OLAP based Scaffolding to support Personalized Synchronous e-Learning

International Journal of Managing Information Technology (IJMIT) Vol.4, No.3, August 2012
DOI : 10.5121/ijmit.2012.4306 73
OLAP based Scaffolding to support Personalized
Synchronous e-Learning
Souvik Sengupta1
,Bulbul Mukherjee2
,Sudipta Bhattacharya3
,Ranajan Dasgupta4
1, 2, 3
Dept of CSE, Bengal Institute of Technology, Kolkata, India
4
Dept of CSE, National Institute of Technical Teachers’ Training and Research, Kolkata, India
1
mesouvik@hotmail.com,2
mebulbulmukherjee@gmail.com,
3
bhattacharya.sudipta@gmail.com,4
ranjandasgupta@hotmail.com
ABSTRACT
The advent of asynchronous web based learning systems has helped the learner in a self paced,
personalized and flexible learning style. It can be even more useful with a supportive synchronous tutorial
(question-answer) session. The challenge is to provide sufficient information to the instructor about the
learner’s experience in that particular course at run time. Online analytical processing (OLAP) is a very
useful technique in producing such run time information in the form of reports. In this paper we have
designed an automated scaffolding technique to hold this vital information about the learner which we have
obtained by OLAP techniques on the log data of the LMS users. We have also proposed an overall
architecture of the scaffolding where this information can be easily accessed and used by the instructor in
the synchronous tutorial session to make the system more adaptive.
KEYWORDS
OLAP, Asynchronous e-Learning, scaffolding, learner portfolio, adaptive learning;
1. Introduction
ontemporary online learning environments are characterized by a growing use of learning
management systems, which enable online access to subject matter content, asynchronous
online discussions, collaborative learning activities, and online assessment. E-learning not only
expands the scope of independent study and provides a variety of synchronous and asynchronous
learning activities.They offer tools for students to engage in asynchronous learning via web based
learning contents, generally compatible with “SCORM [7]” standard and then followed by
asynchronous interaction with the instructor for solving the problems. Instructor interacts with
students over networks and virtually face-to-face at ad hoc times and places. In a traditional class
room scenario teacher student interaction plays a vital role in the teaching learning process. The
teachers are generally aware of the basic competency level and intellect capability of the student
on the basis of day to day class performance. The benefit is that then the teacher can adjust the
level of his answer or explanation in a manner which is best suited for that particular student or
group of students with same level of competency and intellect. In e-learning environment there is
no direct scope to know the basic characteristic of a student. The concept of adaptive e-Learning
[1] enables a learner to be presented with content that matches her level of understanding but in
order to solve the personalized problems the facilitator or the instructor should be supported with
additional information about the learners. The synchronous student-facilitator collaboration was
perceived to be the effective pedagogy [8] that could result in the enhancement of the learning
quality. Our aim in this paper is to provide‘scaffolding’ that contains all the useful learner
information, to the facilitator. A threefold adaptive policy is proposed which first tries to track
C

74
learners’ behaviour from their web learning environment by going through the log data. Then we
have applied OLAP techniques to prepare different analytical reports which are used to construct
the scaffolding at runtime for a cluster of students with similar attribute values. Finally we have
proposed an architecture that adopts and use this scaffolding for aiding the students in a
personalized problem solving in e-Learning.
2. RELATED WORK
Existing LMS software like Moodle, Sakai record learner’s portfolios but they only collect
modest activity log data, without directly helping instructors with sufficient scope for analyzing
learner behaviours. Although Moodle presents several reports about the students ‘activities, but
these are according to the categories specified by the system, not the instructors. In an earlier
work by Lee.Cet al. the investigation process of the learners’ portfolio left in the e-learning
environment is shown and it adopts “data mining” techniques to establish for each cluster of
learners the most adaptive learning path pattern, which can provide a “scaffolding” to guide each
cluster of learners [3].In another work an advanced Petri Net model to analyze the workflow of a
web-based multiple participant’s virtual environment has been proposed. Scaffolding is created
where behaviours of students are supervised by an intelligent control system, which is
programmed by the instructor under a generic interface. The interface is built based on virtual
reality and real-time communication technologies [4]. Use of the scaffolding techniques these
days are also adopted in mobile-learning environments. The portability and immediate
communication properties of mobile devicesindicate the pedagogical potential of mobile devices.
In a work learners’ behaviours on a website are recorded as learning portfolios and analyzed for
behavioural diagnosis or instructional planning. A student model is then built according to the
analytical results of learning portfolios and a concept map of the learning domain [5].
3. TRACKING USER BEHAVIOUR IN ASYNCHRONOUS e-LEARNING
A rich and detailed user profile plays a key role during the learning activities. Adaptive e-learning
takes into account the profile, past behavior, preferences and needs of involved users, as well as
the characteristics of the learning environment that they exploit for these activities. In fact, it
should handle fairly rich and detailed profiles of users to play a key role in the success of the web
based learning programs. The learning portfolios of a learner are collected and stored in learning
database for further analysis to draw a nature of a learner. The following strategies are used for
this portfolio collection process:
A. User log:
Most contemporary e-Learning systems have the capability to collect, organize and report data on
learners’ activities. These may include data on time spent on a learning activity, when it was
started and completed, and number of attempts at an assessment item. It may be total time spent
on a particular course or average time spent on a particular page etc. In the e-learning system, a
learner simultaneously performs multiple learning activities. For example, learners may browse
the learning materials, while simultaneously searching for more information on a particular term.
So to correctly calculate the reading time, the system subtracts the time that a learner spends in
interaction, searching and downloading from the total time spent in the system. While a learner is
browsing a learning material, he may leave his browser on the same section and attend other
learning-related activities on different topics. Such behavior tends to be associated with a very
long reading time. Similarly, learners may browse a section find that they are not interested and
skip to other sections. In this case the reading time is very short. To prevent the reading times
from being recorded in the above cases, we propose to setlower and upper bounds on the reading

75
time. Restated, if a reading time is less than the lower bound or exceeds the upper bound, it is not
recorded in the system.
B. Learner’s preference:
The learner’s searching history within the portal and also on the web is recorded. The searching
style like searching of a particular key word repeated time implies that the place of interest is of
highest order. The navigation techniques that whether the learner skips some modules and wants
the portfolio to remember that indicates the competence level on that topic is high. Similarly
‘favourite’s’ list maintained by the learner can play a vital role to understand the learner’s desire
to use and apply the learning.
C. Learner’s interaction record:
There are three important actors in a formal e-learning context students, instructors, and content.
A runtime learning environment sees the interaction among the three actors and these information
are very important to the construction of the scaffolding.
a) Learner to Content interaction:
Today’s learning content in e-learning world is very much interactive with the user with
association of different multimedia data, animations, instructions, simulations, micro worlds,
and presentation creation tools. So, the nature of the request of content by learner, submission
of presentation and frequencies of file download and upload by a student help to draw the
graph of performance and confidence level of a learner.
b) Learner to Learner interaction:
The student-peer collaboration enabled the students to link up different ideas and to share
knowledge and it induced motivation that contributed to positive learning outcomes.
Collaborative and cooperative learning asynchronously through mail, messaging, blogging or
synchronously by chatting or video conferencing are available in modern adaptive e-learning
systems. The nature of asking questions, sharing answers and frequencies of doing these can
help to analyze a particular learner’s characteristic and interest of study.
Table 1. A diagram of six interactive modes possible
Learner Facilitator Content
Learner Cooperative
And
collaborative
learning
Asking
question,
Doubt
clearance
Self pace
learning
Facilitator Interacting
And
facilitating
Sharing
Knowledge &
experience
Making
Content Interacting
interfacing
Interacting
interfacing
Interacting
interfacing
c) Learner to Instructor interaction:
In e-learning environment one student can interact with the instructor via mail or blog. The
nature of question (e.g if it is subjective or objective), the frequencies of asking question and
the relevancy of it with the subject may judge a learner’s intellect. This type of interactions
can be arranged in a synchronous way by chatting, audio or video conferencing. In order to

76
make this more personalized sufficient information about the students’ learning behavior
should be provided to the Instructor.
D. Learning Path:
The adaptive release of content within e-learning courses enables the learner to experience an
individualized learning path. It also allows the facilitator to create rules for delivering content to
the learner upon specific tasks which leads to the achievement of the course learning objectives. It
can be implemented with prompts and cues that encourage learners to think about their learning
process and to utilize appropriate learning strategies, in essence developing the learners’ meta
cognitive skills. To improve learning performance, a “scaffolding learning paths,” is designed [3]
which provide different learning patterns based on the learning abilities of different learners, and
give learners an adaptive navigational learning map which describes the connected paths of time
sequence of a learner’s visit of the lessons of the teaching materials. Scaffolding learning theory
proposes that instructors should construct different learning scaffolding stands according to level
of ability development and learning progress of different learners [2]
E. Learner’s assessment:
In web-based e-learning system integrated assistance and assessment tools are required that offers
instruction to students while providing a detailed evaluation of their abilities by theinstructors [6].
Many studies have demonstrated that the records of student learning paths and learning
performance in the e-Learning system can be provided as a reference for instructors to evaluate
the learning accomplishments of students and diagnose their learning difficulties, hence, through
“learning portfolio monitoring” in the “intelligent e-learning system,” the investigation records
the learning path data of learners[3]. The assessment of student’s performance enables the
construction of a rich student profile that records student progress and characteristics.
Figure 1. Learner’s information tracking
4. OLAP OPERATIONS
In e-learning environments students’ usage log is the prime subject of performing analysis. The
source of data shown in this paper has come from the LMS server (Moodle) installed at Bengal
Institute of Technology for last four years. We have used an open source Business Intelligence
(BI) tool called GISMO [9] to analyse Moodle log files and to provide different information in
the form of reports and graph representations.
We have record of all the users behaviours as discussed in section 3, stored in the Moodle system
in the form of relational database. This makes our Extraction-Transformation-Load (ETL) process

77
simpler. However, amount of work required for data preparation is less, it is necessary to convert
the relational schema into multidimensional format.
Figure2. Star-Schema Architecture
We have adopted the star schema architecture and created a fact table and five dimension tables
as shown in fig 2. Our objective is to support this scaffolding by producing useful reports in
graphical formats, so that instructor can have some glimpse about the students’ context and
behaviour in the synchronous question answer or doubt clearance session.
5. DESIGNING THE SCAFFOLDING
Our proposed framework creates the scaffolding by investigating the learner’s information using
OLAP methods. Figure 3 shows how different activities are tracked from the asynchronous web
based learning environment and some consolidated reports are produced at run time. Snapshots of
some of the reports that construct the scaffolding are shown in figure 4-7.
Figure3. Proposed Scaffolding Architecture
Resource
Session
Search
Student
Time
Forum
• Resource_ID
Resource_location
:
• St_ID
Name
:
• Topic_ID
Text
:
• St_ID
• Time
:
• Date
• Time
• Srch_Id
Srch_text
:

78
The scaffolding is created only at runtime after the Learner-Instructor interactive session is
initiated. It holds the necessary information about the learner or group of learners about their
subject competence, base- knowledge, weakness, preference etc. The instructor may reframe her
answer or illustration that suits the learner most.
Use of scaffolding in synchronous system
Next we propose to use of this scaffolding for the synchronous interaction session between
learners and instructors. This can have its best result when it is used as tutorial session which is
supplemented to the self regulated learning by the learner. We propose to use different data
mining tools to classify the learners. For numerical data available from average time spent on a
web page, number of uploads and downloads, assessment score etc. the Euclidian distance based
clustering algorithms like k-medoid, DBSCAN etc. can be used, where as for non numeric data
like searching, navigation path, blog data categorical clustering algorithms like STIRR, ROCK,
CACTUS are better suited. The objective of the clustering is to find out students with similar
comprehensive level and expertise which is very helpful for the instructors in conducting live
tutorial sessions for doubt clearance and question answering.
Figure 7
Figure 4 Figure 5
Figure 6

79
Figure 8. Adaptation of Scaffolding
6. CONCLUSION
In this paper we tried to propose a framework of scaffolding which is created using OLAP
application on LMS usage data. It helps the learner as well as the instructor in the live tutorial or
question answer session in the synchronous medium like chatting, video calling, conferencing etc.
The scaffold provides useful information about the learner’s behavior based on his self regulated
learning to the instructor thus make the system more adaptive and personalized. The scaffolding
provides a purely temporary stage and is instantiated with the start of every live session of the
instructor with either a learner or group of learners. This framework is at the higher level of
abstraction, the technical details of implementing different phases of this scaffolding depends on
the choice of the designer of the implementing system, we have given only the outline of how this
can be done.
REFERENCES
[1] Razzaq.L, Heffernan.T.N, “Towards Designing a User-Adaptive Web-Based E-Learning System”
CHI,ACM 978-1-60558-012-8/08/04,Florence, Italy ,April , 2008.
[2] Yang, C.C. “A Study of the Application of Scaffolding Theory to the Acid and Alkaline Chemicals
Website of Primary School.” Master thesis, Graduate School of Applied Chemistry, Providence
University, 2000.
[3] Lee.C.H,Lee.G.G,YunghoLeu.Y, “Analysis on the Adaptive Scaffolding Learning Path and the
Learning Performance of e-Learning ”, WSEAS TRANSACTIONS on INFORMATION SCIENCE &
APPLICATIONS, ISSN: 1790-0832, Issue 4, Volume 5,pp,320-330, April 2008.
[4] 4. Yen.S.H,Lawrence .Y, Deng.L.Y, Chen.Y.H,“Scaffolding for Activity Supervision and Self-
Regulation in Virtual University,Tamkang” Journal of Science and Engineering, Vol. 8, No 2, pp.
133_146 ,2005.
[5] Chen.G.D,Chang.C.K,Wang.C.Y,“Ubiquitous learning website: Scaffold learners by mobile devices
with information-aware techniques”, Computers & Education 50,pp, 77–90,2008.

[6] Razzaq, Heffernan, Koedinger, Feng, Nuzzo
“Blending Assessment and Instructional Assistance. In Nadia Nedjah, LuizadeMacedoMourelle, Mario
Neto Borges and NivalNunesde Almeida (Eds). Intelligent Educational Machines within the Intelligent
Systems Engineering Book Series.Spr
[7] Shareable Content Object Reference Model (SCORM) U.S Government ©Advanced Distributed
Learning. URL: http://www.adlnet.gov/scorm
[8] FauziahSulaiman,HanafiAtan, Rozhan M Idrus&HishamDzakiria
of the Web-Based Synchronous
(MOJIT) Vol. 1, No. 2, pp 58-66 December 200
[9] GIMSO.URL: http://sourceforge.net/projects/gismo/files/GISMO%20for%20moodle%201.9.x/
Authors
Souvik Sengupta is an Assistant Professor in the Computer Science De
of Bengal Institute of Technology, India . He has obtained his Master of
Technology form West Bengal University of Technology, India in 2008. He is
currently performing his research work under the PhD program of the
Department of Information Technology of the university of Calcutta ,India. He
has 10 international journal and conference papers published in last 3 years. He
has also served as a visiting faculty in the M.Tech program of Department of
Computer Science of National Institute of Technical
Research, Kolkata, India. His areas of research interests include Web
Learning, Formal software engineering, Multimedia Database, Data mining, Software engineering,
Artificial Intelligence etc.
Bulbul Mukherjee is in Computer Science Department of Bengal Institute of
Technology, India; her research interest is in Software Engineering, Distributed
Database, Artificial Intelligence, E
Technology in Multimedia & System Software from Wes
Technology, India. She has one paper in ‘Petri
National Conference ETEI-2012 and one paper in ‘Modeling Distributed Active
Database Using CTPN’ published in 12th WSEAS Int. Conf. On Applied
informatics And Communications (Aic'12).
Sudipta Bhattacharya is working as Assistant Professor in the Department of
Computer Science and Information Technology at Bengal Institute of Technology,
Kolkata, India. He received his Bachelor of .Technology,(IT) from West B
University of Technology, India , Master of Technology(IT) from Bengal
Engineering and Science University, Shibpur, India. He has 5 International
Conference Papers and one International Journal Paper published in last 3 years.
His area of research interest Data mining, System &Internet Security and
Networking, E-learning, Artificial Intelligence
Ranjan Dasgupta is Professor & Head, Dept. of CSE. National Institute of
Technical Teachers’ Training & Research, Kolkata, India, Previously he was with
Jadavpur University, and also served several Indian
He received his B. Tech from,Dept. of Radio Physics & Electronics, M. Tech, &
Ph.D from Dept. of Computer Science, University of Calcutta,India. His areas of
research interests include software,engineering, databases, GIS, d
computing. Dr.Dasgupta has published around 30 research papers in different
International Conferences & Journals. He has also served several other
Universities & national & international bodies, World Bank Assisted Projects in various capacities
Razzaq, Heffernan, Koedinger, Feng, Nuzzo-Jones, Junker, Macasek, Rasmussen, Turner &Walonoski
es and NivalNunesde Almeida (Eds). Intelligent Educational Machines within the Intelligent
Systems Engineering Book Series.SpringerBerlin / Heidelberg. pp, 23-49,2007.
Shareable Content Object Reference Model (SCORM) U.S Government ©Advanced Distributed
http://www.adlnet.gov/scorm
FauziahSulaiman,HanafiAtan, Rozhan M Idrus&HishamDzakiria, “Problem-Based Learning: A Study
Based Synchronous Collaboration”, Malaysian Online Journal of Instructional Technology
66 December 2004.
http://sourceforge.net/projects/gismo/files/GISMO%20for%20moodle%201.9.x/
Souvik Sengupta is an Assistant Professor in the Computer Science Department
of Bengal Institute of Technology, India . He has obtained his Master of
Technology form West Bengal University of Technology, India in 2008. He is
currently performing his research work under the PhD program of the
ology of the university of Calcutta ,India. He
has 10 international journal and conference papers published in last 3 years. He
has also served as a visiting faculty in the M.Tech program of Department of
Computer Science of National Institute of Technical Teachers’ Training and
Research, Kolkata, India. His areas of research interests include Web–based
Computer Science Department of Bengal Institute of
Technology, India; her research interest is in Software Engineering, Distributed
Database, Artificial Intelligence, E-learning. She is pursuing Master of
Technology in Multimedia & System Software from West Bengal University of
Technology, India. She has one paper in ‘Petri-Net modeling’ published in a
2012 and one paper in ‘Modeling Distributed Active
Database Using CTPN’ published in 12th WSEAS Int. Conf. On Applied
d Communications (Aic'12).
Sudipta Bhattacharya is working as Assistant Professor in the Department of
Computer Science and Information Technology at Bengal Institute of Technology,
Kolkata, India. He received his Bachelor of .Technology,(IT) from West Bengal
University of Technology, India , Master of Technology(IT) from Bengal
Engineering and Science University, Shibpur, India. He has 5 International
Conference Papers and one International Journal Paper published in last 3 years.
terest Data mining, System &Internet Security and
learning, Artificial Intelligence
Ranjan Dasgupta is Professor & Head, Dept. of CSE. National Institute of
Technical Teachers’ Training & Research, Kolkata, India, Previously he was with
Jadavpur University, and also served several Indian IT companies, for five years.
He received his B. Tech from,Dept. of Radio Physics & Electronics, M. Tech, &
Ph.D from Dept. of Computer Science, University of Calcutta,India. His areas of
research interests include software,engineering, databases, GIS, distributed
computing. Dr.Dasgupta has published around 30 research papers in different
International Conferences & Journals. He has also served several other
Universities & national & international bodies, World Bank Assisted Projects in various capacities
80
Jones, Junker, Macasek, Rasmussen, Turner &Walonoski
es and NivalNunesde Almeida (Eds). Intelligent Educational Machines within the Intelligent
Shareable Content Object Reference Model (SCORM) U.S Government ©Advanced Distributed
Based Learning: A Study
Online Journal of Instructional Technology
http://sourceforge.net/projects/gismo/files/GISMO%20for%20moodle%201.9.x/
Universities & national & international bodies, World Bank Assisted Projects in various capacities.

OLAP based Scaffolding to support Personalized Synchronous e-Learning

More Related Content

What's hot (18)

Similar to OLAP based Scaffolding to support Personalized Synchronous e-Learning (20)

More from IJMIT JOURNAL (20)

Recently uploaded (20)

OLAP based Scaffolding to support Personalized Synchronous e-Learning