ArticlePDF Available

Education data mining with Moodle 2.4

Authors:

Abstract

The goal of e-learning environments is to supply effective learning methods, to enable the users to approach certain resources at any time, to set solutions for certain problems, assessment for the work etc. One of the best known environments of this kind is e-learning system Moodle. These environments like Moodle use and save large amount of data in their databases, but in most cases they don't offer enough information of the course participants and their activities in the system. The aim of this work is, by the use of data mining techniques such as classification, clustering, statistics and regression, to describe the process of selection and acquiring data from the Moodle database, and to create dashboard - web based application, that would communicate with the e-learning system Moodle and supply multilevel approach as: manager, administrator, teacher and user level; and practically will improve the approach to evaluation of larger groups of participants in the learning process. This will help teachers to evaluate web activity of the students, to get more objective feedback and find out more about how the students learn. Also this dasboard will directly solve the teachers problems in the terms of dealing with this kind of platforms and big amounts of data.
105
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, УниверзитетГоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
ɍɇɂȼȿɊɁɂɌȿɌ ȽɈɐȿ ȾȿɅɑȿȼ” - ɒɌɂɉ
ɎȺɄɍɅɌȿɌ ɁȺ ɂɇɎɈɊɆȺɌɂɄȺ
ȽɈȾɂɇȺ 2 VOLUME II
GOCE DELCEV UNIVERSITY - STIP
FACULTY OF COMPUTER SCIENCE
UDC ISSN
ȽɈȾɂɒȿɇ ɁȻɈɊɇɂɄ
2013
YEARBOOK
2013
ISSN:1857-8691
УНИВЕРЗИТЕТ „ГОЦЕ ДЕЛЧЕВ“ – ШТИП
ФАКУЛТЕТ ЗА ИНФОРМАТИКА
GOCE DELCEV UNIVERSITY – STIP
FACULTY OF COMPUTER SCIENCE
ГОДИНА 2 VOLUME IIМАРТ, 2014
ГОДИШЕН ЗБОРНИК
2013
YEARBOOK
2013
2
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
ГОДИШЕН ЗБОРНИК
ФАКУЛТЕТ ЗА ИНФОРМАТИКА
YEARBOOK
FACULTY OF COMPUTER SCIENCE
Издавачки совет
Проф. д-р Саша Митрев
Проф. д-р Лилјана Колева - Гудева
Проф. д-р Владо Гичев
Проф. д-р Цвета Mартиновска
Проф. д-р Татајана Атанасова - Пачемска
Доц. д-р Зоран Здравев
Доц. д-р Александра Милева
Доц. д-р Сашо Коцески
Доц. д-р Наташа Коцеска
Доц. д-р Зоран Утковски
Доц. д-р Игор Стојановиќ
Доц. д-р Благој Делипетров
Редакциски одбор
Проф. д-р Цвета Mартиновска
Проф. д-р Татајана Атанасова - Пачемска
Доц. д-р Наташа Коцеска
Доц. д-р Зоран Утковски
Доц. д-р Игор Стојановиќ
Доц. д-р Александра Милева
Доц. д-р Зоран Здравев
Главен и одговорен уредник
Доц. д-р Зоран Здравев
Јазично уредување
Даница Гавриловаска - Атанасовска
(македонски јазик)
Павлинка Павлова-Митева
(англиски јазик)
Техничко уредување
Славе Димитров
Благој Михов
Редакција и администрација
Универзитет ,,Гоце Делчев“-Штип
Факултет за информатика
ул. ,,Крсте Мисирков“ 10-A
п. фах 201, 2000 Штип
Р. Македонија
Editorial board
Prof. Saša Mitrev, Ph.D
Prof. Liljana Koleva - Gudeva, Ph.D.
Prof. Vlado Gicev, Ph.D.
Prof. Cveta Martinovska, Ph.D.
Prof. Tatjana Atanasova - Pacemska, Ph.D.
Ass. Prof. Zoran Zdravev, Ph.D.
Ass. Prof. Aleksandra Mileva, Ph.D.
Ass. Prof. Saso Koceski, Ph.D.
Ass. Prof. Natasa Koceska, Ph.D.
Ass. Prof. Zoran Utkovski, Ph.D.
Ass. Prof. Igor Stojanovik, Ph.D.
Ass. Prof. Blagoj Delipetrov, Ph.D.
Editorial staff
Prof. Cveta Martinovska, Ph.D.
Prof. Tatjana Atanasova - Pacemska, Ph.D.
Ass. Prof. Natasa Koceska, Ph.D.
Ass. Prof. Zoran Utkovski, Ph.D.
Ass. Prof. Igor Stojanovik, Ph.D.
Ass. Prof. Aleksandra Mileva, Ph.D.
Ass. Prof. Zoran Zdravev, Ph.D.
Managing/ Editor in chief
Ass. Prof. Zoran Zdravev, Ph.D.
Language editor
Danica Gavrilovska-Atanasovska
(macedonian language)
Pavlinka Pavlova-Miteva
(english language)
Technical editor
Slave Dimitrov
Blagoj Mihov
Address of the editorial ofce
Goce Delcev University – Stip
Faculty of Computer Science
Krste Misirkov 10-A
PO box 201, 2000 Štip,
R. of Macedonia
За издавачот:
Проф д-р Владо Гичев
3
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
СОДРЖИНА
CONTENT
CALCULATION OF MULTI-STATE TWO TERMINAL RELIABILITY
Natasha Stojkovic, Limonka Lazarova and Marija Miteva .................................................5
INCREASING THE FLEXIBILITY AND APPLICATION OF THE
B- SPLINE CURVE
Julijana Citkuseva, Aleksandra Stojanova, Elena Gelova ..................................................11
WAVELET APPLICATION IN SOLVING ORDINARY
DIFFERENTIAL EQUATIONS USING GALERKIN METHOD
Jasmina Veta Buralieva, Sanja Kostadinova and Katerina Hadzi-Velkova Saneva .......17
ПРОИЗВОДИ НА ДИСТРИБУЦИИ ВО КОЛОМБООВА АЛГЕБРА
Марија Митева, Билјана Јолевска-Тунеска, Лимонка Лазарова ...............................27
ПРИМЕНА НА МЕТОДОТ CRANK-NICOLSON ЗА РЕШАВАЊЕ НА
ТОПЛИНСКИ РАВЕНКИ
Мирјана Коцалева, Владо Гичев ........................................................................................35
S-BOXES – PARAMETERS, CHARACTERISTICS AND CLASSIFICATIONS
Dusan Bikov, Stefka Bouyuklieva and Aleksandra Stojanova ............................................47
ПРЕБАРУВАЊЕ ИНФОРМАЦИИ ВО ЕРП СИСТЕМИ:
АРТАИИС СТУДИЈА НА СЛУЧАЈ
Ѓорѓи Гичев, Ана Паневска, Ивана Атанасова, Зоран Здравев,
Цвета Мартиновска-Банде, Јован Пехчевски .................................................................53
ЕДУКАТИВНО ПОДАТОЧНО РУДАРЕЊЕ СО MOODLE 2.4
Зоран Милевски, Зоран Здравев .........................................................................................65
ПРЕГЛЕД НА ТЕХНИКИ ЗА ПРЕПОЗНАВАЊЕ НА ЛИК ОД ВИДЕО
Ана Љуботенска, Игор Стојановиќ ....................................................................................77
ИНТЕРНЕТ АПЛИКАЦИЈА ЗА ОБРАБОТКА НА СЛИКИ
СО МАТРИЧНИ ТРАНСФОРМАЦИИ
Иван Стојанов, Ана Љуботенска, Игор Стојановиќ, Зоран Здравев ..........................85
УТАУТ И НЕЈЗИНАТА ПРИМЕНА ВО ОБРАЗОВНА СРЕДИНА:
ПРЕГЛЕД НА СОСТОЈБАТА
Мирјана Коцалева, Игор Стојановиќ, Зоран Здравев ...................................................95
4
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
CALCULATION OF MULTI-STATE TWO TERMINAL
RELIABILITY
Natasha Stojkovic1, Limonka Lazarova2 and Marija Miteva3
1Faculty of Computer Science, “Goce Delcev” University Stip
(natasa.maksimova, limonka.lazarova, marija.miteva)@ugd.edu.mk
Abstract. Traditionally, reliability of the transportation system has been analyzed from
a binary perspective. It is assumed that a system and its components can be in either a
working or a failed state. But, many transportation systems as: telecommunication
systems, water distribution, gas and oil production and hydropower generation systems
are consisting of elements that may operate in more than two states. The problem that
we consider in this paper is known as the multi-state two terminal reliability computation.
The multi state two terminal reliability can be computed with the formula of inclusion
and exclusion, if the minimal path vector or minimal cut vector are known.
Keywords: multi-state systems, network reliability, minimal path vectors, minimal cut
vectors.
1 Introduction
Two-terminal network reliability for binary transportation system has been
studied in various ways. For the binary network it is assumed that a whole
system and its components can be in two states: working or failed state.
However, the binary approach does not completely describe some
transportation systems. Such systems are telecommunication systems, water
distribution, gas and oil production and hydropower generation systems. These
networks and its components may operate in any of several intermediate states
and better results may be obtained using a multi-state reliability approach.[1]
The authors developed a multi-state approach for exact computation of multi-
state two-terminal reliability at demeaned level d (M2TRd). The multi-state two
terminal reliability is defined as the probability that a demand of d units can be
transmitted from source to sink nodes through multi-state edges [2]. The multi
state two terminal reliability can be computed if the minimal path vector or
minimal cut vectors are known. In the literature many algorithms for calculating
on minimal path or cut vectors are known.
Some algorithms for obtaining minimal path or cut vectors are given in [1],
[2], [3] and [4]. In [1] is developed a multi-state approach for exact computation
of multi-state two-terminal reliability. In the paper is proposed algorithm for
obtaining minimal path vector. Disadvantage of this algorithm is that it gives
candidates minimal path vectors that are not minimal. In [2] is proposed
algorithm for obtaining minimal cut vectors for the multi-state two-terminal
transportation system. The disadvantage of this algorithm is that it works only
for weak homogeneous components. The components can have different
number of state, but the first state of the components has to be the same. In [3]
65
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
(лектор)
ЕДУКАТИВНО ПОДАТОЧНО РУДАРЕЊЕ СО MOODLE 2.4
Зоран Милевски1, Зоран Здравев1
1 Факултет за информатика, Универзитет „Гоце Делчев“ - Штип
milevskiz@gmail.com, zoran.zdravev@ugd.edu.mk
Апстракт
Околините за е-учење имаат за цел да обезбедат
ефикасни методи за учење, да овозможат корисниците во кое
било време да пристапат кон одредени ресурси, да постават
решенија за одредени проблеми, да бидат оценети за нивниот
труд и сл. Една од попознатите такви околини е системот за е-
учење Moodle. Овие околини како Moodle користат и складираат
големи количини на податоци, но во повеќето случаи не
задоволуваат поголем дел од барањата за нивна примена и
недоволно ја прикажуваат активноста на учесниците при учењето.
Целта на овој труд е со помош на техники за податочно рударење
како што се класификација, кластерирање, статистики и
регресија, да се опише процесот на селекција и добивање на
податоци од базата на податоци на Moodle и да се креира
контролна табла веб базирана апликација што ќе комуницира со
системот за е-учење Moodle и ќе обезбедува неколку нивоа на
пристап и тоа: менаџерско, администраторско, наставничко и
корисничко ниво, и практично ќе прикажува обработени податоци
и извештаи кои ќе го подобрат пристапот на евалуација на
поголеми групи на учесници во процесот на учење. Со тоа
директно се решава и проблемот на наставниците во поглед на
нивната поддршка при работа со ваков тип на платформи и
големи количини на податоци.
Клучни зборови: далечинско учење, е-учење, едукативно
податочно рударење, Moodle, едукативни контролни табли,
извештаи во повеќе нивоа.
EDUCATION DATA MINING WITH MOODLE 2.4
Zoran Milevski1, Zoran Zdavev1
Faculty of computer science, Goce Delcev University, Stip,
Macedonia
milevskiz@gmail.com, zoran.zdravev@ugd.edu.mk
Abstract
The goal of e-learning environments is to supply effective
learning methods, to enable the users to approach certain resources
at any time, to set solutions for certain problems, assessment for the
work etc. One of the best known environments of this kind is e-
66
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
learning system Moodle. These environments like Moodle use and
save large amount of data in their databases, but in most cases they
don't offer enough information of the course participants and their
activities in the system. The aim of this work is, by the use of data
mining techniques such as classification, clustering, statistics and
regression, to describe the process of selection and acquiring data
from the Moodle database, and to create dashboard - web based
application, that would communicate with the e-learning system
Moodle and supply multilevel approach as: manager, administrator,
teacher and user level; and practically will improve the approach to
evaluation of larger groups of participants in the learning process.
This will help teachers to evaluate web activity of the students, to get
more objective feedback and find out more about how the students
learn. Also this dashboard will directly solve the teachers problems
in the terms of dealing with this kind of platforms and big amounts of
data.
Keywords: Distance Education, E-learning, Educational data
mining, Moodle, Educational Dashboard, Multilevel reports.
1. Introduction
Web-based educational systems and their usage has increased rapidly in the
last few years. The impact on this trend comes from the fact that neither teachers
nor students are limited any longer to be at the same time on the same location,
and additionally these online education-based systems are independent from any
hardware platforms [6]. The approach to these platforms is only through internet
browser and thus the dependence on different operative systems and their
demands is neutralized. These educational systems have been installed in many
universities, and even individual teachers use them with a goal of setting certain
resources that will be easily approachable for certain groups of people.
Moodle (Modular Object Oriented Developmental Learning Environment) as
an educational system is well known and widely used because it is open code
and also satisfies greater part of the needs for its use, and it is also simple to use
both for the teachers and the students as course participants [1], [2], [3]. Moodle
accumulates great amount of different information that is very important when
analyzing the students conduct and represents a gold mine of educational data.
Moodle stores all the data of the activities in which the students are involved.
Moodle also keeps data of the participants profiles, their activity in different
courses, their sent assignments etc.
The e-learning system Moodle is used in 232 countries in the world and at the
moment there are 79429 active Moodle web sites, whereas in Macedonia there
are 39 Moodle web sites most of which belong to the universities in the country
[14].
Although Moodle, as well as the other systems of this type, offers tools for
reports and view of the more important activities of the course participants, when
it comes to a bigger number of students it becomes hard to follow their activity.
On the other hand, although the goal of the e-learning systems is to motivate the
67
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
student by the use of multimedia materials to make studying more interesting and
of higher quality, they don’t always succeed in keeping the student’s focus on the
learning itself. Instead of learning they use the opportunities that the system offers
in the direction of social communication between them (chat) [5]. Whatsoever, to
make studying more effective, it is important to supply personalization of the
contestants, based on their activity, an opportunity to analyze the participants in
different courses, prediction of the results of the participants and better survey of
the activities of the students. A promising area, when it comes to fulfilling this goal
is data mining, and in this case it is educational data mining with the Moodle 2.4
database [3], [6].
Educational data mining means selective extraction of the kept data of large
databases, their processing with the use of several educational techniques of
data mining such as classification, clustering, statistics, regression etc. and
acquiring the processed data that would improve the approach to larger groups
of participants in the learning process [7]. The acquired information can be used
not only by the teachers, but the students themselves too. They can get
recommendations and directions for certain activities and resources that would
improve their learning, where the teacher can get the feedback necessary for the
evaluation of the students activity, separating the students in groups based on
the need for their monitoring, finding the frequent mistakes made both by the
students and the teachers; view into the activities assigned for the students; and
have greater effect than the others [6].
Web based application connected to the active Moodle database can provide
several levels of approach:
Manager level approach,
Administrator level approach,
Teacher level approach and
User level approach.
All of these roles are included in the, so called, dashboard and they will be
explained in this study.
This dashboard is external application made to be very easy to use, because
all needed reports for the various users are simplified in one place and enable a
survey of some information kept in the new dashboard database, while the
standard reports don’t give such view with the use of standard reports.
2. Data analyses in the Moodle database
It is very important what kind of data is kept in the database and the more data
is processed the more information can be acquired. Figure 1 illustrates the
process of educational data mining and the way this process works [2].
68
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
answer, which is the question that was answered by the smallest number of
students, analyses of the results of one student or all the students in several
quizzes etc. [4], [13].
Moodle doesn’t keep these records as text but keeps them in relational MySql
database. Moodle database has around 145 related tables, but all of them are
not necessary to implement educational data mining. Table 1 schemes the more
important tables with their description that would be used for getting raw data so
later be processed with several different techniques [6].
Табела 1. Поважни табели во базата на податоци на Moodle и нивни опис
Table 1. More important tables in Moodle database and their description
Table name
Table description
mdl_assign
Homework data
mdl_assign_submission
Sent homework status for a
student
mdl_assign_grades
mdl_choise, mdl_answers и
mdl_option
Homework score
View of questionnaire
answers
mdl_course
Created courses data
mdl_enrol
Type of log on, course
logging password
mdl_forum и
mdl_forum_discussions
Created course forums and
survey discussions started by a
certain user
mdl_posts
mdl_read
mdl_grades и
mdl_grade_letters
mdl_lesson
mdl_log
mdl_message и
mdl_message_read
mdl_question
mdl_quiz
mdl_user
mdl_enrolments
mdl_user_lastaccess
Forum answers with date,
user, message,
Forum participants activity
Summarizing participant
grades and score criteria and
final grade
Set lessons data
Log for every student action
Survey of sent messages
Question base for the quizzes
with answers
Quiz questions, answers
All users information
Enrolled users in a certain
course
participant’s last course
access
Data preprocessing provides data to be transformed in relevant format for data
mining to be applied. Before using data mining it is important to identify the
necessary user, the course he is enrolled etc [9], [10].
Слика 1. Процес на едукативно податочно рударење
Figure 1. Process of educatonal data mining
Educational data mining is an interactive process in which not only the
processed data can be acquired, but it can also be filtered so that a certain
decision can be made. The process consists of gathering information about the
students’ interaction within the process, than data processing so that they can be
transformed into a relevant format to be mined. Data mining is applied, i.e.
algorithms are used that provide and summarize the acquired interests about a
certain user (teacher, student, manager etc.). Finally, the results are interpreted,
evaluated and represented [2], [6].
Слика 2. Приказ на Moodle извештај
Figure 2 Moodle log report screen
Moodle stores every click of the user and its system navigation. Figure 2 shows
a scheme of modest report record of Moodle about the site activities. Records
can be filtered by course, participant, data and type of activity [15]. Teachers can
use this report to follow the course participants activity, what they do and when.
For activities such as quizzes the report contains data about the results, the time
length of the quiz activity, as well as detailed analyses of every answer of the
student. These reports are useful, but at the same time they are not clear enough.
For a more effective view for the teacher, besides course activity it is important
to be able to see which of the activities attracted greatest attention, which is the
least visited material, in the quiz section, besides detailed analyses of every
69
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
answer, which is the question that was answered by the smallest number of
students, analyses of the results of one student or all the students in several
quizzes etc. [4], [13].
Moodle doesn’t keep these records as text but keeps them in relational MySql
database. Moodle database has around 145 related tables, but all of them are
not necessary to implement educational data mining. Table 1 schemes the more
important tables with their description that would be used for getting raw data so
later be processed with several different techniques [6].
Табела 1. Поважни табели во базата на податоци на Moodle и нивни опис
Table 1. More important tables in Moodle database and their description
Table name
Table description
mdl_assign
Homework data
mdl_assign_submission
Sent homework status for a
student
mdl_assign_grades
mdl_choise, mdl_answers и
mdl_option
Homework score
View of questionnaire
answers
mdl_course
Created courses data
mdl_enrol
Type of log on, course
logging password
mdl_forum и
mdl_forum_discussions
Created course forums and
survey discussions started by a
certain user
mdl_posts
mdl_read
mdl_grades и
mdl_grade_letters
mdl_lesson
mdl_log
mdl_message и
mdl_message_read
mdl_question
mdl_quiz
mdl_user
mdl_enrolments
mdl_user_lastaccess
Forum answers with date,
user, message,
Forum participants activity
Summarizing participant
grades and score criteria and
final grade
Set lessons data
Log for every student action
Survey of sent messages
Question base for the quizzes
with answers
Quiz questions, answers
All users information
Enrolled users in a certain
course
participant’s last course
access
Data preprocessing provides data to be transformed in relevant format for data
mining to be applied. Before using data mining it is important to identify the
necessary user, the course he is enrolled etc [9], [10].
Слика 1. Процес на едукативно податочно рударење
Figure 1. Process of educatonal data mining
Educational data mining is an interactive process in which not only the
processed data can be acquired, but it can also be filtered so that a certain
decision can be made. The process consists of gathering information about the
students’ interaction within the process, than data processing so that they can be
transformed into a relevant format to be mined. Data mining is applied, i.e.
algorithms are used that provide and summarize the acquired interests about a
certain user (teacher, student, manager etc.). Finally, the results are interpreted,
evaluated and represented [2], [6].
Слика 2. Приказ на Moodle извештај
Figure 2 Moodle log report screen
Moodle stores every click of the user and its system navigation. Figure 2 shows
a scheme of modest report record of Moodle about the site activities. Records
can be filtered by course, participant, data and type of activity [15]. Teachers can
use this report to follow the course participants activity, what they do and when.
For activities such as quizzes the report contains data about the results, the time
length of the quiz activity, as well as detailed analyses of every answer of the
student. These reports are useful, but at the same time they are not clear enough.
For a more effective view for the teacher, besides course activity it is important
to be able to see which of the activities attracted greatest attention, which is the
least visited material, in the quiz section, besides detailed analyses of every
70
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
cluster 1 and more than the students in cluster 0. In this way the teacher can use
these information so that he can divide the students into groups of different type
of students for example at least one student from cluster 1 and students from the
other clusters or a group of students from cluster 1 who would work on problem
assignments of higher degree than the others [6], [12].
Classification of participants is used to discover potential students with similar
characteristics for a definite specific pedagogical strategy, to predict the final
results for a group of students, even to identify the students who need motivation
to get better results.
We divide the students into bad, good and excellent by generating decision
trees that involve certain classification rules. Our goal is to classify students in
different groups depending on their activity in Moodle. Table 2 represents the
knowledge by decision tree with if-else rules. This process goes on until all data
are classified perfectly or we run out of attributes. Students with lower number of
passed quizzes are classified as weak students, students with bigger number of
quizzes are classified as excellent and the students with an average number of
quizzes as good and of course taking into account the total time spent on
resources and activities, the number of sent homework assignments etc. [6], [9],
[11].
Табела 2. Множество правила генерирани од одлучувачко дрво
Table 2. Rule set generated by Decision Tree
if(n_quiz=low) then mark=bad
else if(n_quiz=medium) then {
if(total_time=low) then {
if(view_resource=low) then mark=bad
else if view_resource =medium) then {
if(forum_post=low) then mark=bad
else if(view_resource=medium) then {
if(total_assigments=high) then
mark=good
else if(overall_core=high) then
mark=excellent
}
else if(total_time =medium) then {
if(view_resource=low) then mark=bad
else if(view_resource=medium) then mark=good
else if(view_resource=high) then
mark=excellent
if(overall_score=good) then
if(forum_post==good) then
mark=excellent
}
Teachers can use this information from these rules to get an overview of the
course activity and the classification of the course participants. For example, it is
obvious that the main discriminator in this case are the successfully realized
For example, to show the number of resources (figure 3) there is created new
i.e. warehouse that is linked with the original Moodle database and it is filled with
data in sertain time period.
Слика 3. Страница за ресурси во контролната табла
Figure 3. Resources page in the dashboard
To get certain reports it is important to analyze several tables from the database
so that a summary of the system activities can be provided, to get user-friendly
results scheme. That is the aim of this research, with which we consider that
system users (all within their own role) will get view in their activities.
3. Data mining of analyzed data
Besides analyzing the data in the Moodle database, it is very important how
the data will be grouped in order to achieve the required effect.
For that purpose we hold up on data mining and we use some of the known
techniques that can provide us all necessary information and data in the effort to
give the teacher simplified view on the processed knowledge.
In e-learning systems clustering can be useful for finding similar characteristics
students clusters, revealing the user conduct and grouping the students into
several groups: students who are active in the system, discuss in forums, send
homework, spend some time in the system in checking different contents etc. [6].
In this research we will divide students into three clusters as follows: cluster 0
(inactive), cluster 1 (very active), and cluster 2 (active course participants).
Cluster 0 is characterized by students who haven’t sent homework, have read
only few messages, took only few quizzes and spent very little time in checking
the resources, activities and forum participation. Cluster 1 is characterized by
students who have sent at least one message in the forum, have read at least
three messages, have passed successfully at least half of the quizzes and have
finished less than half of them unsuccessfully and have high score and grades.
Cluster 2 is characterized by students who have lower score than students in
71
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
cluster 1 and more than the students in cluster 0. In this way the teacher can use
these information so that he can divide the students into groups of different type
of students for example at least one student from cluster 1 and students from the
other clusters or a group of students from cluster 1 who would work on problem
assignments of higher degree than the others [6], [12].
Classification of participants is used to discover potential students with similar
characteristics for a definite specific pedagogical strategy, to predict the final
results for a group of students, even to identify the students who need motivation
to get better results.
We divide the students into bad, good and excellent by generating decision
trees that involve certain classification rules. Our goal is to classify students in
different groups depending on their activity in Moodle. Table 2 represents the
knowledge by decision tree with if-else rules. This process goes on until all data
are classified perfectly or we run out of attributes. Students with lower number of
passed quizzes are classified as weak students, students with bigger number of
quizzes are classified as excellent and the students with an average number of
quizzes as good and of course taking into account the total time spent on
resources and activities, the number of sent homework assignments etc. [6], [9],
[11].
Табела 2. Множество правила генерирани од одлучувачко дрво
Table 2. Rule set generated by Decision Tree
if(n_quiz=low) then mark=bad
else if(n_quiz=medium) then {
if(total_time=low) then {
if(view_resource=low) then mark=bad
else if view_resource =medium) then {
if(forum_post=low) then mark=bad
else if(view_resource=medium) then {
if(total_assigments=high) then
mark=good
else if(overall_core=high) then
mark=excellent
}
else if(total_time =medium) then {
if(view_resource=low) then mark=bad
else if(view_resource=medium) then mark=good
else if(view_resource=high) then
mark=excellent
if(overall_score=good) then
if(forum_post==good) then
mark=excellent
}
Teachers can use this information from these rules to get an overview of the
course activity and the classification of the course participants. For example, it is
obvious that the main discriminator in this case are the successfully realized
For example, to show the number of resources (figure 3) there is created new
i.e. warehouse that is linked with the original Moodle database and it is filled with
data in sertain time period.
Слика 3. Страница за ресурси во контролната табла
Figure 3. Resources page in the dashboard
To get certain reports it is important to analyze several tables from the database
so that a summary of the system activities can be provided, to get user-friendly
results scheme. That is the aim of this research, with which we consider that
system users (all within their own role) will get view in their activities.
3. Data mining of analyzed data
Besides analyzing the data in the Moodle database, it is very important how
the data will be grouped in order to achieve the required effect.
For that purpose we hold up on data mining and we use some of the known
techniques that can provide us all necessary information and data in the effort to
give the teacher simplified view on the processed knowledge.
In e-learning systems clustering can be useful for finding similar characteristics
students clusters, revealing the user conduct and grouping the students into
several groups: students who are active in the system, discuss in forums, send
homework, spend some time in the system in checking different contents etc. [6].
In this research we will divide students into three clusters as follows: cluster 0
(inactive), cluster 1 (very active), and cluster 2 (active course participants).
Cluster 0 is characterized by students who haven’t sent homework, have read
only few messages, took only few quizzes and spent very little time in checking
the resources, activities and forum participation. Cluster 1 is characterized by
students who have sent at least one message in the forum, have read at least
three messages, have passed successfully at least half of the quizzes and have
finished less than half of them unsuccessfully and have high score and grades.
Cluster 2 is characterized by students who have lower score than students in
72
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
Слика 5. Поминато време во системот од извештајот за корисничка
активност
Figure 5. Time spent in the system from the user activity report
The course menu schemes a list of courses in categories and a total number
of, list of all courses and an opportunity to choose a certain course. After a course
is chosen a view of the number of discussions without content display is acquired,
but a view of the activities of the discussion participants, number of started topics,
number of theme answers. A list of all resources and access to all of them with
additional details for most visited and least visited resource. A survey of all the
activities such as homework and its assessment with a list of students with
highest and lowest score is also available. This part also offers a survey of all
questionnaires and their results as well as a possibility to print the questionnaire
results.
User portfolio enables individual course users’ data preview, as well as table
preview for all the course participants by viewing the activity (inactive, active, very
active), division in categories according to the assessment into bad, good and
excellent, prediction of whether they will complete the course successfully or not
and complete summary.
The user portfolio view gives contrastive analysis of a student’s results in
several courses and all the activities that characterize the student.
b. Administrator level acquired data
The administrator is the user of the application who has all the privileges. The
application administrator has the overall view of the user from the manager level.
There is an additional opportunity to give tasks in precisely defined period of time
(when it is expected not to have any activities in the system) to keep the data in
the new database, as well as to archive the previous preview, because previously
all overviews through cancel procedures are read from the Moodle database that
is linked to the new database, the warehouse. Besides that the administrator has
the right to register new users of the system from external bases for example
excel documents.
quizzes, but there are also other decisive factors that would help the teacher to
decide about the type of activities he would use in the future, to decide which
activities not to use in the future due to the bad results or their insufficient
attractiveness among the other activities and resources. The teacher can and
decide which of the students have difficulties in learning, which topics are more
difficult to overcome, so that he can react on time.
The last mining technigue that will be described in this paper is regresion. It is
the easiest technique to use, but is also probably the least powerful. Regresion
is a data mining function that predicts a number. A regression task begins with a
data set in which the target values are known. In the model build (training)
process, a regression algorithm estimates the value of the target as a function of
the predictors for each case in the build data. These relationships between
predictors and target are summarized in a model, which can then be applied to a
different data set in which the target values are unknown [19]. For example, in
our case study, using regresion we can predict the students grades, based on
observed data for many students activities over a period of time.
For our dashboard, different roles of users (manager, administrator, teacher,
user), display different reports gained with the mining techniques.
a. Manager level acquired data
Manager level enables manager role data survey and getting reports that
enable to follow the activity of all the participants in the system, number of
courses, set materials and resources, realized quizzes etc.
After logging as role manager, the home page (figure 4) gives the manager
quick view of the activity of the users and review of the number of the participants
in the courses.
Слика 4. Почетен поглед по најавување во улога на менаџер
Figure 4. First view after login in the role manager
The menager can look over all system registered users, users that have
confirmed their registration and those that haven’t confirmed the registration,
whereas in the submenu user activities he can see the last system access of the
users, shown for a certain date. An example for how much time the participants
(teachers/students) spent in the system is given at figure 5.
73
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
Слика 5. Поминато време во системот од извештајот за корисничка
активност
Figure 5. Time spent in the system from the user activity report
The course menu schemes a list of courses in categories and a total number
of, list of all courses and an opportunity to choose a certain course. After a course
is chosen a view of the number of discussions without content display is acquired,
but a view of the activities of the discussion participants, number of started topics,
number of theme answers. A list of all resources and access to all of them with
additional details for most visited and least visited resource. A survey of all the
activities such as homework and its assessment with a list of students with
highest and lowest score is also available. This part also offers a survey of all
questionnaires and their results as well as a possibility to print the questionnaire
results.
User portfolio enables individual course users’ data preview, as well as table
preview for all the course participants by viewing the activity (inactive, active, very
active), division in categories according to the assessment into bad, good and
excellent, prediction of whether they will complete the course successfully or not
and complete summary.
The user portfolio view gives contrastive analysis of a student’s results in
several courses and all the activities that characterize the student.
b. Administrator level acquired data
The administrator is the user of the application who has all the privileges. The
application administrator has the overall view of the user from the manager level.
There is an additional opportunity to give tasks in precisely defined period of time
(when it is expected not to have any activities in the system) to keep the data in
the new database, as well as to archive the previous preview, because previously
all overviews through cancel procedures are read from the Moodle database that
is linked to the new database, the warehouse. Besides that the administrator has
the right to register new users of the system from external bases for example
excel documents.
quizzes, but there are also other decisive factors that would help the teacher to
decide about the type of activities he would use in the future, to decide which
activities not to use in the future due to the bad results or their insufficient
attractiveness among the other activities and resources. The teacher can and
decide which of the students have difficulties in learning, which topics are more
difficult to overcome, so that he can react on time.
The last mining technigue that will be described in this paper is regresion. It is
the easiest technique to use, but is also probably the least powerful. Regresion
is a data mining function that predicts a number. A regression task begins with a
data set in which the target values are known. In the model build (training)
process, a regression algorithm estimates the value of the target as a function of
the predictors for each case in the build data. These relationships between
predictors and target are summarized in a model, which can then be applied to a
different data set in which the target values are unknown [19]. For example, in
our case study, using regresion we can predict the students grades, based on
observed data for many students activities over a period of time.
For our dashboard, different roles of users (manager, administrator, teacher,
user), display different reports gained with the mining techniques.
a. Manager level acquired data
Manager level enables manager role data survey and getting reports that
enable to follow the activity of all the participants in the system, number of
courses, set materials and resources, realized quizzes etc.
After logging as role manager, the home page (figure 4) gives the manager
quick view of the activity of the users and review of the number of the participants
in the courses.
Слика 4. Почетен поглед по најавување во улога на менаџер
Figure 4. First view after login in the role manager
The menager can look over all system registered users, users that have
confirmed their registration and those that haven’t confirmed the registration,
whereas in the submenu user activities he can see the last system access of the
users, shown for a certain date. An example for how much time the participants
(teachers/students) spent in the system is given at figure 5.
74
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
dashboard web based application, which is user-friendly and give better control
when it comes to larger groups of students when the standard reports reduce the
control clarity and the ability to evaluate their results at the end of the course [3],
[5], [16].
Nowadays, data mining tools are too complex to be used by the educators and
their futures go beyond the scope of what educator might to do. By creating a
dashboard that would communicate with the e-learning system Moodle, the
teachers can easily evaluate web activity in order to get more objective feedback,
and find out more about students capability in successfully passing the exam.
Also this dasboard will directly solve the teachers problems in supplying support
in dealing with various kind of algorithms. It could also be oriented towards the
academics and administrators responsible in order to obtain parameters about
how to improve site efficiency and adapt it to the behavior of their users, have
measures about how to better organize institutional resources (human and
material) and their educational offer, enhance educational program offers, etc.
Литература (References)
[1] Ramaswami M., and Bhaskaran R.: A Study on Feature Selection
Techniques in Educational Data Mining”, vol.1, Journal Of Computing,
ISSN:2151-9617, https://sites.google.com/site/journalofcomputing (December
2009)
[2] Elatia S., Ipperciel D., Hammad A.:Implications and Challenges to
Using Data Mining in Educational Research in the Canadian Context, Canadian
journal Of Education, pp. 101--119 (2012)
[3] Baradwaj B. K., Pal S.:Mining Educational Data to Analyze Students’
Performance, (IJACSA) International Journal of Advanced Computer Science
and Applications, vol. 2, no. 6 (2011)
[4] Yadav S. K., Bharadwaj B., Pal S.:Mining Education Data to Predict
Student’s Retention - A comparative Study, (IJCSIS) International Journal of
Computer Science and Information Security, vol. 10, no. 2 (2012)
[5] Cocea M., Weibelzahl S.: Disengagement Detection in Online
Learning - Validation Studies and Perspectives, IEEE transactions on learnin
technologies, vol. 4, no. 2 (April-June 2011)
[6] Romero C., Ventura S., García E.:Data mining in course management
systems - Moodle case study and tutorial
[7] BAKER R.S.J.D., YACEF K.:The State of Educational Data Mining in
2009 - Review and Future Visions
[8] Retalis S., Papasalouros A., Psaromiligkos Y., Siscos S., Kargidis T.:
Towards Networked Learning Analytics A concept and a tool
[9] Romero C., Ventura S., Espejo P. G. and Hervás C.: Data Mining
Algorithms to Classify Students, The 1st International Conference on
c. Teacher level acquired data
The teacher has a similar role in the application to the manager, with the little
difference that the manager has an overview of all courses and all users of the
application, whereas the user with the teacher’s role can view only the data that
refers to the courses that he has created and the system users that are
participants only in his courses.
Additionally the teacher has better view of the section that refers to the results
of the student’s homework, quizzes and their activity in the system.
The teacher can see which of the questions the participants have been
answered correctly and which not so that he can direct the participants to find out
the correct answers in the following lessons. For example, in the quiz section, to
be able to see the results according to the standard Moodle report several steps
are required in order to get a table view, so the application enables getting view
with only one click, automatic sorting of the results and a percentage
representation of the score as well as another column with a grade in form of
letter.
In the section that refers to the comparison of the results and the activity of the
students in the other quizzes, the teacher can see only the analyses of the
activities and the processed data only for the courses that he has created, but
not for all the other quizzes in the e-learning system.
d. User level acquired data
User with the user level role in the application is in fact the student who
participates in one or several courses of the e-learning system Moodle. The user
has a username and a password as in the profile he has created on the e-learning
system Moodle itself. The data that the user can see is from the user portfolio of
the manager and the teacher and refer only to the logged user. In this way the
user can view in which subject he participates, to see his activity, results,
comparison of the activities in different courses etc.
Besides these views the user gets certain suggestions by the teacher for the
necessity to pay more attention and to be more active in the working obligations
within the course in order to motivate him to accomplish better final results.
4. Conclusion
This work gives analyses of the data from the database in the e-learning system
Moodle and gives survey of the results from the data mining with the use of
several techniques applied on the application that offers several levels of
approach. It is necessary to integrate the data mining tools in the e-learning
environments which is the goal of this research, because in this way all these
data mining techniques will be applied in a single application and the feedback
and the acquired results will be directly applied on the e-learning environments
[16].
Here are several data mining techniques that can be used for acquiring
processed results and reports in the process of learning, and they are not
complicated to be used by the teachers. That is why this approach of creating
75
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
dashboard web based application, which is user-friendly and give better control
when it comes to larger groups of students when the standard reports reduce the
control clarity and the ability to evaluate their results at the end of the course [3],
[5], [16].
Nowadays, data mining tools are too complex to be used by the educators and
their futures go beyond the scope of what educator might to do. By creating a
dashboard that would communicate with the e-learning system Moodle, the
teachers can easily evaluate web activity in order to get more objective feedback,
and find out more about students capability in successfully passing the exam.
Also this dasboard will directly solve the teachers problems in supplying support
in dealing with various kind of algorithms. It could also be oriented towards the
academics and administrators responsible in order to obtain parameters about
how to improve site efficiency and adapt it to the behavior of their users, have
measures about how to better organize institutional resources (human and
material) and their educational offer, enhance educational program offers, etc.
Литература (References)
[1] Ramaswami M., and Bhaskaran R.: A Study on Feature Selection
Techniques in Educational Data Mining”, vol.1, Journal Of Computing,
ISSN:2151-9617, https://sites.google.com/site/journalofcomputing (December
2009)
[2] Elatia S., Ipperciel D., Hammad A.:Implications and Challenges to
Using Data Mining in Educational Research in the Canadian Context, Canadian
journal Of Education, pp. 101--119 (2012)
[3] Baradwaj B. K., Pal S.:Mining Educational Data to Analyze Students’
Performance, (IJACSA) International Journal of Advanced Computer Science
and Applications, vol. 2, no. 6 (2011)
[4] Yadav S. K., Bharadwaj B., Pal S.:Mining Education Data to Predict
Student’s Retention - A comparative Study, (IJCSIS) International Journal of
Computer Science and Information Security, vol. 10, no. 2 (2012)
[5] Cocea M., Weibelzahl S.: Disengagement Detection in Online
Learning - Validation Studies and Perspectives, IEEE transactions on learnin
technologies, vol. 4, no. 2 (April-June 2011)
[6] Romero C., Ventura S., García E.:Data mining in course management
systems - Moodle case study and tutorial
[7] BAKER R.S.J.D., YACEF K.:The State of Educational Data Mining in
2009 - Review and Future Visions
[8] Retalis S., Papasalouros A., Psaromiligkos Y., Siscos S., Kargidis T.:
Towards Networked Learning Analytics A concept and a tool
[9] Romero C., Ventura S., Espejo P. G. and Hervás C.: Data Mining
Algorithms to Classify Students, The 1st International Conference on
c. Teacher level acquired data
The teacher has a similar role in the application to the manager, with the little
difference that the manager has an overview of all courses and all users of the
application, whereas the user with the teacher’s role can view only the data that
refers to the courses that he has created and the system users that are
participants only in his courses.
Additionally the teacher has better view of the section that refers to the results
of the student’s homework, quizzes and their activity in the system.
The teacher can see which of the questions the participants have been
answered correctly and which not so that he can direct the participants to find out
the correct answers in the following lessons. For example, in the quiz section, to
be able to see the results according to the standard Moodle report several steps
are required in order to get a table view, so the application enables getting view
with only one click, automatic sorting of the results and a percentage
representation of the score as well as another column with a grade in form of
letter.
In the section that refers to the comparison of the results and the activity of the
students in the other quizzes, the teacher can see only the analyses of the
activities and the processed data only for the courses that he has created, but
not for all the other quizzes in the e-learning system.
d. User level acquired data
User with the user level role in the application is in fact the student who
participates in one or several courses of the e-learning system Moodle. The user
has a username and a password as in the profile he has created on the e-learning
system Moodle itself. The data that the user can see is from the user portfolio of
the manager and the teacher and refer only to the logged user. In this way the
user can view in which subject he participates, to see his activity, results,
comparison of the activities in different courses etc.
Besides these views the user gets certain suggestions by the teacher for the
necessity to pay more attention and to be more active in the working obligations
within the course in order to motivate him to accomplish better final results.
4. Conclusion
This work gives analyses of the data from the database in the e-learning system
Moodle and gives survey of the results from the data mining with the use of
several techniques applied on the application that offers several levels of
approach. It is necessary to integrate the data mining tools in the e-learning
environments which is the goal of this research, because in this way all these
data mining techniques will be applied in a single application and the feedback
and the acquired results will be directly applied on the e-learning environments
[16].
Here are several data mining techniques that can be used for acquiring
processed results and reports in the process of learning, and they are not
complicated to be used by the teachers. That is why this approach of creating
76
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, Универзитет „Гоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
(лектор)
ПРЕГЛЕД НА ТЕХНИКИ ЗА ПРЕПОЗНАВАЊЕ НА ЛИК ОД ВИДЕО
Ана Љуботенска1, Игор Стојановиќ2
1 Факултет за информатика, УниверзитетГоце Делчев“ - Штип
(ana.ljubotenska, igor.stojanovik)@ugd.edu.mk
Апстракт
Во областа на анализирање на слики, значаен проблем претставува
препознавањето на лик чија основна цел е да се открие или потврди идентитетот на
личност од внесена слика, ако се дадени слика од лик како влезен податок и база со
слики од познати ликови. Оваа проблематика стана особено актуелна во последните
години, пред сè поради големата примена што ја има во различни домени, како на
пример во биометриската верификација. Техниките кои се користат за оваа цел се
класифицирани во три групи, во зависност од методологијата за добивање на
податоците за ликот: методи кои обработуваат видео и аудио секвенци
(аудиовизуелно препознавање на лик), интензитет на слика или други клучни
податоци, како што се инфрацрвена слика, 3Д или 2Д податоци. Методологиите
можат да се комбинираат, така што ќе се работи за бимодално препознавање на лик
или мултимодално препознавање. Фокусот во овој труд е кон тоа како да се
комбинираат различни биометриски карактеристики за биометриската верификација
за да се направи посигурно препознавањето на лик. Главната идеја е да се направи
преглед и споредба на перформансите на клучните техники кои се користат за
препознавање на лик од видео, укажувајќи на нивните основни карактеристики и
предности, со што овој труд ќе им послужи на авторите како основа за понатамошни
истражувања и продлабочувања во оваа област.
Клучни зборови: биометриска идентификација, препознавање на лик,
препознавање на говор, видео.
PREVIEW OF METHODS FOR IMAGE RESTORATION FROM VIDEO
Ana Ljubotenska1, Igor Stojanovik2
1Faculty of computer science, Goce Delcev University, Stip, Macedonia
(ana.ljubotenska, igor.stojanovik)@ugd.edu.mk
Abstract
In the field of image processing, significant problem is face recognition. Main goal is to
determine or validate person identity from the entered image, if we have image with person
as input and database with recognized faces. This issue has become particularly topic
current years, primarily due to large scale applications that has in various domains, such as
in biometric verification. The techniques which are used for this purpose are classified in
three groups, depending on the methodology for obtaining data for person: methods that
process video and audio sequences (audio - visual character recognition), intensity image,
or other means data such as infrared image, 3D or 2D data. Methodologies can be
combined, so face recognition can be bimodal or multimodal. The focus in this paper is how
to combine different biometric features for biometric verification to make face recognition
safer. The main idea is to review and compare the performance of the means techniques
used for face recognition from video, showing their basic features and advantages. This
paper will be base of our future research in this topic.
Kew words: biometric identifications, face recognition, voice recognition, video.
1. Вовед
Биометриски базираните технологии се покажаа како најсоодветно решение за
препознавање на личности, со што се овозможува автентикација на личности и дозволен пристап
до виртуелни и физички уреди. Ова е овозможено со користење на паметни картички, токени,
лозинки, пинови и слично. Лозинките и пиновите, иако се често користени, лесно можат да се
заборават, а исто така можат да бидат откриени од друго лице кое нема овластен пристап.
Токените и картичките, пак, можат лесно да се изгубат или да се направат нивни дупликати. Овие
недостатоци ја навестуваат потребата од пронаоѓање на друг начин за идентификација на
личности. Најсоодветен начин е оној кој се базира на индивидуалните биолошки карактеристики,
бидејќи тие не можат да бидат заборавени, изгубени или украдени. Биометриски базираните
технологии вклучуваат идентификација што се базира на физичките и логички карактеристики на
Educational Data Mining, Montréal, Québec, Canada, pp. 8-18 (June 20-21,
2008)
[10] Yadav S. K., Pal S.: Data Mining - A Prediction for Performance
Improvement of Engineering Students using Classification, World of Computer
Science and Information Technology Journal (WCSIT) ISSN: 2221-0741 vol. 2,
no. 2, pp. 51-56 (2012)
[11] Dash M., Liu H.:Feature Selection for Classification, An International
Journal of Intelligent Data Analysis, vol. 1, no. 3,2006, pp. 131-156 (1997)
[12] Chen G., Liu C., Ou K., Liu B.: Discovering decision knowledge from
web log portfolio for managing classroom processes by applying decision tree
and data cube technology, Journal of Educational Computing Research, pp.
305332 (2000)
[13] Anozie N., Junker B.W.: Predicting end-of-year accountability
assessment scores from monthly student records in an online tutoring system,
Educational Data Mining AAAI Workshop, pp. 1-6, California, USA (2006)
[14] Moodle org. LMS Moodle official site. web. 11 Apr. 2013,
http://moodle.org
[15] High school Dobri Daskalov, E-learning Moodle, Kavadarci, R. of
Macedonia: n.p., 2009. web. 11 Apr. 2013, http://moodle.dobridaskalov.edu.mk
[16] Darrell M. W.:Big Data for Education - Data Mining, Data Analytics,
and Web Dashboards, U.S. Department of Education Office of Educational
Technology, Enhancing Teaching and Learning Through Educational Data
Mining and Learning Analytics”, pp. 36 (2012)
[17] Oracle. Oracle Data Mining Concepts. Release 1 (11.1), Oracle Data
Mining Concepts , web.7 Sept. 2013
104
Годишен зборник 2013
Yearbook 2013
Факултет за информатика, УниверзитетГоце Делчев“ – Штип
Faculty of Computer Science, Goce Delcev University – Stip
... A selected review of recent works in [32,33,34], on learning analytics reveals the following issues. The list is non-exhaustive. ...
Conference Paper
Personalized learning models are developed to cater for the differentiation in learner styles and needs. Tutors determine the most appropriate learning components for each student. The learning units (LUs) are adapted to learners based on their contexts. However, there are no methods that adapt learning objects to learners based on their personalized learning styles. There also does not exist appropriate techniques that employ decision making approaches to evaluate the LUs. This study presents a model that uses learning styles to determine the appropriate learning information by employing learning analytics. Its proposed evaluation model facilitates evaluation of how suitable, acceptable and useful- ness of personalized learning in the LUs. To test the model, varying evaluation criteria weights are employed. It is proposed that the model can be used by tutors to assist learners in creating and applying LUs that are most suitable for their needs thereby improving the quality of learning.
... A selected review of recent works in [32,33,34], on learning analytics reveals the following issues. The list is non-exhaustive. ...
Conference Paper
Personalized learning models are developed to cater for the differentiation in learner styles and needs. Tutors determine the most appropriate learning components for each student. The learning units (LUs) are adapted to learners based on their contexts. However, there are no methods that adapt learning objects to learners based on their personalized learning styles. There also does not exist appropriate techniques that employ decision making approaches to evaluate the LUs. This study presents a model that uses learning styles to determine the appropriate learning information by employing learning analytics. Its proposed evaluation model facilitates evaluation of how suitable, acceptable and useful- ness of personalized learning in the LUs. To test the model, varying evaluation criteria weights are employed. It is proposed that the model can be used by tutors to assist learners in creating and applying LUs that are most suitable for their needs thereby improving the quality of learning.
... After thorough analysis of related research on applying learning analytics in education (Baker and Yacef 2009;Baradwaj and Pal 2011;Campagni et al. 2015;Milevski and Zdravev 2013;Romero, Ventura, and García 2008;Romero et al. 2013;Srivastava and Srivastava 2013;Verma, Thakur, and Jalori 2012), the author came to conclusion that basic learning analytics/educational data mining techniques applicable in this case should be (but not limited for): (1) Classification to classify each item in a set of data into one of predefined set of learners group; (2) Clustering to determine groups of students that need special course profiling; (3) Association rules to discover interesting relations between course elements which were used by particular students; (4) Prediction to predict dependencies of using learning environment's activities/tools and final student's learning outcomes; (5) Decision tree of students' actions. Decision trees are one of the most common methods in data mining, because the application or such kind of model is easy to understand. ...
Article
Full-text available
The aim of the paper is to present methodology to personalise learning using learning analytics and to make further decisions on suitability, acceptance and use of personalised learning units. In the paper, first of all, related research review is presented. Further, an original methodology to personalise learning applying learning analytics in virtual learning environments and empirical research results are presented. Using this learning personalisation methodology, decision-making model and method are proposed to evaluate suitability, acceptance and use of personalised learning units. Personalised learning units evaluation methodology presented in the paper is based on (1) well-known principles of Multiple Criteria Decision Analysis for identifying evaluation criteria; (2) Educational Technology Acceptance & Satisfaction Model (ETAS-M) based on well-known Unified Theory on Acceptance and Use of Technology (UTAUT) model, and (3) probabilistic suitability indexes to identify learning components’ suitability to particular students’ needs according to their learning styles. In the paper, there are also examples of implementing the methodology using different weights of evaluation criteria. This methodology is applicable in real life situations where teachers have to help students to create and apply learning units that are most suitable for their needs and thus to improve education quality and efficiency.
Article
Full-text available
We give a brief overview of feature selection methods used in statistical classification. We cover filter, wrapper and embedded methods.
Article
Full-text available
We present methodology for developing functions that predict student scores on end of year state accountabil- ity exams from dynamic testing metrics developed from intelligent tutoring system log data. Our results con- firm the findings of Heffernan et al. (2006) that on- line tutoring log based metrics provide better predic- tions than using paper and pencil benchmark tests. Our approach provides a family of prediction functions to be used throughout the year, in order to provide timely and valid feedback to teachers and schools about stu- dent progress. Since the same dynamic testing metrics are used in all prediction functions in the family, we can also begin to understand the changing influence of these metrics on prediction over time.
Article
Full-text available
In conventional classrooms, teachers attempt to enhance instruction by monitoring students' learning processes and analyzing their performances by paper records and observation. Similarly, distance learning systems on the Web should be designed to record students' behaviors to assist teachers in assessing performance and making decisions related to curriculum. Recent developments in Web server systems can record the students' access to the learning systems in Web logs. Information processing analysis on the historical classroom processes can help teachers to develop knowledge for applying proper teaching strategies according to available information in Web logs. However, teachers cannot easily infer the pedagogical meaning of Web logs and discover the pedagogical rules of students' behavior patterns in the Web logs to refine teaching strategies. Therefore, to use Web logs for pedagogical purposes, this article adopts decision tree and data cube information processing methodologies to observe students' behaviors and discover the pedagogical rules on students' learning performance from Web logs. The architecture and guidelines of utilizing the data cube and decision tree methodologies for pedagogical purposes are also presented. Consequently, teachers can efficiently estimate and explain the effectiveness of pedagogical strategies, ultimately improving instruction with decision tree and data cube software.
Article
Full-text available
Learning environments aim to deliver efficacious instruction, but rarely take into consideration the motivational factors involved in the learning process. However, motivational aspects like engagement play an important role in effective learning-engaged learners gain more. E-Learning systems could be improved by tracking students' disengagement that, in turn, would allow personalized interventions at appropriate times in order to reengage students. This idea has been exploited several times for Intelligent Tutoring Systems, but not yet in other types of learning environments that are less structured. To address this gap, our research looks at online learning-content-delivery systems using educational data mining techniques. Previously, several attributes relevant for disengagement prediction were identified by means of log-file analysis on HTML-Tutor, a web-based learning environment. In this paper, we investigate the extendibility of our approach to other systems by studying the relevance of these attributes for predicting disengagement in a different e-learning system. To this end, two validation studies were conducted indicating that the previously identified attributes are pertinent for disengagement prediction, and two new meta-attributes derived from log-data observations improve prediction and may potentially be used for automatic log-file annotation.
Article
Full-text available
Now-a-days the amount of data stored in educational database increasing rapidly. These databases contain hidden information for improvement of students' performance. Educational data mining is used to study the data available in the educational field and bring out the hidden knowledge from it. Classification methods like decision trees, Bayesian network etc can be applied on the educational data for predicting the student's performance in examination. This prediction will help to identify the weak students and help them to score better marks. The C4.5, ID3 and CART decision tree algorithms are applied on engineering student's data to predict their performance in the final exam. The outcome of the decision tree predicted the number of students who are likely to pass, fail or promoted to next year. The results provide steps to improve the performance of the students who were predicted to fail or promoted. After the declaration of the results in the final examination the marks obtained by the students are fed into the system and the results were analyzed for the next session. The comparative analysis of the results states that the prediction has helped the weaker students to improve and brought out betterment in the result.
Article
Full-text available
The main objective of higher education is to provide quality education to students. One way to achieve highest level of quality in higher education system is by discovering knowledge for prediction regarding enrolment of students in a course. This paper presents a data mining project to generate predictive models for student retention management. Given new records of incoming students, these predictive models can produce short accurate prediction lists identifying students who tend to need the support from the student retention program most. This paper examines the quality of the predictive models generated by the machine learning algorithms. The results show that some of the machines learning algorithms are able to establish effective predictive models from the existing student retention data.
Article
Canadian institutions of higher education are major players on the international arena for educating future generations and producing leaders around the world in various fields. In the last decade, Canadian universities have seen an influx in their incoming international students, who contribute over 3.5 billion to the Canadian economy (Madgett & Bélanger 2008, p. 195). Research in Canadian post-secondary institutions is booming, especially in education (SSHRC, 2011)-for the academic year 2010-2011, of the 12 subject areas, the total SSHRC funding for projects in education, ranked fourth, exceeding 27 million. All of these variables place Canadian higher education in a leading and strategic position in several educational research fields. One can imagine the wealth of knowledge about trends in higher education that could be revealed if the large amount of data generated by Canadian universities were systematically analyzed and handled using techniques such as data mining. However, not much can be achieved from the unharnessed knowledge accumulated on a daily basis, as the advancement of data mining research that would provide the ultimate tool to learn about trends and changes in Canadian institutions is often held back by inadequate data warehousing, as well as by privacy, confidentiality, and copyright regulations. In this paper, we engage in a critical discussion/analysis of the interface between data mining research in higher education and the legal implications of such a tool.
Article
Feature selection has been the focus of interest for quite some time and much work has been done. With the creation of huge databases and the consequent requirements for good machine learning techniques, new problems arise and novel approaches to feature selection are in demand. This survey is a comprehensive overview of many existing methods from the 1970's to the present. It identifies four steps of a typical feature selection method, and categorizes the different existing methods in terms of generation procedures and evaluation functions, and reveals hitherto unattempted combinations of generation procedures and evaluation functions. Representative methods are chosen from each category for detailed explanation and discussion via example. Benchmark datasets with different characteristics are used for comparative study. The strengths and weaknesses of different methods are explained. Guidelines for applying feature selection methods are given based on data types and domain characteris...
Article
Educational data mining is an emerging discipline, concerned with developing methods for exploring the unique types of data that come from the educational context. This work is a survey of the specific application of data mining in learning management systems and a case study tutorial with the Moodle system. Our objective is to introduce it both theoretically and practically to all users interested in this new research area, and in particular to online instructors and e-learning administrators. We describe the full process for mining e-learning data step by step as well as how to apply the main data mining techniques used, such as statistics, visualization, classification, clustering and association rule mining of Moodle data. We have used free data mining tools so that any user can immediately begin to apply data mining without having to purchase a commercial tool or program a specific personalized tool.