Monitoring the Development of University Scientific Schools in
University Knowledge Management
Gulnaz Zhomartkyzy and Tatyana Balova
Department of Information System, D. Serikbayev East Kazakhstan State Technical University, 69 Protozanov A.K.,
Ust-Kamenogorsk, Kazakhstan
Keywords: Knowledge Management System, Model of a Specialist, Monitoring of Scientific Schools, Clustering,
Scientific Community.
Abstract: This paper proposes a technological approach to university scientific knowledge management which
integrates the ontology based knowledge model and the methods of university scientific resource intellectual
processing. The process-oriented On-To-Knowledge methodology is used as the basis for university
scientific knowledge management. Some models and methods of university scientific knowledge
management have been studied. The developed model of a specialist that reflects the level of scientific
activity productivity and overall assessment of the employee's scientific activity has been described. A
specialist’s competence in knowledge areas is based on the processing of information resources. The
approach to the university scientific school identification based on the clustering of university academic
community common interests has been described.
1 INTRODUCTION
A large amount of accumulated information
resources and the high speed of new information
arrival impose increasingly high requirements to
modern systems designed to provide information
support to university scientific processes.
The intellectual capital or intangible assets of the
university are the source of new scientific
knowledge. At the level of organization scientific
knowledge or intellectual resources is a complex
category which combines intellectual capital, people,
and various forms of intangible assets which
concentrate knowledge and professional skills
(Klimov, 2002).
Russell Akkof, one of the classics of Operational
Research, proposed the following hierarchy of
knowledge: data - information - knowledge -
understanding - wisdom (Ackoff, 1989). There are
different approaches to the classification of
knowledge in organizations. The most common is
the division of knowledge into explicit and tacit
knowledge. The transformation of knowledge in an
organization occurs through explicit and tacit
knowledge interaction. The knowledge conversion
or transformation results in its qualitative and
quantitative increase.
The notion of knowledge management in an
organization is determined by the authors in (Zaim,
2007) as the strategy and the transformation of
knowledge.
The main focus in knowledge-intensive
organizations is on the creation, transfer and
development of knowledge, so effective knowledge
management is the matter of survival for such
organizations ( Miles, 2005; Scarso et al, 2010).
There arises a need in processes, infrastructure and
organizational procedures at a higher education
institution that would allow its employees to use its
corporate knowledge base.
This paper considers the problem of monitoring
the development of university scientific schools,
which is one of functional components of univertity
SKMS. The paper discusses some models and
methods of university scientific knowledge life cycle
support.
Section 2 describes the "University Scientific
Knowledge Management ". Section 3 describes "The
Monitoring of University Scientific School
Development." Section 4 describes "The
Development of a University Researcher Model",
section 5 presents "The Approach to the University
Scientific School Identification", Section 6 is the
conclusion.
222
Zhomartkyzy G. and Balova T..
Monitoring the Development of University Scientific Schools in University Knowledge Management.
DOI: 10.5220/0005464202220230
In Proceedings of the 17th International Conference on Enterprise Information Systems (ICEIS-2015), pages 222-230
ISBN: 978-989-758-097-0
Copyright
c
2015 SCITEPRESS (Science and Technology Publications, Lda.)
2 THE STRUCTURE OF
UNIVERSITY SCIENTIFIC
KNOWLEDGE MANAGEMENT
SYSTEM
In this paper the university’s scientific knowledge
management system (SKMS) is considered as an
aggregate of information, software, technical means,
and organizational solutions aimed at efficient
management of the university's available intellectual
resources and training specialists who meet the
modern requirements.
According to the above definition, university
scientific knowledge management (SKM) can be
understood as:
the aggregate of processes associated with the
creation, distribution, processing and use of
university scientific knowledge;
an established systematic process of working
with information resources and knowledge,
scientists and specialists in certain areas in
order to facilitate access to knowledge and re-
use them with modern information technology
at the university.
The purpose of university SKMS is the
formation of a unique ontology-based integrated
intellectual environment to improve the
competitiveness of the university's science and
education. The university SKMS is the technological
component of the university SKM, which provides
the creation, organization and dissemination of
scientific knowledge among the university staff.
There are following approaches to knowledge
management: organizational and technological
(Tuzovskiy, 2007). The technological approach puts
the application of IT-technologies in line with the
organizational measures. The model of technological
approach to knowledge management is shown in
Figure 1.
The process-oriented On-To-Knowledge
methodology is used as the basis for university
scientific knowledge management (Sveiby,1989;
Staab et al, 2001). The methodology of KMS
development and support is based on the process and
metaprocess of working with knowledge
(KnowledgeMetaProcess and KnowledgeProcess).
The basis of the metaprocess of working with
knowledge (KnowledgeMetaProcess) is the
development of an ontology, which consists of the
following steps: a feasibility study, the beginning,
clarification, evaluation, support and evolution.
The ontology is the link/(linking element) of
knowledge objects and a connecting bridge between
different steps of knowledge transformation
processes (KnowledgeProcesses). The development
of the ontology is the important aspect of knowledge
management solution support. The development and
deployment of applications of knowledge
management takes into account the requirements of
"KnowledgeProcess" and considers such processes /
issues as:
metaprocess of working with knowledge
(KnowledgeMetaProcess);
software engineering (software development
and design– Software engineering);
the corporate culture of the organization.
The process of working with knowledge
(Knowledge Process) focuses on the use of KM-
solutions, i.e. after KM-application are fully realized
and implemented in the organization, the cycle of
knowledge transformation is performed. The
knowledge transformation cycle consists of the
following steps: creation, storage, search and access,
use.
The developed model of technological approach
to knowledge management based on the
methodology described above is shown in Figure 1.
Figure 1: Technological approach to university scientific
knowledge management.
The proposed technological approach integrates a
functional component and knowledge management
tools. The functional component includes:
classification of information resources,
a scientist’s/a specialist’s scientific profile
formation and saving;
MonitoringtheDevelopmentofUniversityScientificSchoolsinUniversityKnowledgeManagement
223
identification of university scientific schools
and research directions;
ensuring the availability, search and
navigation.
Knowledge management tools are:
the ontology based university scientific
knowledge model,
the procedure of university information
resources processing,
semantic queries,
semantic Portal of university scientific
knowledge.
The following sections describe the
implementation of the task of monitoring scientific
school development in university knowledge
management.
3 MONITORING THE
DEVELOPMENT OF
UNIVERSITY SCIENTIFIC
SCHOOLS
The priority task of integrating research and
education is the development of scientific schools
which must be the main result of fundamental
science and education interaction. Introduction to the
research undertaken by scientists of scientific
schools is the best school for young people.
Scientific schools provide constant growth of
qualification of their participants; the presence of
several generations in bundles of "teacher-student"
ensures the continuity of generations (
NC STI RK).
The development of scientific schools and scientific
and pedagogical teams is the basis for the
development of fundamental scientific research and
training quality improvement of research and
educational personnel.
Scientific schools form that dynamic unit of
science which ensures the continuity of scientific
knowledge and creates optimal conditions for its
development. Scientific school is the key element of
collective preservation and multiplication of
knowledge, one of the conditions to maintain the
quality of research, and hence the quality of training
scientific personnel. Scientific school is a clearly
defined direction of scientific research carried out in
the framework of specific scientific specialties
(Trubina and Zabelina, 2011).
Identification of scientific schools is becoming
increasingly important in recent years in connection
with the development of mechanisms for
organization effectiveness assessment in tenders for
financing projects, their certification and
accreditation. The availability of scientific schools is
one of the most important criteria for foreign
scientific funds which conclude contracts on joint
research and grants as well as the criteria taken into
account in establishing the rating of organizations.
One of the qualitative characteristics of a
particular scientific direction’s overall development
and potential is the state of scientific schools. The
creation, reorganization and coordination of
scientific schools are regulated by universities.
Monitoring the development of scientific schools
remains a major issue in university scientific and
innovation activity management. Thus,
identification, recording, development, and
monitoring the development of scientific schools is
one of the priorities of science and education.
We have studied and developed a model of
scientific community and the method of its
intelligent processing to implement the functions of
monitoring. The overall structure of monitoring the
development of university scientific schools is
presented in Figure 2.
Figure 2: The main functions of monitoring the
development of university scientific schools.
The scheme of monitoring shows the following
functions: determining the researcher's competence
level, identification of scientific schools and
research directions.
The following sections describe the model of a
specialist which reflects the level of scientific
activity productivity based on the calculation of
entropy and overall scientific activity evaluation; as
well as the approach to the university scientific
school identification based on clustering the
university scientific community for common
interests (university scientific schools).
Monitoring the development of university
scientific schools
1 Determining the
scientific activity level
of a researcher
2 Identification of
scientific schools and
directions
ICEIS2015-17thInternationalConferenceonEnterpriseInformationSystems
224
4 THE DEVELOPMENT OF A
UNIVERSITY RESEARCHER
MODEL
One approach to human capital management is to
develop a model of a university researcher (the
model of a specialist).
Currently, there are two ways to create and
support the model of a researcher: by a survey
(qualification audit in an organization) and by
monitoring their work in the knowledge
management system (scientific papers, projects)
(Tuzovskiy, 2007). The paper supports the definition
given in Tuzovskiy , 2007), where the model of a
researcher refers to a sound set of interrelated
properties of a specialist, which can be formally
described and used to support the efficient work
with implicit knowledge.
In the scientific knowledge ontology the model
of a specialist has the following formal description
which includes a set of contextual and content
metadata:
M
M
,M
(1)
where: M
– is contextual metadata of a
specialist description; M
– is content
metadata, which describe the specialist’s
competence.
Contextual metadata
of a specialist
include such parameters as:
identification (a name, a photo, the date of
birth, the place of birth, a login, a password);
contact information (postal and email
addresses, a personal web page, phone
numbers;
education (diplomas, certificates, etc.);
professional achievements (prizes in
competitions, awards, medals, etc.).
Content metadata M
provide the
description of the specialist's competence as a set of
his competence characteristics:
M
C
,C
,
(2)
where:
– the competence of a specialist in
fields of knowledge relevant to rubrics which are
described as classes in the scientific knowledge
ontology О
.
– a measure of specialist’s scientific activity
efficiency (the level of the specialist's competence
dispersion);
- the overall assessment of scientific work.
The model of an individual researcher's scientific
activity is determined by the factors of scientific
activity (Figure 3).
In ontological information model these factors
are grouped into the following classes: Event,
Project, Publication.
Figure 3: Classes of the information model used to
simulate a university researcher’s activities.
4.1 A Specialist's Competence in Areas
of Knowledge based on the
Classifications of Information
Resources
Automatic processing of scientific electronic
resources by the methods of TekstMining text
processing are required to implement the described
above model.
A specialist's competence in areas of knowledge
is formed on the basis of the specialist’s
scientific profile. The specialist’s scientific profile is
based on the classification factors of his scientific
activity (publications) by scientific areas
(Zhomartkyzy, 2014).
The specialist’s scientific profile is determined
as the profile of all his publications: the profiles of
documents are formed by processing the university
scientific resources. The profile of a document is
determined as the vector of all its relevant classes:
,…,
(3)
where:
с
are relevant classes .
Accordingly, the specialist's competence in areas
of knowledge
, relevant to the specialist’s
scientific profile is determined as the profile of all
his publications.
MonitoringtheDevelopmentofUniversityScientificSchoolsinUniversityKnowledgeManagement
225
C
,…,
(4)
where
are all the author's documents.
The final step of text classification is the
formation of the document's semantic profile by
creating the class of "Information Resources"
individuals of scientific activity ontology.
4.2 The Calculation of the University
Specialist’s Scientific Activity
Efficiency
To assess the university specialist’s scientific
activities efficiency we suggest using the level of
dispersion of his competence.
The model of an individual researcher’s
scientific activities is determined by the factors of
scientific activities. These factors are grouped into
ontology classes in the model of ontology.
The connection between the Person class and the
class factor of scientific activities is shown below:
≡∩∃.
≡ ∩∀publHasDivis.
where, – is persons, – иis information
resources,
is the field of knowledge..
Each researcher works in at least one field of
knowledge (VINITI rubricator, VINITI - All-
Russian Institute of Scientific and Technical
Information). Therefore, the classification of
scientific activity factors is carried out by means of
the VINITI rubricator of fields of knowledge up to
level 3.
Cybernetics →Artificial Intelligence →
Knowledge engineering
Cybernetics → Artificial Intelligence → Expert
systems
Cybernetics→ Theory of modeling →
Mathematic modeling
This paper proposes the method for calculating
the efficiency index of a specialist’s scientific
activities C
to analyze the competence of
employees in a particular field. A specialist’s
scientific activity efficiency C
is calculated using
the entropy.
The more papers of a specialist (researcher) are
grouped by a certain category, the lower entropy and
the higher the specialist’s scientific activity
efficiency are (Adamic et al, 2008), (Baesso et al,
2014). A specialist who has a high entropy works, as
a rule, in several fields of knowledge, i.e., the
specialist has a lower scientific activity efficiency
(scientific competence):
с
с
(5)
(6)
– is the number of the researcher's papers by
heading 1,
;
- the total number of the researcher's papers.
An example of calculations using formula 3 is
shown in Table 1.
Table 1: A Specialist’s Scientific Activity Efficiency
Calculation.
A
researcher
Fields of
knowledge
The
number of
scientific
resources
The
efficiency
index of a
specialist’
s scientific
activities
(
)
specialist
1.
Physics of
Atom and
Molecule
1
1,68
General Physics 3
Solid State
Physics (nano-
sized objects,
the structure of
solids, general
issues of Solid
State Physics)
5
Physics of
Gases and
Liquids
1
specialist
2
Solid State
Physics
5 0
specialist
3
General Physics
6 0
specialist
4
Nuclear Physics 1
0,91
General Physics 2
specialist
5
General Physics 7
0,54
Physics of
Gases and
Liquids
1
Threshold values of the specialist’s scientific
activity efficiency
were determined empirically:
0
1 - a high level;
1
2 - a medium level;
2 - a low level.
The analysis of results of personal calculation
for leading university scientists by knowledge
areas "General Physics", "Physics of Solids",
"Physics of Atoms and Molecules," "Physics of
Gases and Liquids" and "Nuclear Physics" confirms
the applicability of the formula for calculating the
entropy of the researcher’s scientific competence.
ICEIS2015-17thInternationalConferenceonEnterpriseInformationSystems
226
4.3 The Calculation of Total
Assessment of the University's
Specialist’s Scientific Activity
The specialist's scientific activity efficiency is a
quantitative indicator of knowledge, skills and
abilities in a scientific field of a corresponding
specialty.
The qualitative analysis, i.e. the definition of a
specialist’s professional competence, is also needed
for making management decisions in the context of
the university’s different departments.
Each researcher has a trajectory of educational
and scientific activities (Figure 4): scientific
activities, educational activities, participation in
competitions and grant projects, international
mobility.
This scheme allows us to calculate the qualitative
characteristics of the overall assessment of the
specialist’s scientific activity.
Figure 4: Directions of scientific activities at the
university.
The calculation of a specialist’s overall scientific
activity evaluation C
can be shown
schematically as follows (Figure 5).
A university specialist with a high overall
assessment of scientific activity:
Works in all areas of scientific activities;
Has a high index of scientific competence C
.
A university specialist with a medium overall
assessment of scientific activity:
Only develops educational courses or is only
involved in projects;
Has a medium index of scientific competence
C
.
A university specialist with a low overall
assessment of scientific activity:
Only develops educational courses;
Has a low index of scientific competence C
.
Figure 5: The scheme of the calculation of a specialist’s
overall scientific activity evaluation.
5 THE APPROACH TO THE
UNIVERSITY SCIENTIFIC
SCHOOL IDENTIFICATION
BASED ON THE UNIVERSITY
COMMUNITY
CLUSTERIZATION BY
COMMON INTERESTS
Each scientific school or scientific direction forms a
scientific community by interests and develops in
accordance with some specific rubrics of knowledge
areas (Cantador and Castells, 2011). The VINITI
rubricator of knowledge areas is used as the
rubricator. In the proposed approach, a model of the
scientific community is described as follows:
.
,
.
,
.
,
(7)
where
are rubrics corresponding to specific
areas of science and technology (one subrubric may
be in several scientific fields).
The model of a scientific community is shown in
Figure 6.
The proposed approach requires to carry out the
university scientific community clustering based on
its memebers’ common interests to identify
scientific schools and research directions.
To identify scientific schools and research
directions DBSCAN clustering method is used in the
scientific community model.
The principal advantages of this method served
as the basis of the choice of DBSCAN density
clustering method:
A specialist publishes scientific
articles, monographs, books
A specialist develops
educational resources (lecture
notes, tutorials, guidelines)
A specialist participates in
projects
Educational and scientific activity of the
university’s specialist
(directions of activities at the university)
International mobility of a
specialist
MonitoringtheDevelopmentofUniversityScientificSchoolsinUniversityKnowledgeManagement
227
Identification of the number of clusters (based
on the notion of point density);
The clustering algorithm is able to detect
clusters of different shapes;
Resistance to noise objects.
The idea underlying the algorithm is that within each
cluster there is a typical dot density (of objects),
which is significantly higher than the density outside
the cluster (Figure 7). The density in the areas with
noise is lower than the density of any of the clusters.
For each dot of the cluster its neighborhood of a
given radius must be at least a certain number of
points, this number of dots is specified by a
threshold value. (Bolshakova et al, 2011),
(Marmanis et al, 2011).
Figure 6: A university scientific community by interests.
Figure 7: Example of a cluster of arbitrary shape.
In Fig.7 А is a core point. В, С are border
points.Cluster
is not the empty subset of objects
satisfying the following conditions, at givenand
, where Eps is the maximum distance
between adjacent points, MinPt is the minimum
number of neighboring points:
∀,: if
is density-connected
from,then
, at given Eps and MinPt;
∀,
:isdensityconnectedwith, at
given Eps и MinPt.
Figure 8 is given below for detailed description
of the points.
Thus, a cluster is a set of closely-related points.
Each cluster contains at least MinPt of documents.
To perform clustering the model of a scientific
community is translated into a binary matrix (Figure
9). The values of matrix elements correspond to the
presence or absence of work on the appropriate
rubric.
a) density-reachable points
b) density-reachable points
and density-connected
points of the class
Figure 8: Types of points that form classes in DBSCAN
algorithm.
Members of the scientific
community
Rubrics
0 0 0 0 0 1 0 0 1
1 0 0 0 0 1 0 0 1
1 0 0 0 0 1 0 0 1
1 0 1 0 0 0 1 0 1
1 0 0 0 0 1 0 1 1
0 0 0 0 0 1 0 0 1
1 0 0 0 0 1 0 0 1
Figure 9: The matrix of the scientific community model
description.
The clustering algorithm based on the density of
points is described below (Bolshakova et al, 2011).
Input: a set of objects Q, parameters - Eps
(the distance between the objects of the
class),.
1. Determination of directly density-
reachable points:
∈
,|
|
where, q is a core point, p is a border point.
Point p is directly density-reachable from
point q,
2. Determination of all density-reachable,
density-connected points of the current class:
←
←
←
∈
,|
|
Output: a set of clusters
Noise is a subset of objects that do not belong
to any cluster,
∈|∀
∉,
1,
|
|
.
ICEIS2015-17thInternationalConferenceonEnterpriseInformationSystems
228
The value of Eps is determined as the distance
between the researchers’ profiles. A researcher's
profile consists of all its relevant rubrics and is
presented as a vector. VINITI rubricator of
knowledge areas is used as a rubricator. The set of
vectors forms a matrix of researchers's profiles . To
calculate the distance a cosine measure of adjacency
is used. The value MinPt is the minimum number of
the subjects of scientific school, i.e. the subjects of
"the communities of interest" in the model of a
scientific community.
For approbation of the proposed approach we
chose scientific communities of D. Serikbayev
EKSTU and Ioffe Physical-Technical Institute of the
Russian Academy of Science (Ioffe Institute). Papers
and research adirections of their scientifc
communities were examined. The results of
numerical experiments confirmed the efficiency of
the clustering algorithm used.
6 CONCLUSIONS
This paper describes the realization of monitoring
the development of university scientific schools,
which is one of functional components of the
technological approach to university scientific
knowledge management. Some models, methods,
and technologies of university scientific knowledge
life cycle support processes are considered.
The paper describes the developed model of a
specialist which reflects the level of scientific
activity productivity based on the calculation of
entropy and overall scientific activity evaluation.
The approach to identification of university
scientific schools based on the clustering of
university scientific community by common
interests is proposed.
The next stage of this work is to address the
problem of assessment of university scientific
activities and the degree of its integration with
educational process.
The work was performed under grant "The
development of an e-university's ontological
knowledge base”, state registration number
0213RK00305
REFERENCES
Klimov, S. M., 2002. Intellectual resources of society.
SPb: IFEREL, Znanie.
Ackoff, R.L., 1989. From Data to Wisdom. In Journal of
Applied Systems Analysis, vol. 16. pp. 3-9.
Zaim H., 2007. Performance of Knowledge Management
Practices: a causal analysis. Knowledge Management,
vol. 11(6), pp. 54-67. url:http://dx.doi.org/
10.1108/13673270710832163.
Miles, I., 2005. Knowledge intensive business services:
Prospects and policies. Emerald Group Publishing
Limited, vol. 7(6). pp. 39-63.
DOI 10.1108/14636680510630939.
Scarso, E., Bolisani, E., 2010. Knowledge-based strategies
for knowledge intensive business services: A multiple
case-study of computer service companies . Electronic
Journal of Knowledge Management. url: www.ejkm
com, vol. 8(1). pp. 151-160.
Tuzovskiy, A. F., 2007. The development of knowledge
management systems based on a single ontological
knowledge base. In Bulletin of the Tomsk Polytechnic
University, vol2(310), pp. 182-185.
Sveiby, Karl-Erik., 1989. The Invisible Balance Sheet.
Stockholm, 138 p.
Staab S., Schunurr H-P., Studer R., Sure Y., 2001.
Knowledge processes and ontologies. IEEE Intelligent
Systems, Special Issue on Knowledge Management,
vol. 16(1), pp. 26-34.
NC STI RK. Scientific schools and priorities for the
development of the country. URL:
http://exclusive.kz/bez-rubriki/22068.
Trubina, I. O., Zabelina, I.N., 2011. Creating personnel's
positive motivation in higher education institutions in
the process of education and development of scientific
schools. In Creative Economy, vol. 1(49), pp. 30-36.
Tuzovskiy, A. F., 2007. Creating and using a knowledge
base of specialists' competence profiles at
organizations. In Bulletin of the Tomsk Polytechnic
University, vol. 310(2), pp. 186-189.
Zhomartkyzy, G., Balova, T., Milosz, M., 2014.
Information Models and Methods of the University's
Scientific Knowledge Life Cycle Support. Proceedings
of the 2014 Federated Conference on Computer
Science and Information Systems, pp. 279 – 285.
All-Russian Institute of Scientific and Technical
Information, url:
http://scs.viniti.ru/rubtree/main.aspx?tree=RV.
Adamic, L., Zhang, J., Bakshy, E., Ackerman, M.S., 2008.
Knowledge sharing and yahoo answers: everyone
knows something. Proceedings of the 17th
international conference on World Wide Web. pp. 21-
25.
Baesso, P. T.,Wolfgand, Matsui Siqueira S. and Cristina
Vasconcelos de Andrade L., 2014. Finding Reliable
People in Online Communities of Questions and
Answers.-AnalysisofMetricsandScopeReduction. In
Proceedings of the 16th International Conference on
Enterprise Information Systems, pp. 526-535, DOI:
10.5220/0004954005260535.
Cantador, I., Castells P., 2011. Extracting multilayered
Communities of Interest from semantic user profiles:
Application to group modeling and hybrid
recommendations . Computers in Human Behavior,
vol. 27(4), pp. 1321-1336.
Bolshakova, E. I., Klyshinsky, E. S., Lande, D. V.,
MonitoringtheDevelopmentofUniversityScientificSchoolsinUniversityKnowledgeManagement
229
Noskov, A. A,. Peskova, O. V., Yagunova, E. V.,
2011. Automatic processing of natural language texts
and computational linguistics: Textbooks. - M.:
MIEM, 272 p.
Marmanis, H., Babenko, D., 2011. Algorithms of
Intellectual Internet. Best practices for collecting,
analyzing and processing data. Trans. from English. -
SPb .: Symbol-Plus, 480 p.
Ioffe Physical-Technical Institute url: http://www.ioffe.ru/
ICEIS2015-17thInternationalConferenceonEnterpriseInformationSystems
230
