CC BY
24
Прикладная информатика
UNIVERSITY KNOWLEDGE DOMAIN APPLICATION FOR EDUCATIONAL CONTENT UPDATING
УДК 338.46
Nikolai Nikolaevich Gorbachev
Head of IT Department, Minsk Branch of MESI Tel.: +375-17-328-12-86, E-mail: ngorbachev@mfmesi. ru
Anatolii Solomonovich Greenberg
Doctor of science, technical; Professor, chair of mathematics and information science, Minsk Branch of MESI Tel.: +375-17-291-45-56, E-mail: [email protected]
Method of permanent updating of educational informational resources based on formalized university knowledge domain ontology is suggested. The method is applied as superstructure for settling the economic tasks of optimization of educational content updating.
Keywords: information economy, domain knowledge ontology, e-learning, knowledge management, educational information resources, economic-mathematical modeling.
Николай Николаевич Горбачёв
Начальник отдела ИТ Минского филиала МЭСИ
Тел.: +375-17-328-12-86, E-mail: ngorbachev@mfmesi. ru
Анатолий Соломонович Гринберг
д.т.н., профессор, Кафедра математики и информатики, Минский филиал МЭСИ Тел.: +375-17-291-45-56, E-mail: [email protected]
ИСПОЛЬЗОВАНИЕ МОДЕЛИ
ПРЕДМЕТНОЙ ОБЛАСТИ ДЛЯ УПРАВЛЕНИЯ КОНТЕНТОМ ВУЗА
Предлагается метод регулярного обновления образовательных
информационных ресурсов ВУЗа, основанный на использовании формализованной модели предметной области ВУЗа в виде онтологии. Метод применен в качестве структурной основы для решении экономических задач оптимизации обновлений
образовательного контента.
Ключевые слова: информационная экономика, модель онтологии, электронное обучение, управление знаниями, образовательные информационные ресурсы, экономико-математические модели.
1. Introduction
Implementation of contemporary research and practice results into educational process is fundamental condition for positive competition of the university at the educational market. Instant updating of content becomes an important issue of capital investment. Since 2000, volume of scientific knowledge doubles every two years period [1] ; in 2012 volumes of technical information for some sectors of the economy will double every 18 months.
Using electronic educational informational resources instead of hard copies speeds-up the revision. However, expenses on technical support, development and content updating are 85% of total amount of knowledge management system (KMS) maintenance expenses. However, investments into development of new content and e-learning technologies results only 15% of KMS maintenance costs [2].
Classification and verification of incoming data and information requires development of specific KMS instruments. According to valid Russian norms, updated content should develop the competences listed in State Educational Standards (GOS), revising of which takes considerable time and makes some of the competences out-of-date at the moment of GOSpublication. Problem of content updating specifically for university means balancing between the state requirements and current research and practice results available to students from external knowledge sources.
The task of this article is development of knowledge management methods and tools allowing for effective updating of educational content using requirements of the State (using SES) and employers (current research and practice results).
2. Educational Content as Object of Management
University content development is an instant, meaning that current students and lecturers receive the educational information used by predecessors and deliver it to the followers. Within the educational process current generation defines the discrepancies between the current research and available content, discovers the prospective information sources to be used in future. According to [3],perspective information sources may be discovered using two approaches:
-Project approach, which considers discovery and implementation of current research results into university educational process for rational and economically effective content updating using external knowledge sources. It is effective for updating the content for special syllabus disciplines.
-Distributed resources approach that utilizes the advantages of grid system of content actualization and requires verification and systematization of updated data and information obtained of various students and lecturers. This method becomes effective in case of considerable volume of correspondents and application of current data processing tools.
Combination of both approaches involves the results of personal and collaborative work with external knowledge sources into development of prospective information sources of the university. Such combination may be proceeded using dual control model that allows for transferring the content to optimal condition in the process of university knowledge domain research. This idea is confirmed with research of Morrison and Weiser [4] that also motivated application of dual control to KMS.
DeLone and McLean [5] suggested a model that divides personal and organizational contribution to knowledge development and updating. Organizational (university) contribution means systematization and verification of information resources applied in educational process. Such systematization and verification is limited with university knowledge domain, which may be formalized in a form of ontology.
Suggested structure of university educational content management system is described on Figure 1. Information resources (IRs) of the university include traditional and electronic educational IRs. Electronic IRs are real-time updatable. KMS users (lecturers, students) working with external knowledge sources and defining correspondence between university IRs and current research and practice results. In case of discrepancy they attract the new information to university prospective information sources repository. Knowledge domain in a form of ontology guides the KMS users in course of studying the external knowledge sources and allows for classification, verification and systematization of knowledge incoming to university prospective infor-
Figure 1. Structure of university educational content management system
mation sources repository.
3. University Knowledge Domain Modeling
Knowledge domain modeling consists of two stages:
- Mathematical modeling of university knowledge domain using methods of graph theory
- Implementation of mathematic model to ontology building using modeling software
Graph ontology model of university knowledge domain
On preliminary stage of modeling major taxonomy objects and relations between them were identified. Official Russian educational standards (GOS) and university syllabus were taken as basic source of information. Following the analysis, graphs of taxonomy substances and relations were identified for describing the relations between the ontology elements.
Since taxonomy graphs generation is possible for both ontology and single substances, the following sets of graphs were described:
- set of educational products (specialties) EP = {EP,,..., EP}
- set of syllabus SB = {SB,,..., SB}
- set of didactic units (GOS official descriptions of syllabus disciplines) DU = {DU,,..., DU}
- graph Rel of describing relations between elements of set DU = {DU DU}, including non-hierarchic relations between nodes of graph
- graph Rel of describing relations between elements of set TEZ = {TEZ,,...,
TEZ }, including non-hierarchic relations between nodes of graph for concretiza-tion of didactic units at the level of thesaurus within university knowledge domain.
University ontology may be recorded as OEDU = {Def, Atr, Rel, Func}, where:
f Op Osb, ODU where Def = {defr..,def } - is finite set of ontology definitions, where:
OEP -ontology of educational products (specialties) provided by the university;
0SB - ontology of syllabus of educational products (specialties) of university;
ODU - ontology based on didactic units (GOS official descriptions of syllabus disciplines);
0TEZ - ontology based on thesaurus developed by tutors (experts) forming keywords for concretization of didactic units.
Atr = {atr,., atr} - is finite set of attributes of Def, required for formation of ontology;
Rel = {rel,., rel} - set of relations between definitions adjusting relations between them;
Func = {func ..., func} - functions defining dependencies between the definitions.
Implementation graph model to domain knowledge ontology building
Actual university domain ontology development for economics specialties was proceeded in the following way. Upon GOS analysis for economics specialties
(OGOS) elements of a sequence of disciplines urn:Disciplines were identified and divided into subclasses depending of educational levels and types of disciplines (sets F.00 - federal obligatory sets of disciplines; R.00 - regional sets of disciplines, V.00 - optional disciplines). Educational products provided by the university may be graded according to levels of education. Attributes, relations and functions are described with four-levels axioms (class, subclass, annotation and individual axioms).
Visually, part of university domain knowledge ontology, build upon analysis of GOS forms an oriented planar graph. In course of analysis and formalization only part of GOS limited with economics domain knowledge was used (official Russian Federation standards of higher professional education for economics specialties). However, developed approach to GOS ontology OGOS formation allows for extending ontology for other specialties.
Upon analysis of standard syllabus located at official educational portal http:/ /www.edu.ru classes of didactic units (forming major definitions describing topics of disciplines) were identified (class urn:DidacticUnits {urn:DE0001, ... DE0924}). Detailed elaboration of domain knowledge is continued in university ontology (Oedu).
Formalized elements of GOS ontology
OGOS forms the basis for university domain knowledge ontology OEDU. University educational products limited ontology (university specialties with educational
Figure 2. University domain knowledge ontology architecture
that case, SN = {1,..., GJ- value of tez
levels) OEP has intersection with sets weighting coefficients of indicators. In urn:Disciplines, urn:Qualifications and urn: Specialities of OGOS. From elements of set urn:Disciplines OGOS a limited ontology OEP is formed for proceeding the educational process for university speciali-
ties O Using sets Atr = {atr
atr},
indicator, and GN - power of set of value indicator, and N = {1,..., R} - number of TEZ value.
Values of indicators form the space containing the objects:
Rel = {rel,..., rel} and Func = {func,... func} various relations between disciplines may be established (equivalence, incidence, overlapping etc.). Final description of syllabus disciplines is made with thesaurus limited ontology OTEZ detailing the limited ontology of didactic units (ODU). Developed functions of university domain knowledge ontology OEDU allows to content management using technologies on the principle of object-oriented principle of multiple application (reusable learning objects - RLO).
Elements of set OTEZ forms set of re-
TEZ
lated content metadata for utilization in KMS META = {META^yy...,MET^}, basing on which final metadata required for indexing are generated for RLO subset RLO = {obj, ..., objf ..., obj}, where i = k~n . Objects of RLO set is described with j indicators belonging to TEZ. Each indicator {TEZ1..., TEZ}is characterized with certain values of , corre-
sponding to metadata, being formed basing on definitions of OTEZ, and may be either numerical or qualitative; numerical metadata may be used for defining the
TEZj = META,..., METAN
TEZ,,
META*
=METAy..
Object of RLO set (obj) may be described as a point META. iny-dimension-al space, formed in the result of multiplication of sets TEZ = TEZ, x ... x TEZ.
1 y
Function from y arguments of TEZ
(META yy..., META N1) is collated with each RLO object. In case if value indicators of some objects (obji) is equal, then point in space META will correspond to some elements of tez.
Suggested architecture of university domain knowledge ontology is indicated on Figure 2. In general, proposed structure of domain knowledge ontology adapted for e-EIR management is illustrated on diagram 2.
University builds its knowledge domain ontology OEDU (acting as meta-on-tology in hierarchy) basing on state educational standards O„„_, which also defines university domain knowledge as licensed grades and specialties OOP. Minimal requirements to syllabus (OSB) and
major topics of disciplines (didactic units - ODE) are also defined with OGOS. For de-
DE GOS
scription of major topics of disciplines ODE university develops thesaurus (OTEZ). Relations, functions and attributes of ontology definitions form hierarchic and non-hierarchic associations.
Final purpose of ontology is management of IR and IS as RLOs using object-oriented technologies for multiple publication of updated content. Domain knowledge ontology acts as superstructure of content management tools within KMS toolset. Regarding content actualization and development at the level of university educational portal, ontology allows:
-effective actualization of multiple IRs using prospective IS, providing labor economy and decreasing the period of actualization;
-development of official and non-official IRs of the university within the domain knowledge, increasing market costs of intangible assets.
4. Integration of University Knowledge Domain Ontology into Knowledge Management Tools
University knowledge domain ontology may be integrated into KMS on four levels. Ontology has superstructure role forming the basis for educational content management using metadata, providing: - interdisciplinary integration at the levels of didactic units and thesaurus (ar-
Figure 3. Suggested architecture of university KMS based on domain knowledge
rangement of cross-references, managing e-EIRs related with various disciplines etc.),
- forming the basis for individual and collaborative work of KMS users with external information sources forming prospective IS and discovering discrepancy in current IR
- actualization of existing content, managing additional IRs.
Application of domain knowledge ontology for content management allows for interdisciplinary integration at the level of thesaurus, effective content indexing, analysis of available e-EIR and educational content (tests, e-books etc.) on the subject of correspondence with educational standards, actualization and development of KMS content in the result of users' activity.
Suggested architecture of university KMS based on domain knowledge is mentioned on figure 3. On the level of information domain knowledge ontology acts as preparation tool for RLO-based content of IR and IS for multiple application in electronic courses. On knowledge level domain knowledge ontology acts as modeling tool for effective content indexing. On processes level domain knowledge ontology is used as integration tool performing interdisciplinary integration on
the levels of ODE and OTEZ. Finally, on interface level domain knowledge ontology acts as source for content updating and development tools of semantic web-portal.
Objective function of developed knowledge domain ontology and suggested four-level KMS ontology-based toolset is efficient updating of educational content and development of information resources of the university. Application of content updating and development methods results the following effects:
- changes in structure of labor expenditures related with content processing;
- optimization of content updating periodicity;
- optimization of content updating technology steps;
- optimization of updated content publication using metadata.
5. Economic Application of Ontology-based Knowledge Management Tools
Application of suggested domain knowledge ontology as superstructure to educational content updating within KMS collaboration work toolset allows for development of various mathematical optimization models. Implementation of such models in the university causes decreasing of content updating costs, defining best possible periodicity of content up-
dating, increasing content effectiveness using object-oriented principle of multiple publication.
Since KMS is software complex, regulations in field of information technologies are used for content updating valuation. The following levels of educational content maintenance are suggested referring to [6]:
- prophylactic maintenance of educational content: updating adding necessary revisions into existing content (insignificant corrections due to data or information actualization) ;
- corrective maintenance of educational content: modification of existing content for providing the correspondence with external knowledge sources (corrections due to availability of new results of research for actuality);
- adaptive maintenance of educational content: revising (modification) providing content actuality due to changed or revised results of research, which allows for modification of existing data and information;
- complete maintenance of educational content: considerable modification of content due to availability of considerable amount of changed external knowledge sources allowing for complete revision of existing educational content.
Figure 4. Graphic view of economic-mathematical model of content updating periodicity task
This gradations is used for improving the calculation of labour expenses for evaluation of economic effectiveness of developed KMS content updating tools and optimization economic-mathematical models development.
Economic-mathematical model of content updating periodicity
A0 - price of new content element;
Acm (t ) - depreciated cost of content element (in Russian Rubles) without performed updating at the time moment t;
Aj (t) - volume of educational services provided using j element of content at the time moment t (in Russian Rubles);
C6 (t ) - regulatory costs of content updating, each type of updating procedure may be applied severally within the current period, for example, prophylactic content updating may be applied several time within one year;
k = 1, 2, ..., n - stages (years) of content use;
t = t, t2,...t5 - content maintenance periods in correspondence with levels of educational content maintenance; after
the level 5 content shall be proceeded with complete maintenance.
University provides educational services usingj element of content for n-year period with basic price A0. With changing t time parameter quality of content is decreasing, Aj - amount of educational services provided is decreasing. Difficulty and expenses of content actualization -C6 - increase. The task of model is to define the stages of optimal content actualization maximizing the volume of provided educational services. Stages of content maintenance k are defined with sequence number of educational semester k = 1, 2, ..., n. Possible variants of decisions at k-stage are alternatives: continue updating existing content or move current content to retrospective IS and use new content.
Condition of content at k-stage is period of content maintenance t={t1,...t5} at the beginning of k-semester. ffn(t) - maximal volume of educational services provided by the university for stages k1, k2, ..., kn, considering that at the beginning
of k-semester content developed (purchased) /-years ago is available to the university. Graphically the task is presented on Figure 4.
If Aj (t) - volume of educational services provided applyingj-element of content being used for /-years, and - regulated expenses for content element actualization. Decisions on continuation of content use are taken in the beginning of k-semester on the basis of evaluation. In the beginning of semester one of the following action shall be taken: RLOUpdate -continue updating existing content or RLORetr - move current content to retrospective IS and use new content.
Optimal action 3-OHPopt means aggregation of decisions: RLO°pt,RLO°ptRLO°p[, in the result of which the systems condition transfers from condition kto final condition. Price of purchasing new content element is equal to A0 and means its basic accounting value as intangible asset. If Acm (t) is residual value of content element which was used in KMS for t years, recurrent
Прикладная информатика
equation according to Bellman function [7] shall be written as:
' A (')-YC6(f)+fk+1(t+A)
RLOop' = RLOUpdate
fn (t) = max •
Aj (0) - AL (') - A - Ce(0) +
+fk-1(A0) RLOop' = RLORetr
Optimal decision 3-OHPopt is defined in a way of calculation of all possible values of function fn (') = Aj (') ~YjC6 (') + + fk+1 ('+A0), which shall be compared. Optimal decision corresponds with maximal value at k-step.
The other economic-mathematical models of content updating optimizations may be settled on the basis of knowledge domain ontology: task of optimal update publications applying R. Floyd-S.Warshall algorithm in combination with L. Dice coefficient, task of updating technology time optimization using network planning and management methods etc.
6. Conclusions
1. Developed method of formal presentation of university knowledge domain and university ontology development using mathematical apparatus of graph theory allows for systematization the relations between syllabus and single terms of educational disciplines. The method is applied to educational content development and actualization within KMS toolset for settling the optimization tasks
of periodicity of content updating.
2. University content was divided into current educational information resources (IR) and prospective and retrospective information sources (IS), which increased the effectiveness of university information system maintenance and allowed for integration of domain knowledge ontology into content management toolset for monitoring, verification and permanent actualization of educational content using collaboration work of KMS users (students, lecturers).
3. Division of content updating levels was proposed which includes multiple content update publication facility using object-oriented principle based on knowledge domain ontology.
4. Domain knowledge ontology is applied to development of economic-mathematical modeling for settling optimization tasks related with content updating.
Obtained scientific results of the present research will be developed in additional optimization tasks based on university knowledge domain ontology: multi-criteria optimization of periodicity and terms of content updating for logically connected syllabus theoretical and practical disciplines; content selection for development of desirable competency using quality and price parameters.
References:
1. Appiah A., Building Post Roads
of a New Millennium // Almanac, University of Pennsylvania, 2007, Volume 53, No. 34
2. Garrot T. et al. Un marco teyrico para la economía del e-learning // Revista de Univestidad y Sociedad del Conocimiento, Vol. 5, n.e 1, с. 57-71
3. Stenmark D., Integrating Knowledge Management Systems with Everyday Work: Design Principles Leveraging User Practice. [Электронный ресурс]. URL: http:// dlib2.computer.org/conferen/hicss/ 143 5/pdf/143 50104.pdf (дата обращения: 20.09.2007)
4. Morrison J., Weiser M. A research framework for empirical studies in organizational memory // Proceedings of the Twenty-Ninth Annual Hawaii International Conference on System Sciences. IEEE Computer Society Press, 1996.
5. DeLone W.H., McLean E.R., Information Systems Success: The Quest for the Dependent Variable, in: Information Systems Research, Vol. 3, No. 1, 1992, 60-95.
6. ISO/IEC 14764:2006 Software Engineering - Software Life Cycle Processes - Maintenance [Electronic resource] // http://www.iso.org/iso/ iso_catalogue/catalogue_tc/ catalogue_detail. htm?csnumber=39064
7. Bellman, R.E. 1957. Dynamic Programming. Princeton University Press, Princeton, NJ. Republished 2003 : Dover
