International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)

Volume 4 Issue 3, March 2015

Clustering Moodle Data As a Tool

For Profiling Students

Gawali Kavita Bapu Mrs.Pratiksha R. Deshmukh

More Bharati Ashok (Guide)
Salunke Priyanka Shamrao
Sul Gauri Tanaji

Department of Computer engineering,Government college of engineering and research,

Awasari(kd), Ambegaon,pune,India

Abstract
We wish to create system that will gather
information related to e-learning courses and These e-learning systems accumulate a vast
performing analysis on the data. In this system amount of information which is very valuable for
analysis, clustering and prediction on data is analyzing student's behavior. Learning management
performed to generate results, which makes systems accumulate a great deal of log data about
monitoring data easier. The aim of our project is students activities. They can record whatever student
to use the methods of data mining, machine activities are involved, such as reading, writing,
learning to given students better tools to help taking tests, performing various tasks, and even
their e-learning courses.In this system processing communicating with peers . They normally also
of Moodle data is to be carried out.This project provide a database that stores all the system's
includes online tests, quizzes, displaying information: personal information about the users,
notifications, assignments submission ,calculating academic results, user's interaction data, etc.
grades,attendance sheet generation etc. However, due to the vast quantities of data these
systems can generate daily, it is very difficult to
Keywords:Data- manage manually, and authors demand tools which
mining,Clustering,Prediction,Machine learning assist them in this task, preferably on
a continuous basis.
Problem Definition Although some platforms offer some
reporting tools, when there are a great number of
To make monitoring easier, we will provide students, it becomes hard for a adminstrator to extract
reliable statistics, behavior groups and predicted useful information. They do not provide specific
results.In this step of the project, we are clustering tools which allow educators to thoroughly track and
students by mining Moodle log data. A first objective assess all the activities performed by their learners
is to define relevant clustering features. A second and to evaluate the structure and contents of the
objective is to determine if our students show course and its effectiveness in the learning process.
different learning behaviours. We will experiment
whether there is an overall,ideal number of clusters Proposed System
and whether the clusters show mostly qualitative or
quantitative differences. Experiments in clustering The proposed system will be generic enough
were carried out using real data obtained from to collect and analyse data from any LMS with a
various institute using Moodle platform. logging system. However, our current
implementation connects with only one LMS so far,
Existing System because in our context, all of the available data

618
ISSN: 2278 – 1323 All Rights Reserved © 2015 IJARCET
 International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
Volume 4 Issue 3, March 2015

comes from a Moodle [1], [2] platform where the

courses are located.We also want to see if we are able
to predict their end results at their exams just from
their curriculum data, which would mean we could
henceforth advise students on how they are doing.
Students basically have their own preference
in Moodle. Some students like to visit only one
module in Moodle but may be others students tend to
visit more than one module at a time. There are a few
features in Moodle that students occasionally visit
and use.
It is important to analysis this random data in future
development of Moodle and also to learn the students
learning behavior in Moodle. For example, if the data
extracted shows that the user will visit Course
module if they visit Assignment module, then the
administrator need to figure out the way to make
navigation between the modules become easier.
In the last few years, researchers have begun
to investigate various data mining methods to help
teachers improve e-learning systems. Data mining
can be applied to explore, visualize and analyze e-
learning data in order to identify useful patterns , to
evaluate web activity to get more objective feedback
for teacher's instruction and to find out more about
how the students learn etc. These methods allow us to
Fig 1.0
discover new, interesting and useful knowledge based
on student's usage data.
Features
Product Scope
The features we have selected are:
The scopes of this research are: _ Login information
i. The method use is Data Mining. _ Last login by user
ii. Data used is login information, user's information _ Time spent online by user
,login time, action and information of the action. _ Number of PDF downloaded to read later
iii. Algorithm used for clustering students's learing _ Number of quizzes, assignments, etc.done
behavior is K-means Algorithm. _Average grade Calculations
_Attendance Sheet generation
_Feedback from both student and teachers
Objectives _Displaying notifications
The main objectives on why this research is done are:
i. To study the learning behavior of the students that
visiting Moodle. Clustering
ii. To apply the Data Mining for clustering student's
learning behavior. Clustering[5] is the unsupervised grouping
iii. To develop the application in implementation of of objects into classes of similar objects. In
Data Mining in grouping Moodle data. e-learning, clustering can be used for finding clusters
iv.To develop the application that provides online of students with similar behavior patterns. In the
attendance facility. example of forums, a student can be active or a
v. To develop application which helps teachers to lurker[3] . These patterns may in turn reflect a
give advise to the student by analyzing students difference in learning characteristics, which may be
learning behavior and Moodle log's. used to give them differentiated guiding. Data mining
in general can also be used to better inform teachers
about what is going on , or to predict a student's
chance of success , which is the final aim of our
System Architecture project.

619
ISSN: 2278 – 1323 All Rights Reserved © 2015 IJARCET
 International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
Volume 4 Issue 3, March 2015

The pre-processing serves to eliminate others. We have used MySQL because it is the
useless information, select the data we want to study world's most popular open source database .
and, in our case, shape it into features.Our aim with
this analysis will be to determine if there is an overall
ideal number of clusters and whether the clusters Mathematical Model
show mostly qualitative or quantitative differences.
The clustering will be made by Weka tool, so our Set Theory
experiments will output clusters of students. Hence,
we will try to interprete the results in terms of Our system can be represented as a set
differences of behaviour between students.
System S = {I,O,C}
Preprocessing Moodle data
Where,
Moodle is an open-source learning course
I=set of inputs
management system to help educators create
effective online learning communities. Moodle is an O=set of outputs
alternative to proprietary commercial online learning
solutions, and is distributed free under open source C = set of constraints
licensing. Moodle has been installed at universities
and institutions all over the world. An organigation Input
has complete access to the source code and can make
changes if need be. Moodle can range from a single- Input I = {Login, Assignments, Academic Details}
teacher site to a 40,000 student university . Moodle's
modular design makes it easy to create new courses, Login = {Username, Password}
adding content that will engage learners. This style of
learning believes that students learn best when they Assignments = {Assignments1, Assignments2,…,
interact with the learning material, construct new Assignmentsn}
material for others, and interact with other students
about the material. Moodle does not require the use Username = {Username1, Username2,…,Usernamen
of this style in the courses but this style is what it best }
supports. Moodle keeps detailed logs of all activities
Password = {Password1, Password2,…,passwordn}
that students perform.
Logging is record keeping that can keep Output
track of what materials students have accessed.
Moodle logs every click that students make for Output O = {Result}
navigational purposes and has a modest log viewing
system built into it . Log files can be filtered by Result = {Reports, Notifications}
course, participant, day and activity.The teacher can
use these logs to determine who has been active in Constraint
the course, what they did, and when they did it. For
activities such as quizzes, not only the score and C = {C1, C2}
elapsed time are available, but also a detailed analysis
of each student's with each other. Where,
Teachers can easily get full reports of the activities of
individual students, or of all students for a specific C1 = “User should perform login operation before
activity. Activity reports for each student are using the system”.
available and details about each module as well as a
C2 = “User should have working internet
detailed story of each students involvement.
connection”.
This can be useful to check to see if
everyone has done a certain task, or spent a required
amount of time online within certain activities.
Conclusion
Moodle does not store logs as text files. Instead, it
stores the logs in a relational database. So, data are We propose to create tools that are novel
stored in a single database. MySQL is best supported, and really needed by training managers. Our
but it can also be used with Oracle, Access and application uses data mining and machine learning
methods to solve the problem of student monitoring

620
ISSN: 2278 – 1323 All Rights Reserved © 2015 IJARCET
 International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
Volume 4 Issue 3, March 2015

in e-learning obtained from Moodle courses. We

think that these features are comprehensive and
generic enough to be reused by othering.These
features are then used to conduct a clustering of the
data followed by an analysis.

References
[1] Moodle Trust, \Moodle offcial
site,"2013,http://moodle.co
[2] J. Cole and H. Foster, Using Moodle: Teaching with the
popular open source course management system. O'Reilly
Media, Inc., 2007.
[3] J. Taylor, \Teaching and Learning Online: TheWorkers,
The Lurkers and The Shirkers,"Journal of Chinese Distance
Education, no. 9, pp. 31{37, 2002.
[4] G. Cobo, D. Garc'a, E. Santamar'a, J. A. Mor'an, J.
Melench'on, and C. Monzo, \Modeling student's activity in
online discussion forums: a strategy based on time series
and agglomerative hierarchical clustering,"in Educational
Data Mining Proceedings, 2011.
[5] A. K. Jain, M. N. Murty, and P. J. Flynn, \Data
clustering: Areview,"1999.

621
ISSN: 2278 – 1323 All Rights Reserved © 2015 IJARCET