Problem of Trust in e-Learning Environment
Vladislav Petrov, Natalia Miloslavskaya, Victor Gorbatov and Anatoliy Durakovskiy
The National Research Nuclear University «MEPhI», 31 Kashirskoye Shosse, Moscow, Russia
Keywords: Distance Learning, e-Learning, Information Security, Problem of Trust, Trusted e-Learning Environment.
Abstract: The problem of trust and methodological approaches to resolve it for one of the most widespread types of
open information systems – an e-Learning environment (ELE) – are discussed. For that purpose the state of
trust implementation in distance learning (DL) is analyzed and its peculiarities are shown. An information
security (IS) threats model for ELE is proposed. The methodological foundations of trust building for ELE
are described. The results obtained allow to determine the goals and objectives for further research,
including in particular the formulation of the task of developing a formalized (unified) model of building
trust for learning management system (LMS) information resources using an integrated (complex) approach
to IS insurance.
1 INTRODUCTION
Technological IT advances are considerably
modifying traditional forms of activity, in particular,
in the field of educational services. E-learning
systems (ELS), using distance learning technologies
(DL) for achieving the best possible performance,
became an integral part of the modern educational
process as its specific form or a separate part of
broader blended educational process. Teaching
Internet-based technologies are widely represented
on the market of IT products and are commonly used
by educational institutions in practice.
However, despite of a particular attention to the
technological component of insuring information
security (IS) while applied to DL, there is still a
number of outstanding issues obligatory in
implementation of ELS. At present the main goal, in
our opinion, is a solution of the so-called problem of
trust and building a trusted e-Learning environment
(TELE).
In accordance with the terminology set for the IS
field a concept of "building trust" in ELS can be
defined as a fulfilment of the generally known triad
of IS requirements:
availability of ELS resources that seems not to
require special comments;
confidentiality/privacy of ELS resources in
accordance with the law or another restrictions.
For example, the personal data provided by
ELS users should have such a property in the
Russian Federation. The necessity to fulfil this
requirement determines the current
development period of the technological
component of IS ensuring in ELS;
integrity of ELS resources as no no-authorized
modification provides legal value of learning
outcomes on the basis of their trustworthiness
and/or non-repudiation from these results.
Which properties and in what combination are
required to perform depends on the IS threat and
intruders models designed for the specific protected
object.
The paper shows that the above mentioned
problem of trust, due to the peculiarities of its
fulfilment for the third condition, still not having a
satisfactory solution, at least, on the existing market
of IT products.
The remainder of the paper is organized as
follows. Section 2 analyses the state of trust
implementation in DL. Section 3 shows the
peculiarities of trust in DL. An IS threat model for e-
Learning environment (ELE) is proposed in Section
4. The methodological foundations of trust building
for ELE are discussed in Section 5. In conclusion
main results of the work are shown and future
research is pointed out.
2 RELATED WORKS
Many of the tasks of securing information systems
424
Petrov V., Miloslavskaya N., Gorbatov V. and Durakovskiy A..
Problem of Trust in e-Learning Environment.
DOI: 10.5220/0005477604240429
In Proceedings of the 7th International Conference on Computer Supported Education (CSEDU-2015), pages 424-429
ISBN: 978-989-758-107-6
Copyright
c
2015 SCITEPRESS (Science and Technology Publications, Lda.)
(ELS as an example), in particular maintaining
confidentiality of their information assets, are
performed by applying user identification,
authentication and authorization mechanisms,
providing the legal nature of their interaction with
the system. However, even this aspect is difficult to
be adjusted in DL in terms of anywhere recognized
international norms and laws. Today only one
methodological document – the international
standard ISO/IEC 24703:2004 – Information
technology – Participant identifiers (ISO/IEC
24703:2004) and its Russian harmonized analogue
GOST R ISO/IEC 24703-2011 – can be specified,
where only data types used to identify the
participants of the educational process are defined.
And aspects of ensuring IS, in particular, the
personal data security concerning participants’
identifiers usage as well as their authentication are
beyond the scope of this standard.
There are a lot of research works devoted to
investigations relating to different aspects of IS in
ELE. But none of them considers the trust in DL as a
systemic problem of IS in the above mentioned goal-
setting.
The necessity of IS ensuring for DL while using
the Internet as an open type communication was
investigated in (Furnell and Karweni, 2001). But the
paper does not provide a description of the specific
approaches to the problem solution.
Other researchers (Nickolova and Nickolov,
2007), (Eibl, 2010) confined only to build up the so-
called threat models or a list of potential dangers.
The majority of works (Weippl, 2005), (Ullah at
al, 2012), (Kumlander, 2008), (Inaba, Watanabe and
Kodate, 2003) are connected with the study of the
issues on effective legal access control, in other
words of such a service which guarantees the right to
use the system only by its authorized users. That is,
the problem of countering unauthorized access in its
traditional understanding was solved, including
using of biometric identification/authentication for
DL process participants and monitoring compliance
with the passage of various kinds of control
activities.
All these definitely important aspects of ensuring
IS for DL, in particular, the access restriction for
illegal users, nevertheless, do not realize all the
above mentioned conditions for trust building as a
complex problem. For example, the possibility of
countering action against such a real threat as a non-
verbal substitution of a legal trainee on the distance
progress testing procedure with his consent.
Traditional access control mechanisms do not
provide an effective mechanism to counter the
known IS threat called masquerading. The reason is
that a signature carrier (for example token) and the
software that generates these signatures can be
considered as alienable. And this fact does not
enhance the level of trust in ELS.
3 PECULARITIES OF DISTANCE
LEARNING AS AN OPEN
SYSTEM
An additional aspect, in a certain way complicating
comprehension of the problem of trust, is the
fundamentally open nature of ELE. The traditional
understanding of ELE as an open system is based on
the definition suggested by the Committee of IEEE
POSIX 1003.0 (1003.0-1995 – Guide to the
POSIX(R) Open System Environment). It is
concerned as a computer environment that
implements an open interface specifications and
services (environment services). This environment
supports the data formats sufficient to provide the
following properties:
extensibility is the system ability to add new
application functions or to modify some
functions from the amount of already
performed without modifying the rest of the
subsystems;
scalability is the system ability to increase its
productivity while expanding resources;
portability is the ability to transfer the system
to a more advanced hardware and software
platforms while their upgrading or
replacement;
interoperability is the system ability to interact
with other systems, if necessary, referring to
information resources (databases, knowledge
bases, and etc.) of these systems or performing
specific tasks using their computing resources
when their own resources are insufficient;
integration is the ability to combine several
systems/databases/applications for different
purposes in a single multifunctional
system/data store/multi-tier client/server
architecture.
Such a common representation of the open
system, created on the basis of a cloud-based
technology according to SaaS model (Software as a
Service) (Docebo.com), (ProProfs.com), is widely
used in DL and other network technologies
implemented in the Intranet, Internet, or their
combination shown in Figure 1.
ProblemofTrustine-LearningEnvironment
425
Figure 1: Typical LMS’s structure.
It is interesting to note that such a learning
management system (LMS) service deployment
model as a public cloud, in other words an
infrastructure for free mass use, is frequently used in
the technology of cloud computing applied here to
DL purposes. Such an infrastructure may be in
ownership, management and maintenance of
commercial, academic and government
organizations (or of any combination thereof). The
public cloud physically exists in the jurisdiction of
the owner (server provider) and the problem of
integrated security compliance of information
resources (the problem of trust) becomes more
obvious than in the untrusted environment in the
setting of local installation (Miloslavskaya, Petrov
and Tolstoy, 2014), (IITO UNESCO, 2013).
In that context trust is a (positive) decision on
the admissibility of interaction with an IT system
and acceptance of the results of its functioning.
We formally define the open trusted IT
environment as a set of hardware and software,
providing creation, application and development of
the system in accordance with its purpose and
having a full set of software, design and in-line
documentation, including program source codes
meeting IS requirements and confirmed by
certificates of compliance audit reports) by the
relevant legislative regulation systems.
Analysis of some sources (Hameetha Begum,
Sheeba and Nisha Rani, 2013), (Gunasekar and
Anirudh, 2011), (Wenan Tan et al, 2012) showed
that the use of the public cloud as a basis for ELE is
not safe in the context of our definition of trust.
There are currently no generally recognized
requirements to cloud IS assurance, in particular a
detailed (particular) IS threat model for cloud
computing environments, as well as verification
mechanisms able to unambiguously define a user.
The reason is that even using of integrated circuit
card or USB key does not warrant that the access has
been gained by the very legitimate user because they
are a removable media of key information.
Consequently, the use of the cloud infrastructure has
high risks and more limited access control
capacities. Therefore, one of the main problems of
cloud computing is the users’ trust formation in
above mentioned goal-setting in relation to cloud
providers and their possibilities to form TELE.
4 IS THREATS MODEL FOR
E-LEARNING ENVIRONMENT
Analysis of (Miloslavskaya, Petrov and Tolstoy,
2014), (Najwa Hayaati Mohd Alwi and Ip-Shing
Fan, 2010) shows that none of the IT products
available at the LMS market (both cloud-based and
local) does not currently warrant the necessary IS
level for LMS resources as it is provided by usage of
the specific information protection tools (IPT).
However, if consider the existing research papers on
IS threats in DL (Nickolova and Nickolov, 2007),
(Najwa Hayaati Mohd Alwi and Ip-Shing Fan,
2010), then the major threats are described there in
some detail including those using users’ (DL process
participants) software. Nevertheless, the list of IS
threats described there is not complete. So, it can be
said that the question of regulation of requirements
to DL process participants’ (users’) working
stations/computers/mobile devices (WS) has not
been discussed up to date. The absence of such
requirements to the DL process leads the DL
participants and LMS (educational institution)
interaction environment to objective distrust. For
example, during e-assessment (progress testing) the
DL process participants have a possibility to use
third-party Internet-resources that are not permitted
within the traditional progress testing activities.
Thus the extended set of key IS threats is
represented as follows:
1) an unauthorized access to LMS information
resources, including an unauthorized access to the
answers to the control data (tests, quizzes, etc.).
2) a possibility of LMS breaking by hackers as
well as legal users in order to substitute the author's
answers to the progress testing and, as a
CSEDU2015-7thInternationalConferenceonComputerSupportedEducation
426
consequence, changing/replacement of the testing
results.
These IS threats are realized in the form of
attacks against LMS server in order to progress
testing data compromise, unauthorized copying of
learning materials and personal data theft or attacks
against DL instructor/teacher WS in order to obtain
data/edit the results of academic performance rating,
spoofing of user identity while accessing ELE and
interference at all DL stages (Miloslavskaya, Petrov
and Tolstoy, 2014), (Siciliano, 2013).
The IS threats model for TELE in case of
vulnerabilities allowing LMS’s IS violation is shown
in the Table 1.
Table 1: IS threats model for TELE.
Reasons –
Absence of
trust in
IS threat –
Violation of
Description
WS Integrity,
availability and
confidentiality
of LMS (with its
resources) and
WS
Use of undocumented
features of custom and
malicious software with
the purpose of substitution
of access permission,
information leakage and
modification and
p
erforming attacks against
WS and LMS
LMS
information
(including
progress
testing data
)
transfer
paths
LMS resources
integrity and
confidentiality
and DL process
participants’
privacy
Capturing information
circulating between DL
process participants with
the purpose of
compromise both
authorization and progress
testing data
Authenticity
of DL
process
participants
DL process
participants’
privacy
Substitution of legal DL
process participants for
falsification of progress
testing results
The result of both intentional and unintentional
actions of any intruder is a violation of LMS
properties and, as a consequence, the distortion of
the real (true) information: deletion, substitution and
modification of rights, system’s hacking, the
replacement of progress testing results, etc. All this
information is described in a separate document
entitled “Intruders’ Model”. The model contains a
formalized intruders’ classification, including
description of their experience, knowledge, available
resources needed for IS threats implementation,
possible motivation of their actions and IS threats
implementation techniques used by the given
intruders.
The extended IS threat model analysis leads to a
conclusion that the current situation with ensuring IS
for DL is far away from ideal. In other words DL
process participants do not have trust in DL process.
One of the main reasons is the fact that the main DL
process participants – trainees – can be interested in
falsification of their progress testing results. And as
it follows from analysis of the Table 1, actually all
existing DL information-sharing environment and its
basic structure do not meet above mentioned
requirements of building trust (Figure 2).
Figure 2: Typical LMS structure with intruders (in red)
and their attacks’ objects.
5 METHODOLOGICAL BASIS
FOR BUILDING TRUST IN
E-LEARNING ENVIRONMENT
Thus, the scenario of above mentioned trust building
in ELE requires essential attention to all of the
following areas: upgraded (non-traditional) user
identification and authentication; access control; and
connection protection and et al (Miloslavskaya,
Petrov and Tolstoy, 2014).
In order to determine the methodological
approaches for building TELE for the new DL
forms, some necessary conditions in terms of well
known IS requirements (such as availability,
confidentiality and integrity) to LMS information
resources (assets) can be formulated.
Definition 1. TELE is determined as the
automated information system with a typical
document of compliance setting IS requirements.
That means the fulfillment of a set of the following
conditions:
ProblemofTrustine-LearningEnvironment
427
the presence of the reference software and
operating systems with control and
differentiation of starting processes and tasks;
the presence of certified IPT and cryptographic
protection facilities;
building of trust in circulating information;
the presence of a complete set of software and
hardware documentation prepared in
accordance with applicable standards;
the presence of an environment support system
at all stages of the life cycle from software
design process to compliance evaluation
certification.
Definition 2. The relatively trusted ELE means
an ELE that does not fulfill at least one of the
conditions of Definition 1.
Definition 3. The untrusted ELE means an ELE
that does not fulfill the first three conditions of
Definition 1.
Limitation 1. The relatively trusted ELE for DL
process participants is considered as one of the
untrusted ELE components.
For normal TELE functioning a property
precluding the possibility of implementing a variety
of IS threats Th = {Th
}, emerging and ongoing in
interaction with this environment via untrusted LMS
channels, should be carried out. DL process
participant’s WS is a collection of finite disjoint sets
of trusted {ET} and untrusted {EUnt} components
being distributed across hierarchical layers of
embedded components L
, where i [0, N], and L
is a basis component of this hierarchical structure, in
which the physical access to hardware resources of
DL process participants in distributed ELE is
delegated by the corresponding means.
Based on that the following statements can be
introduces.
Statement 1. A component L
realized only by
the trusted computing facilities can be allocated in
TELE. Therefore, the computer environment built
on L
will be protected from all hypothesized IS
threats Th.
Statement 2. The components L
, where i
[2, N],are not safe or/and are untrusted by definition,
as meanwhile realization of both external and
internal unauthorized access is possible, during
which the properties of confidentiality, integrity and
availability for WS of DL process participants can
be violated.
Statement 3. Trusted components of an arbitrary
component are functioning within this component’s
security boundaries and must interact with
neighboring components via secure interfaces.
Statement 4. To ensure availability and integrity
the component L
must support recursive methods
of control, management, integrity monitoring of the
trusted components constituting L
, wherei [1, N],
and information circulating between the
components. For all the components L
, where i
[0, N], an access differentiation model of software
components to protected resources must be
supported.
The above conditions allow to correctly
formulate the non-existent at the moment task of
developing a formalized (unified) model of building
trust for LMS information resources based on an
integrated approach to IS insurance. Formulation
and solution of this problem will be the subject of
our further research.
6 CONCLUSIONS
Thus the task of building trust in ELE can be
formulated on the basis of the criterion of "Three
trusts" when it is necessary to provide the
widespread trust in all three key DL elements:
WS of DL process participants;
information transfer channel between DL
process participants that will allow to introduce
an additional authentication parameter such as
biometrical characteristics for DL process
participant;
upgraded authenticity of DL process
participants including usage of biometric
methods.
Compliance with these requirements may be
achieved through the use of some biometric
characteristics or their combination. They should
have the fewest false positives and do not require
further action by the user, and additional technical
support.
The implementation of this criterion in LMS by
combining the known mechanisms of organizational
and technical IS ensuring allows to significantly
reduce the probability of having false progress
testing results and to detect messages and items non-
repudiation violation in a single TELE among all DL
process participants.
While beginning the research the authors have
already tested a variety of methods and means
supporting the implementation of the given criteria.
The results obtained are a testimony to the fact that
we can achieve different compromise levels between
openness and a desired trust level. The future work
is connected with the search of an acceptable more
rational approaches.
CSEDU2015-7thInternationalConferenceonComputerSupportedEducation
428
REFERENCES
1003.0-1995 – Guide to the POSIX(R) Open System
Environment (OSE). IEEE Standard.
Docebo.com. Discover Docebo’s Features An ecosystem
of features and modules to extend your LMS.
https://www.docebo.com/elearning-platform-saas-lms/
(last access date 22/03/2015).
Eibl, C.J., 2010. Discussion of Information Security in E-
Learning, Dissertation, Doktor der Ingenieur-
wissenschaften (Dr.-Ing.), Universität Siegen, Siegen,
Germany. http://dokumentix.ub.uni-siegen.de/opus/
volltexte/ 2010/444/pdf/eibl.pdf (last access date
22/03/2015).
Furnell, S.M., Karweni, T., 2001. Security issues in Online
Distance Learning, VINE, Vol. 31, Iss: 2, pp.28-35.
Gunasekar, K., Anirudh, C., 2011. Analysis of security
issues in cloud based e-learning. http://bada.hb.se/
bitstream/2320/9271/1/2011MAGI23.pdf (last access
date 22/03/2015).
Hameetha Begum S., Sheeba T., Nisha Rani S.N., 2013.
Security in Cloud based E-Learning. International
Journal of Advanced Research in Computer Science
and Software Engineering. Vol. 3, Iss. 1, pp. 270–278.
IITO UNESCO, 2013. Information and communication
technologies in education: Monography / Ed.
Badarcha Dendeva. Moscow, IITO UNESCO, 2013
(in Russian).
Inaba, R., Watanabe, E., Kodate, K., 2003. Security
Applications of Optical Face Recognition System:
Access Control in E-Learning. Optical Review, Vol.
10, Iss. 4, pp. 255–261.
ISO/IEC 24703:2004 — Information technology —
Participant identifiers.
Kumlander, D., 2008. Soft Biometrical Students
Identification Method for e-Learning. Advances in
Computer and Information Sciences and Engineering,
pp. 114–118.
Miloslavskaya, N., Petrov, V., Tolstoy, A., 2014. Security
Aspects for E-Learning Portals. In Proceedings of the
6th International Conference on Computer Supported
Education (CSEDU 2014). Spain, Barcelona. pp. 427–
432.
Najwa Hayaati Mohd Alwi, Ip-Shing Fan, 2010.
Information Security Threats Analysis for E-Learning.
Technology Enhanced Learning. Quality of Teaching
and Educational Reform. First International
Conference, TECH-EDUCATION 2010, Athens,
Greece, May 19-21, 2010. Proceedings. pp. 285–291.
Nickolova, M., Nickolov, E., 2007. Threat Model for User
Security in E-Learning Systems. International Journal
of Information Technologies and Knowledge, Vol. 1,
pp.341–347.
ProProfs.com. SaaS elearning platform: features &
benefits. http://www.proprofs.com/c/e-learning/saas-
elearning-platform-features-benefits/ (last access date
22/03/2015).
Siciliano, R., 2013. Distance Learning Poses Serious Data
Security Issues. http://www.huffingtonpost.com/
robert-siciliano/distance-learning-poses-
s_b_3938096.html (last access date 22/03/2015).
Ullah, A., Xiao, H., Lilley, M., Barker, T., 2012. Using
Challenge Questions for Student Authentication in
Online Examination. International Journal for
Infonomics (IJI), Vol. 5, Iss. 3/4, pp. 631–639.
Weippl, E.R., 2005. Security in E-Learning, Springer US.
Wenan Tan, Jingxian Li, Anqiong Tang, Tong Wang,
Xiaoming Hu, 2012. Trust Evaluation Model Based on
User Trust Cloud and User Capability in E-Learning
Service. Communications and Information Processing.
International Conference, ICCIP 2012 Aveiro,
Portugal, March 7-11, 2012 Revised Selected Papers,
Part I, pp. 583–590.
ProblemofTrustine-LearningEnvironment
429