Conference PaperPDF Available

FORMALIZATION AND MODELLING OF SECURE ACCESS AT E-LEARNING ENVIRONMENT

Authors:

Abstract and Figures

The paper presents an investigation of secure access to resources and personal data at combined e-learning environment. Conceptual model of the structure is designed after formalization based on state transition diagram. The proposed approach is based on the Markov chain (MC) theory so in this connection analytical definition of Markov model is presented. Several analytical and statistical experiments based on partial factor plan have been carried out. Collected results are processed by using " Develve " software and obtained statistical assessments are discussed.
Content may be subject to copyright.
PROCEEDINGS of the International Conference InfoTech-2015
78
Proceedings of the International Conference on
Information Technologies (InfoTech-2015)
17-18 September 2015, Bulgaria
FORMALIZATION AND MODELLING OF SECURE
ACCESS AT E-LEARNING ENVIRONMENT
Radi Romansky, Irina Noninska
Technical University of Sofia
e-mails: {rrom; irno}@tu-sofia.bg
Bulgaria
Abstract: The paper presents an investigation of secure access to resources and personal
data at combined e-learning environment. Conceptual model of the structure is designed
after formalization based on state transition diagram. The proposed approach is based on
the Markov chain (MC) theory so in this connection analytical definition of Markov
model is presented. Several analytical and statistical experiments based on partial factor
plan have been carried out. Collected results are processed by using Develve software
and obtained statistical assessments are discussed.
Key words: authentication & authorization, data protection, e-learning processes
evaluation, formalization and modelling, Markov chain.
1. INTRODUCTION
Nowadays educational technologies for the global network Internet are used by
constantly increasing number of learners which is due to easy access to information
resources, vireaty of virtual educational environments, collaboration and forums, etc.
Different forms as e-learning, d-learning, m-learning have been created to help learners
and teachers in their educational activities. In addition development of cloud
computing and mobile cloud computing extends concept of Internet importance for e-
learning. A new model called collaborative learning based on cloud computing is
proposed in [1]. The goal of this model is to solve the problem concerning gap, formed
by rapidly increasing number of learners on one hand and slowly growing of university
teachers on the other. The authors declare that an empirical evaluation of a prototype
system is able to give a high level of the proposed approach efficiency. Another point
of view for cloud computing at e-learning is presented in [2]. The authors defend their
idea that cloud services are applicable for e-learning processes, so they propose a rule-
based expert consultant system (SCCeLE) which goal is to help any e-learning
17-18 September 2015, BULGARIA
79
structure to determine the best cloud model. In this direction [3] empasises that
selecting partners of cloud computing could be releifed by developing a consortium, so
authors discuss its structure and evaluate several indicators defined by a model based
on Markov chain and AnalyticHierarchy Process.
Cloud computing (since 2006) and mobile cloud computing (after 2009-2010)
permit new architectural models developing and their implementation at e-learning
environment. For example, an analysis of cloud computing concept is made in [4] and
an architecture of cloud computing platform by combining features of e-learning is
proposed. This article introduces cloud computing to e-learning by developing an e-
learning cloud. The article [5] proposes an architecture for mobile based e-learning
system. The authors define the goal as “to develop e-learning system that performs
personalized delivery of course content according to learner contextual information
such as learning style and characteristics of the learning device”.
The purpose of this paper is to present an investigation of secure access to
resources of combined e-learning environment, proposed by the authors in [6]. The
main goal is to formalize and evaluate internal information and system resources’
accessing and using. In this reason different opportunities for investigation in the field
of e-learning are discussed in the next section 2. Sections 3 and 4 deals with
formalization of the processes by state transition network (STN) and stochastic
analytical modelling organization based on the Markov Chain (MC) theory. Section 5
extends evaluation by statistical analysis of calculated probability sets by using the
Develve software (http://develve.net/).
2. MODELLING OF SECURITY AND PRIVACY ASPECTS OF
E-LEARNING (RELATED WORK)
Evaluation of structure, processes and characteristics of different e-learning
environments, including security and privacy protection schemes could rely on
modelling methods. All popular and frequently used modelling techniques are based on
discrete or stochastic apparatus (state transition diagrams, colored Petri nets, decision
trees, Markov chains, Bayesian networks, hidden Markov models, component
analysis).
An important part of a modelling process is formal description of an investigated
process and defining corresponding conceptual model. Each process could be
described as a set of discrete states and transitions between them represented by graph
theory. For example Bicher at all [7] propose an idea to apply event-graph modelling
method for blended learning concept and propose a platform, named MMT
(Mathematics, Modelling and Tools). Event-graphs are very simple for defining and
flexible for behavior presentation. They could be used for deterministic and stochastic
modelling as well.
Another approach based on graph theory for e-learning processes modelling is
presented in [8]. A direct-hypergraph is used to build a model of a process connecting
three main objects: learner, knowledge and learning resources. Graph technique is
PROCEEDINGS of the International Conference InfoTech-2015
80
used to formalize different parts of e-learning, represented it as a structure of following
components: e-learning processes; personalized resource requirements; relationship
between processes, learner and learning resources. The model has been proposed to
optimize combined resources’ interdependence and to reveal dynamic of processes.
A combination of abstraction and probabilistic learning able to enhance statistical
model checking performance is discussed in [9]. This article proposes a formal
definition of set of traces generated from the original system and an abstract model is
designed. The model is verified based on combination of Monte-Carlo simulation with
statistical techniques. Authors validate this approach by Herman’s Self Stabilizing
protocol. Experimental results show reduction of abstract’s model size, verification
time and possibility to estimate probability of satisfaction by statistical model as well.
Another solution proposed in [10] is application of stochastic model and Bayesian
belief networks for learner assessments. A Bayesian network is a directed acyclic
graph that maps relationship called hypothesis-evidence. The authors discuss
integration at an application of mobile agent technology with case-based reasoning tool
for content delivery at e-learning system.
Markov chain (MC) apparatus is a method qiute often used for modelling of e-
learning processes, especially those, oriented to investigation of different approaches
for security and privacy protection and their efficiency. For example, articles [11] and
[12] present Markov chain and classification algorithms as a tool, applied to
investigate database collecting learner’s answers to asked questions. The authors from
Graz University of Technology (Austria) focus on a specific application named 1x1
trainer” designed for primary school children. Their goal is to understand whether
learners’ answers to already asked questions can affect the way they will answer the
subsequently asked questions. Another goal is identifying influential structures in the
history of learners’ answers considering MC of different orders.
Another application of MC presented in [13] regards to establishment of usage
profiles which target is e-learning oriented web site. With these profiles web based
platforms administrators can personalize sites according to preferences and behavior
of students, promoting easy navigation, functionalities and better abilities to meet their
requirements.
A review of Information Security Modelling dimentions for e-learning strategies
and environments is made in [14] where after analyses of threats, risks and
vulnerabilities at e-learning environments an Information Security framework for e-
learning technologies is proposed. The authors pointed out that people stay connected
each other and able to access services globally. In this reason, as the Internet is open
for all users to access and share information, security e-learning technologies are of
utmost important. This article describes e-learning as a combination of different
techniques, technologies and tools, including collaborative software, e-portfolios,
virtual environments, audio & video, PCs, tablets, notebooks, smart-boards, webcams,
screen-casting, discussion forums, chat, file sharing, video conferences, shared
17-18 September 2015, BULGARIA
81
whiteboards blogs/microblogs, wikis. All activities via the network must be guaranteed
by security measures which efficiency could be estimated by modelling.
Different approaches for information security modelling and investigation of
security aspects are proposed in articles, devoted to online learning. For example, a
trustworthiness model for secure learning assessment design in online web
collaborative learning groups is proposed in [15]. It could be accepted as a
representative of new group holistic security models that have not been completely
carried out yet.
A smart grid information security risk assessment model on the basis of set pair
analysis improved by Markov chain is discussed in [16]. Method, proposed in it avoids
shortcomings of traditional Expert Score with significant subjective factors and also
considers links between information system components, which make risks’ index
system closer to the reality. A stochastic modelling technique using Attack Graph for
security state of a network analysis is discussed in [17]. An integrated view of security
state of a network based on security attributes aggregation at each node is presented in
the article. This concept is called Cyber-Security Analytics and a cyber-situational
awareness model built on 4 levels is presented. The first level named “Perception”
includes security monitoring and intrusion detection techniques. Level 2 named
“Comprehension” supports risk analysis security visualization. Levels 3 and 4 are
“Mitigation” (risk mitigation) and “Forecasting” (predictive models).
3. FORMAL DESCRIPTION OF THE PROCESSES
Formalization of an investigated process is a compulsory stage which must
precede model developing. The authors of [7] specify using of event-graph apparatus
as “the most popular ways to describe a discrete-event model in figurative form”. This
is true because each process could be regarded as a sequence of discrete events with
transitions between them. Formalization by using graph structure could be determined
(binary or labelled directed / undirected graph) or stochastic (the transitions are
probabilities).
Investigated processes at combined e-learning architecture are formalized and
resulting graph is shown in fig. 1. The choice of graph apparatus is made based on
following conditions:
Number of investigated events at the structure is finite and this permits defining
a finite set of states presented by nodes in a directed graph.
Each observed event in the architecture has constant behaviour presented by
fixed procedure (algorithm), but realization is made under a stochastic flow.
Graph nodes present all observed events (sub-processes) and transitions
between nodes describe possible realization of an e-learning process.
PROCEEDINGS of the International Conference InfoTech-2015
82
Access to
external
resources
Input point
Authentication
Registration
PD
Student
level
Tutor
level
Administrative
level
Learning
resources
Educational
DB
Personal Data
Administrative
DB
End of work
Authorization
(checking right for
access)
Request
processing
No
Yes
Success
New user
New user’s profile
is created
DRMS
DB
Figure 1. Graph formalization of the access to the system resources
Following groups determine a finite set of states in graph formalization:
a) Functionalities of the Front Office sub-system: input point (access to official
portal of e-learning environment); registration (every new user should make a
registration before accessing and using different resources and services);
authentication (procedure able to guarantee secure and reliable access to environment
for registered users);
b) Functionalities of the Back Office sub-system: request analysis (preliminary
analyses of user’s request based on checking user’s profile with Personal Data PD
and level of rights for accessing internal or external learning resources based on Digital
Right Management System DRMS principles); authorization (determining user’s
status and defined rights to use requested information learning resource, evaluation
information, administrative information, personal data processing, etc.); request
processing (realization of a permitted request after succesfull authorization);
c) Functionalities of collaborated external resources: they unite spaces in the field
of cloud computing (services IaaS, SaaS, PaaS, data centres), social computing (social
media, social networks, social aggregators, blog/microblogs, forums, etc.) and
personal specialized web sites. All these spaces and technological opportunities could
be used by rent, so it is accepted that they are reliable protected, applying
recommendable information security and privacy protection standards as an obligation
of service owner and/or service provider.
The functionality of formalized combined e-learning environment can be
generalized as a family of discrete-time stochastic processes each of them described by
a random process X(t,
,
), were t is an ordered increased set of discrete time points
17-18 September 2015, BULGARIA
83
t0 < t1 < < tn T (T is a parametric space),
represents all random factors of the
system and
is a set with all possible design parameters. The realization of discrete-
time stochastic process is a sequence of states X(ti) for discrete moments tiT, for
i=0,1,2,…,n. This permits MC using for model design and if it is assumed that a state
transition from the current time ti to the next time ti+1 is independent on the time, the
MC becomes a stationary MC and the final probabilities could be determined based on
the MC theory.
4. MODELING AND INVESTIGATION
Modelling is made on a MC with 8 discrete states which are defined below. The
Markov model with its analytical definition and graph of states is shown in fig. 2.
Definition of a Markovian chain
a) States: S = {S1, …, S8} b) Matrix of transition probabilities P = {pij}
S1 Access to the front office;
S2 Registration procedure;
S3 Authentication procedure;
S4 Request analysis;
S5 Access to external resources;
S6 Authorization procedure for
access to internal resource;
S7 Using educational resource;
S8 Using system resource &
personal data processing
1
2
3
4
5
6
7
8
1
0
a
1-a
0
0
0
0
0
2
0
1
0
0
0
0
0
0
3
1-b
0
0
b
0
0
0
0
4
1-c-d
0
0
0
c
d
0
0
5
1
0
0
0
0
0
0
0
6
0
0
0
1-e-f
0
0
e
f
7
0
0
0
1
0
0
0
0
8
0
0
0
1
0
0
0
0
c) Vector of initial probabilities: P0 = {1, 0, 0, 0, 0, 0, 0}
Description of the conditions: a probability for unregistered (new) user; b
probability for correct authorization of registered user; c probability for a request to
access and using external educational resources; d probability for authorization
(determining the right) for using an internal resource; e probability for an internal
educational resource using (after successfull authorization); f probability for using
system resources and personal data processing.
A preliminary analysis of procedures is made and some assumptions are accepted
to simplify the analytical investigation. For example, it is assumed that the probability
of access to the system by new unregistered user is no more than 0,3 (a≤0,3) and the
unauthorized access to the resources (including attacks from external nodes) is in the
frame [10%, 30%], i.e. 0,1 b 0,3. Next assumption is that refusal of service (in
“Request Analysis” and “Authorization” states) is no more than 10% of all cases and
this permits to determine values: (1 - c - d) 0,1 (c + d) 0,9 and
(1 - e - f) ≤ 0,1 (e + f) ≤ 0,9.
PROCEEDINGS of the International Conference InfoTech-2015
84
S1
S2
S4
S5
S3
S6
S7
1-a
b
a
1-c-d
c
e
d
1-e-f
1-b
S8
f
a) Graph of states
1)9
.)8
.)7
.)6
.)5
).1(.)4
).1()3
.)2
).1(3).1()1
87654321
68
67
46
45
87634
23
12
5421
pppppppp
pfp
pep
pdp
pcp
pppfepbp
pap
pap
ppdcpbpp
b) Analytical model definition
Fig. 2. Markov model definition
Analytical solution of the model:
The system of equations, presented in fig. 2(b) has been solved by
transformations and presentations based on probability p1:
12 app
;
13 )1( pap
;
14 1)1( p
d
ab
p
;
145 1)1( p
d
ab
ccpp
;
146 1)1( p
d
ab
ddpp
;
167 1)1( p
d
ab
edepp
;
168 1)1( p
d
ab
fdfpp
.
The final substitution is made in equation (9):
1
)1( )1(
)1( )1(
)1( )1(
)1( )1(
)1( )1(
)1(
111
11111
p
dabdf
p
dabde
p
dabd
p
dcbc
p
d
ab
paapp
After solving this equation the following equations for final probabilities are
obtained (the substitution
= [2-2d+b(1-a)(1+c+d+de+df)] is used):
)1(
;
)1(
;
)1(
;
)1(
;
)1(
;
)1)(1(
;
)1(
;
)1(
8765
4321
abdf
p
abde
p
abd
p
abc
p
ab
p
da
p
da
p
d
p
5. ANALYTICAL INVESTIGATION AND EXPERIMENTAL RESULTS
Assumptions made in the previous section 4 permit to define working frame for
analytical investigation of the proposed MC-model by determining concrete
probabilities values: a{0,2; 0,25; 0,3}; b{0,7; 0,75; 0,8; 0,85; 0,9}; c{0,35; 0,4;
0,45}; d{0,45; 0,5; 0,55; 0,6} and (1-e-f)=0,1 e = f = 0,45 (equal probabilities for
access after authorization to educational or system resource).
17-18 September 2015, BULGARIA
85
Multi-factor experimental plan. The selected limited environment factor permits
constructing of partial factor experimental plan on the base of all combinations of
accepted probabilities values. Statistical realization of this plan is made by software
Develve and experimental results are shown in figure 3. Different statistical
assessments are generalized in table 1.
Figure 3. Statistical assessments for the probabilities
Table 1. Generalization of the statistical assessments for the probabilities
p1
p2
p3
p4
p5
p6
p7
p8
n
108
108
108
108
108
108
108
108
Mean
0,1995
0,0501
0,1493
0,2505
0,1001
0,1318
0,0593
0,0593
Median
0,1992
0,0492
0,1490
0,2512
0,1003
0,1313
0,0591
0,0591
MIN
0,1501
0,0300
0,1151
0,2198
0,0787
0,0989
0,0445
0,0445
MAX
0,2522
0,0757
0,1885
0,2777
0,1216
0,1666
0,0750
0,0750
=[max-min]
0,1021
0,0457
0,0733
0,0579
0,0429
0,0677
0,0305
0,0305
Variance
0,0006
0,0001
0,0003
0,0002
0,0001
0,0003
0,0001
0,0001
St.Dev.
0,02391
0,01139
0,01757
0,01293
0,01051
0,01778
0,008
0,008
Conf.Int. (T)
0,005
0,002
0,003
0,002
0,002
0,003
0,002
0,002
Conf. Int. (N)
0,005
0,002
0,003
0,002
0,002
0,003
0,002
0,002
Experimental results show that the probabilities have very small assessments for
the variance and standard deviation (St.Dev.). The confidence intervals (Conf.Int.) by
using Student’s T-distribution (T) and by using normal distribution (N) are equal. The
PROCEEDINGS of the International Conference InfoTech-2015
86
difference between maximal and minimal probabilities is largest for the probability
p1 that presents loading front office portal by access of different remote users.
Average values obtained for final probabilities on the base of realized factor
experimental plan are summarized graphically in figure 4. These results show that the
highest value has utilization of the state “Request analysis” as a main part of the back
office. At the same time the assessments for two security procedures authentication
(p3) and authorization (p6) have a quite little difference (about 0,017) that is
determined by the access to external resources. A comparative statistical analysis of
using these two procedures with all accesses to the front office input point (probability
p1) is made by Develve” and is shown in figure 5. Difference between cases of
process initialization by access to the Front Office (input point p1) and using the
Back Office procedure Request Analysis” (p4) could be explained by the possibility
of many internal requests processing by other states (authorization, educational
resources, system resources, personal data processing).
Figure 4 Average vallues for the final probabilities
Figure 5. Comparative analysis of authentication and authorization with all user’s accesses
17-18 September 2015, BULGARIA
87
Time series for the obtained data sets of selected probabilities are shown in fig. 6.
They could be related to the average values (fig. 4) and difference from table 1.
(a) Procedures in the Front Office
(b) Utilization of external (p5) and
internal (p6) resources
Figure 6. Time series for some data sets of selected final probabilities
Additional statistical analysises of mutualy dependancy between final
probabilities are provided and results are shown in the following figures. Figure 7
presents results from correlation analasis between selected procedures (a) and
calculation of all correlation coefficients between procedures in stationary regime (b).
One way Anova analysis for states ‘1’, ‘3‘and ‘4’ is shown in fig. 8(a) and calculated
assessments by linear multi-regressinon analysis for the state “authorization” (p6) from
the states “input” (p1) and “authentication” (p3) is presented in fig. 8(b).
a) Correlation between selected pairs
b) Multi correlation calculated by Develve
Figure 7. Coefficients of correlation between final probabilities for different states
PROCEEDINGS of the International Conference InfoTech-2015
88
a) One way ANOVA analysis
b) Linear multi-regression
Figure 8. Dependency analysis by Develve
One factor experimental plan has been made to evaluate dependencies between
activities of procedures. The functional dependency of the final probabilities from the
level of new unregistered users accessing the input point of front office is investigated
by consecutively increasing factor a” from 0,02 to 0,70 by step 0,02 (number of
registrations n=36). For each other factor (b, c, d, e, f) its average value calculated
during the multi-factor experiments is fixed. Fig. 9 shows implementation level of
procedures “authenticationand “authorization and their relation with all access
attempts to e-learning environment (input point of the front office).
Figure 9. Changing of the security procedures utilization with increasing the new user access
17-18 September 2015, BULGARIA
89
The diagram shows nearly equivalent assessments (0,160,01) for authentication
and authorization in small range of factor “a, but the difference between them and the
level of front office accessing increases to a large degree (difference about 0,30) for
large scale of factor “a.
Fig. 10 shows variation of accessing external and internal resources as a function
of factor a”. Two types of access are similar and dependence on a is in limited
borders but front office accessing (input point) increases slowly.
Figure 10. State of the access to the external and internal resources at increasing “a”
Diagrame in fig. 11 presents a comparison between utilization of registration”
procedure and requesting any external or internal resource. Factor a accepts the
same values and as it is expected registration utilization increases in the large range of
the set, but the maximal level is no more than 0,22.
Figure 11. Relation between registration and access to the resourses
PROCEEDINGS of the International Conference InfoTech-2015
90
CONCLUSION AND FUTURE WORK
At present, when new network technologies are intensively used by people, cloud
services and social computing could extend e-learning environments with new
functionality. According to published trends cloud computing will be rapidly
developed in the future [3]. Many publications emphasise this obstacle and different
authors suggest elements of cloud and social computing to be included in all e-learning
environments. Since network communication and remote access to different resources
will continue to threaten data integrity, strong measures for information security and
personal data protection must be implemented.
The paper presents an approach for developing and investigation of combined e-
learning environment based on MC theory. For this purpose a MC-model has been
designed and different experiments based on multi-factor and one-factor plan have
been carried out. Obtained stochastic assessments are processed in addition by using
statistical means and tools in order to evaluate borders of changes and tendencies in
the procedures employment. The paper presents full illustrations of calculated
assessments. All results analysed during this investigation could be used as a basis for
further experimental work, based on simulation.
REFERENCES
[1] Liao, J. et all. (2014). Collaborative Cloud: a New Model for e-Learning. Innovations in
Education and Teaching International, vol. 51, No. 3, pp. 338-351. Available at:
http://www.tandfonline.com/doi/abs/10.1080/14703297.2013.791554
[2] Gamalel-Din, S., R. Salama, M. Al-Sowaiel (2014). An Expert Consultant for Cloudifying E-
Learning Environments, 2nd Int’l Conference on Future Internet of Things and Cloud
(FICLOUD), Barcelona, Spain, 27 August, pp. 308-315.
[3] Liang Hong, Changyuan Gao (2014). Partner Selection of Cloud Computing Federation Based
on Markov Chains. Computer Modelling & New Technologies, vol. 18 No. 12B, pp.590-594.
(http://www.cmnt.lv/en/on-line-journal/2014/2014-volume-18-12/part-b-information-and-
computer-technologies/partner-selection-of-cloud-computing-federation-based-on-markov-
chains)
[4] Masud, M. A. H., X. Huang (2012). An E-learning System Architecture based on Cloud
Computing. World Academy of Science, Engineering and Technology, 62, pp.74-78.
[5] Nehru, V., M. Anto Bennet (2014). An Architectural-Model for the Mobile based E-Learning
Environment. International Journal of Computer Science and Mobile Applications,
ISSN: 2321-8363, Vol.2, No. 11 (November), pg. 41-48. Available at:
http://www.academia.edu/9186926/An_Architectural-Model_for_the_Mobile_based_E-
Learning_Environment
[6] Romansky, R., I. Noninska (2015). Security and Privacy Principles Realization in e-Learning
Architecture. Proc. 29th Int’l Conference on Information Technologies (InfoTech-2015),
17-18 November, St. St. Constantine and Elena, Bulgaria (in current issue).
[7] Bicher, M. et all. (2014). Support Event-Graph Lectures by the MMT E-learning System.
Simulation Notes Europe (SNE), vol. 24, No. 1 (April), pp.47-50.
[8] Sun X., Z. Li, S. Hu (2014). Directed-Hypergraph Based E-Learning Process Modeling
Supporting Dynamic-Personalized-Combined Resource Optimization, Proc. of 5th Int’l
Conference of Digital Home (ISBN 978-1-4799-4285-5), 28-30 November, pp.324-330,
17-18 September 2015, BULGARIA
91
[9] Nouri, A. et all (2014). Faster Statistical Model Checking by Means of Abstraction and Learning.
RV, September, Toronto, Canada (www-verimag.imag.fr/~jcombaz/misc/rv14/pdv)
[10] Mangalwede, A. R., D. H. Rao (2010). Application of Bayesian Networks for Learner
Assessment in E-Learning Systems. International Journal of Computer Applications, vol. 4,
No.4 (July), pp.23-27.
[11] Taraghi, B. et al. (2014). On using Markov Chain to Evidence the Learning Structures and
Difficulty Levels of One Digit Multiplication. Proc. of the 4th Int’l Conf. on Learning Analytics
and Knowledge (LAK’14), Indianapolis, USA, 24-28 March, pp.68-72
[121] Taraghi, B. et al. (2014). Markov Chain and Classification of Difficulty Levels Enhances the
Learning Path in One Digit Multiplication. Lecture Notes in CS 8523, LCT 2014, part I,
pp. 322-333.
[13] Marques, A, O. Belo (2011). Discovering Student Web Usage Profiles Using Markov Chains.
The Electronic Journal of e-Learning, vol. 9, No. 1, pp. 63-74, available online at
http://files.eric.ed.gov/fulltext/EJ930258.pdf
[14] Baby, A., A. Kannammal (2014). Information Security Modelling in an E-learning Environment.
International Journal of Computer Science, vol. 11, No. 1 (January), pp. 195-200.
[15] Miguel, J et al. (2015). Security in Online Web Learning Assessment. Providing an Effective
Trustworthiness Approach to Support e-Learning Teams. World Wide Web, 25 January (22 p.)
DOI 10.1007/s11280-014-0320-2
[16] Jianye, Z. et al. (2014). Information Security Risk Assessment of Smart Grid Based on
Absorbing Markov Chain and SPA. International Journal of Electric Power Systems, vol. 15,
No. 6, pp.527-532.
[17] Abraham, S., S. Nair (2014). Cyber Security Analytics: A Stochastic Model for Security
Quantification using Absorbing Markov Chains. Journal of Communications, vol. 9, No. 12
(December), pp.899-907.
... The proposed structure unites two sub-systems where functions for security and privacy protection are divided (Romansky, 2015b): Front office is designed for user's authentication and personal profiles creation; Back office is responsible for access control based on authorization, digital rights management and personal data protection. A preliminary conceptual defining is made and a formalization of components of the proposed combined e-learning environment by using state transition network (STN) is realized (Romansky, 2015c). A Markov model for investigation of processes in this heterogenic e-learning environment is designed. ...
... All popular and frequently used modelling techniques are based on discrete or stochastic apparatus (state transition diagrams, coloured Petri nets, decision trees, Markov chains, Bayesian networks, hidden Markov models, component analysis). Extended discussion is made in (Romansky, 2015c). ...
... A formal description of processes in proposed combined e-learning environment with heterogenic structure is presented in figure 3 (Romansky, 2015c). The choice of graph apparatus is made based on following conditions: ...
Article
Full-text available
The contemporary digital world based on network communications, globalization and information sharing outlines new important targets in the area of privacy and personal data protection which reflect to applied principles of secure access to proposed information structures. In this reason the aim of secure access to all resources of an e-learning environment is very important and adequate technological and organizational measures for authentication, authorization and protection of personal data must be applied. Strong security procedures should be proposed to protect user's profiles, designed after successful registration and all personal information collected by educational processes. The goal of this article is to present an idea to combine traditional e-learning technologies with new opportunities that give mobile applications, cloud services and social computing. These technologies can endanger data security since they make possible remote access to resources, sharing information between participants by network communications. In order to avoid data vulnerabilities users must be identified and authenticated before, i.e. to be allowed to access information resources otherwise integrity and confidentiality of e-learning system could be destroyed. In order to propose solution basic principles of information security and privacy protection in e-learning processes are discussed in this article. As a result, an organizational scheme of a system for information security and privacy is proposed. Based on these principles a graph formalization of access to the system resources is made and architecture for combined (heterogenic) e-learning architecture with secure access to the resources is designed. Analytical investigation based on designed Markov chain has been carried out and several statistical assessments delivered by Develve software are discussed.
Chapter
Full-text available
The contemporary digital world based on network communications, globalization and information sharing outlines new important targets in the area of privacy and personal data protection which reflect to applied principles of secure access to proposed information structures. In this reason the aim of secure access to all resources of an e-learning environment is very important and adequate technological and organizational measures for authentication, authorization and protection of personal data must be applied. Strong security procedures should be proposed to protect user’s profiles, designed after successful registration and all personal information collected by educational processes. The goal of this article is to present an idea to combine traditional e-learning technologies with new opportunities that give mobile applications, cloud services and social computing. These technologies can endanger data security since they make possible remote access to resources, sharing information between participants by network communications. In order to avoid data vulnerabilities users must be identified and authenticated before, i.e. to be allowed to access information resources otherwise integrity and confidentiality of e-learning system could be destroyed. In order to propose solution basic principles of information security and privacy protection in e-learning processes are discussed in this article. As a result, an organizational scheme of a system for information security and privacy is proposed. Based on these principles a graph formalization of access to the system resources is made and architecture for combined (heterogenic) elearning architecture with secure access to the resources is designed. Analytical investigation based on designed Markov chain has been carried out and several statistical assessments delivered by Develve software are discussed.
Article
Full-text available
This paper proposes a trustworthiness model for the design of secure learning assessment in on-line web collaborative learning groups. Although computer supported collaborative learning has been widely adopted in many educational institutions over the last decade, there exist still drawbacks which limit their potential in collaborative learning activities. Among these limitations, we investigate information security requirements in on-line assessment, (e-assessment), which can be developed in collaborative learning contexts. Despite information security enhancements have been developed in recent years, to the best of our knowledge, integrated and holistic security models have not been completely carried out yet. Even when security advanced methodologies and technologies are deployed in learning management systems, too many types of vulnerabilities still remain opened and unsolved. Therefore, new models such as trustworthiness approaches can overcome these lacks and support e-assessment requirements for e-Learning. To this end, a holistic security model is designed, implemented and evaluated in a real context of e-Learning. Implications of this study are remarked for secure assessment in on-line collaborative learning through effective trustworthiness approaches.
Conference Paper
Full-text available
This paper investigates the combined use of abstraction and probabilistic learning as a means to enhance statistical model checking performance. We are given a property (or a list of properties) for verification on a (large) stochastic system. We project on a set of traces generated from the original system, and learn a (small) abstract model from the projected traces, which contain only those labels that are relevant to the property to be verified. Then, we model-check the property on the reduced, abstract model instead of the large, original system. In this paper, we propose a formal definition of the projection on traces given a property to verify. We also provide conditions ensuring the correct preservation of the property on the abstract model. We validate our approach on the Herman’s Self Stabilizing protocol. Our experimental results show that (a) the size of the abstract model and the verification time are drastically reduced, and that (b) the probability of satisfaction of the property being verified is correctly estimated by statistical model checking on the abstract model with respect to the concrete system.
Article
Full-text available
Most of the current research in security analysis has been centered on identifying threats and vulnerabilities and providing suitable defense mechanisms to improve the robustness of the network and systems. While this approach is attractive, it provides limited insight into understanding the impact these attacks have on the overall security goals of the network and the system. Attack Graph as a model lends itself nicely to the analysis of the security state of a network. Most of the Attack graph based metrics proposed in the literature are one dimensional; however the research community has acknowledged the fact that security needs to be treated as a multidimensional concept. In this paper, we utilize stochastic modeling techniques using Attack graphs to define a complementary suite of quantitative metrics to aid the security engineer in visualizing the current as well as future security state of the network and optimizing the necessary steps to harden the enterprise network from external threats. We present experimental results from applying this model on a sample network to demonstrate the practicality of our approach.
Article
Full-text available
Nowadays, Web based platforms are quite common in any university, supporting a very diversified set of applications and services. Ranging from personal management to student evaluation processes, Web based platforms are doing a great job providing a very flexible way of working, promote student enrolment, and making access to academic information simple and in an universal way. Students can do their regular tasks anywhere, anytime. Sooner or latter, it was expected that organizations, and universities in particular, begin to think and act towards better educational platforms, more user-friendly and effective, where students find easily what they search about a specific topic or subject. Profiling is one of the several techniques that we can use to discover what students use to do, by establishing their user navigation patterns on Web based platforms, and knowing better how they explore and search the sites" pages that they visit. With these profiles Web based platforms administrators can personalize sites according with the preferences and behaviour of the students, promoting easy navigation functionalities and better abilities to response to their needs. In this article we will present the application of Markov chains in the establishment of such profiles for a target eLearning oriented Web site, presenting the system we implemented and its functionalities to do that, as well describing the entire process of discovering student profiles on an eLearning Web based platform.
Article
Full-text available
The number of learners using e-learning has been increasing at an enormous rate in the past decade due to easy access to higher educational resources via the Internet. On the other hand, the number of teachers in most universities is growing slowly. As a result, instructional problems have emerged due to the lack of sufficient support to learners in the e-learning process. Collaborative learning has been suggested as a solution to this problem. Current collaborative learning puts a heavy emphasis on teamwork and group discussion, but provides inadequate support for individual learners during the whole learning process. This paper presents a new model of collaborative e-learning using cloud computing technology to solve this problem. A prototype system has been developed with empirical evaluations to demonstrate the effectiveness of the proposed approach.
Article
To assess and prevent the smart grid information security risks more effectively, this paper provides risk index quantitative calculation method based on absorbing Markov chain to overcome the deficiencies that links between system components were not taken into consideration and studies mostly were limited to static evaluation. The method avoids the shortcomings of traditional Expert Score with significant subjective factors and also considers the links between information system components, which make the risk index system closer to the reality. Then, a smart grid information security risk assessment model on the basis of set pair analysis improved by Markov chain was established. Using the identity, discrepancy, and contradiction of connection degree to dynamically reflect the trend of smart grid information security risk and combining with the Markov chain to calculate connection degree of the next period, the model implemented the smart grid information security risk assessment comprehensively and dynamically. Finally, this paper proves that the established model is scientific, effective, and feasible to dynamically evaluate the smart grid information security risks.
Conference Paper
The need for e-Learning is becoming apparent due to the scarceness of resources of the traditional education. Therefore, many education and training organizations became aware of the importance of adopting e-Learning. In addition, new purely virtual learning organizations start emerging intensively. On the other hand, new technology-namely, cloud computing-has emerged. Business enterprises adopted this technology and have developed Cloud-based business models in which both infrastructures as well as software applications are offered for renting, which make it possible to use expensive infrastructures and software applications without high setup expenses yet at affordable low operating costs. This research studied how to employ Cloud Computing in e-Learning business. The different types of e-Learning organizations are identified and the e-Learning services offered by each type are classified. The Cloud Model that best fits the nature of each of these services according to the e-Learning organizations' affordability of resources as well as the abilities of the technological environment is then recommended. Accordingly, taxonomy of Cloud Models mix is defined for each type of those e-Learning organizations and a rule-based expert consultant system (SCCeLE) is developed to aid them in determining the best cloud model mix that best fits their current situations and targets. On the other hand, this developed taxonomy, is believed, will encourage the Cloud Service Providers (CSP) to conquer the business of e-Learning as it aids them in drawing a full picture on the e-Learning market and supports them in identifying their target market and in developing their e-Learning support business packages.
Article
In order to recommend the combined e-learning resources to a learner according to his objective and knowledge background automatically and dynamically, an e-learning process model was presented. The directed-hyper graph was used to model the process which connected the learner, the knowledge and the learning resources. The structure of an e-learning process, the personalized resource requirement corresponding to different kinds of knowledge backgrounds, and the relationship among the process, the learner and the resource environment were modeled graphically and formally. Then how the model support the combined resource optimization personally and dynamically was discussed by using the hyper graph theory and the semantics appended to the model. A set of modeling and optimizing rules and theorems were given too. Through contrasting with the method provided by Acampora, it showed the effectiveness of this method. Finally, the application and limitation of this method was discussed.