in the set of educational
components and the distribution of ECTS credits allocated to these educational components. In
addition, higher education institutions independently determine how disciplines ensure the
formation of program outcomes.</p>
        <p>Based on the documents submitted by candidates for vacant positions in the organization, it is
possible to reflect the quality of their educational trajectory.</p>
        <p>The mathematical model of the selection process is based on the KFS (Knowledge Flow Structure)
model. The relationship between competencies, program outcomes, and educational components can
be represented as a KFS model (Fig. 1).</p>
        <p>For each educational program, a matrix can be created in the following format
( ) = (  )  ,
(8)
where: i – number of the educational component;
j – program outcome number.</p>
        <p>The elements of the matrix of type (8) reflect the number of credits of the European Credit
Transfer and Accumulation System aimed at achieving the j-th program outcome through the i-th
educational component.</p>
        <p>Based on matrix (8), the overall profile of the educational program is calculated as a vector
An Example of equation</p>
        <p>= (  ),  = 1, … ,  ,
where vj=Sum(mij, i=1,...,k), j=1,...,n.</p>
        <p>The formalization of the educational program profile of type (9) can be represented as a vector of
ECTS credit distribution by program outcomes:
An Example of equation</p>
        <p>= ( ),  = 1, … ,  ,
where s is the number of credits for the l-th program outcome.</p>
        <p>The credit distribution vector of type (10) can be an important tool for modeling and studying
interactions when modeling educational program profiles.</p>
        <p>Both cardinal and ordinal measurement scales can be used to develop the described model. In
particular, experts can set weighting coefficients for the importance of all model components—
competencies, program outcomes, educational components, etc. The relative importance of these
model components can be measured using alternative ranking, pairwise or multiple comparisons, as
well as various methods of clustering results.</p>
        <p>Thanks to the described model, it is possible to determine the degree of similarity between
educational components, analyze the characteristics of program outcomes and other components of
the model proposed by the authors.</p>
        <p>The proposed model for formalizing educational programs is universal and can be applied to
various specialties and different levels of education. It is particularly relevant for fields associated
with rapid technological change, such as information technology.
7.2.</p>
      </sec>
      <sec id="sec-7-2">
        <title>Quality of educational trajectory</title>
        <p>The quality of the educational trajectory can be reflected on the basis of the documents submitted
by candidates for a vacant position.</p>
        <p>Heuristic H3. The educational trajectory of a candidate for a vacant position in an organization
adequately reflects the knowledge acquired and competencies mastered by the applicant as a result
of studying at higher education institutions.
(9)
(10)</p>
        <p>The quality of the educational trajectory can also be thoroughly and comprehensively examined
and confirmed based on testing candidates for a vacant position in the organization.</p>
      </sec>
    </sec>
    <sec id="sec-8">
      <title>8. Conclusions</title>
      <p>In this paper, the authors consider the current problems of automated candidate selection using a
documentary approach to assessing their competence. The problem statement and mathematical
model of automated candidate selection using a documentary approach are presented, and the
formalization of this problem is described.</p>
      <p>To supplement the information when constructing a mathematical model and studying the
relevant issues, additional heuristics are introduced, on the basis of which the logic of the study on
assessing the quality of the educational trajectory is provided.</p>
      <p>The structure and sequential representation of the stages of decision-making regarding the
recruitment and selection of candidates for positions are considered. The most popular approaches
to personnel assessment at the selection stage are also presented – assessment, testing, and ensuring
a sound procedure for formalized ranking.</p>
      <p>The paper describes the formalization of educational trajectory as the main formalism in modeling
the task of selecting candidates and reasonably determining their ratings.</p>
      <p>All approaches to modeling recruitment processes during personnel selection described in this
work are relevant, justified, and promising. The authors plan to continue the research described in
this work and apply new analysis tools.</p>
    </sec>
    <sec id="sec-9">
      <title>Declaration on Generative AI</title>
      <p>The author has not employed any Generative AI tools.</p>
    </sec>
  </body>
  <back>
    <ref-list>
      <ref id="ref1">
        <mixed-citation>
          [1]
          <string-name>
            <surname>Skrbinjek</surname>
            ,
            <given-names>V.</given-names>
          </string-name>
          &amp;
          <string-name>
            <surname>Dermol</surname>
            ,
            <given-names>V.</given-names>
          </string-name>
          (
          <year>2016</year>
          ).
          <article-title>Designing a Programme Profile: An Example of a Bachelor Business Study Programme</article-title>
          .
          <source>International Journal of Management</source>
          ,
          <article-title>Knowledge and Learning</article-title>
          . PP.
          <volume>123</volume>
          -
          <fpage>136</fpage>
          . https://www.researchgate.net/publication/308787379_Designing_
          <article-title>a_Programme_Profile_An_E xample_of_a_Bachelor_Business_Study_Programme#fullTextFileContent</article-title>
        </mixed-citation>
      </ref>
      <ref id="ref2">
        <mixed-citation>
          [2]
          <string-name>
            <surname>Alain de Janvry</surname>
            , Finan,
            <given-names>F.</given-names>
          </string-name>
          and
          <string-name>
            <surname>Sadoulet</surname>
            ,
            <given-names>E.</given-names>
          </string-name>
          (
          <year>2005</year>
          )
          <article-title>Using a Structural Model of Educational Choice to Improve Program Efficiency</article-title>
          . Working Paper.
          <volume>38</volume>
          p. https://gspp.berkeley.edu/assets/uploads/research/pdf/ProgresaWP-2005.pdf.
        </mixed-citation>
      </ref>
      <ref id="ref3">
        <mixed-citation>
          [3]
          <string-name>
            <surname>Rommel</surname>
            <given-names>AlAli</given-names>
          </string-name>
          , Ali Al-Barakat.
          <article-title>Using Structural Equation Modeling to Assess a Model for Measuring Creative Teaching Perceptions and Practices in Higher Education</article-title>
          .
          <source>Education Sciences</source>
          .
          <year>2022</year>
          ,
          <volume>12</volume>
          (
          <issue>10</issue>
          ),
          <volume>690</volume>
          . pp.
          <fpage>1</fpage>
          -
          <lpage>17</lpage>
          ; https://doi.org/10.3390/educsci12100690
        </mixed-citation>
      </ref>
      <ref id="ref4">
        <mixed-citation>
          [4]
          <string-name>
            <surname>Lagrange</surname>
            ,
            <given-names>J.-B.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Huincahue</surname>
            ,
            <given-names>J.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Psycharis</surname>
            ,
            <given-names>G.</given-names>
          </string-name>
          <article-title>Modeling in Education: New Perspectives Opened by the Theory of Mathematical Working Spaces</article-title>
          .
          <source>Mathematical Work in Educational Context</source>
          ,
          <volume>18</volume>
          , Springer International Publishing, pp.
          <fpage>247</fpage>
          -
          <lpage>266</lpage>
          ,
          <year>2022</year>
          ,
          <string-name>
            <surname>Mathematics</surname>
          </string-name>
          <article-title>Education in the Digital Era</article-title>
          ,
          <year>ff10</year>
          .
          <volume>1007</volume>
          /978-3-
          <fpage>030</fpage>
          -90850-8_
          <fpage>11</fpage>
          . https://hal.science/hal-03636706/document).
        </mixed-citation>
      </ref>
      <ref id="ref5">
        <mixed-citation>
          [5]
          <string-name>
            <surname>Mulesa</surname>
            ,
            <given-names>O.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Geche</surname>
            ,
            <given-names>F.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Batyuk</surname>
            ,
            <given-names>A.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Myronyuk</surname>
            ,
            <given-names>I.</given-names>
          </string-name>
          (
          <year>2018</year>
          )
          <article-title>Using a system approach in the process of the assessment problem analysis of the staff capacity within the health care institution</article-title>
          .
          <source>In: Computer science and information technologies (CSIT</source>
          <year>2018</year>
          ). PP.
          <volume>177</volume>
          -
          <fpage>180</fpage>
          .
          <fpage>10</fpage>
          .1109/STCCSIT.
          <year>2018</year>
          .
          <volume>8526749</volume>
          .
        </mixed-citation>
      </ref>
      <ref id="ref6">
        <mixed-citation>
          [6]
          <string-name>
            <surname>Tsyganok</surname>
            ,
            <given-names>V.</given-names>
          </string-name>
          <article-title>Providing sufficient strict individual rankings' consistency level while group decision-making with feedback Journal of</article-title>
          Modelling in Management,
          <year>2013</year>
          ,
          <volume>8</volume>
          (
          <issue>3</issue>
          ), pp.
          <fpage>339</fpage>
          -
          <lpage>347</lpage>
          .
        </mixed-citation>
      </ref>
      <ref id="ref7">
        <mixed-citation>
          [7]
          <string-name>
            <surname>Polska</surname>
            <given-names>O.V.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Kudermetov</surname>
            <given-names>R.K.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Zolotukhina</surname>
            <given-names>O.A.</given-names>
          </string-name>
          ,
          <string-name>
            <surname>Shkarupylo</surname>
            <given-names>V.V.</given-names>
          </string-name>
          (
          <year>2021</year>
          ).
          <article-title>A UML profile for quality-based web service selection using logic scoring of preference method</article-title>
          .
          <source>Telecommunication and information technologies. No</source>
          <volume>1</volume>
          (
          <issue>70</issue>
          ), PP.
          <fpage>65</fpage>
          -
          <lpage>78</lpage>
          .
        </mixed-citation>
      </ref>
    </ref-list>
  </back>
</article>