DISTANCE LEARNING BY INTELLIGENT TUTORING SYSTEM
Part II: Student/teacher adaptivity in an Engineering Course
José Manuel Gascueña, Antonio Fernández-Caballero
Computer Science Reseach Institute of Albacete, University of Castilla-La Mancha, Albacete, Spain
Enrique Lazcorreta , Federico Botella
Operations Research Centre, University Miguel Hernandez of Elche, Elche, Spain
Keywords: Distance learning, Intelligent Tutoring System
s, E-learning, E-teaching, Education, Adaptivity.
Abstract: Intelligent Tutoring Systems (ITS) have proven their worth in multiple ways and in multiple domains in
Education. In this article the application of an Intelligent Tutoring System to an Engineering Course is
introduced. The paper also introduces an explanation of how the course adapts to the students as well as to
the teachers. User adaptation is provided by means of the so called pedagogical strategies, which among
others specify how to proceed in showing the contents of the matter for a better assimilation of the
knowledge by the student. Thus, in this paper the adaptation mechanisms implemented in the ITS, which
permit that the students learn better and the professors teach better, are explained in extensive.
1 INTRODUCTION
One of the main problems in Intelligent Tutoring
Systems (ITS) consists in adapting to the needs of
the user who interacts at each moment. A way to
provide user adaptation is by means of the so called
pedagogical strategies, which specify how to
sequence the contents, what kind of feedback has to
be given during education, when and how the tutor’s
contents (problems, definitions, examples, and so
on) have to be shown or explained (Murray, 1999).
ITS have proven their worth in multiple ways
an
d in multiple domains in Education (Anderson,
Corbett, Koedinger & Pelletier, 1995; Woolf, Beck,
Eliot & Stern, 2001). ITS are growing in acceptance
and popularity for several reasons, including: (i) an
increased student performance, (ii) a deepened
cognitive development, and, (iii) a reduced time for
the student to acquire skills and knowledge (Sykes
& Franek, 2003). Currently, ITS can be found in
core databases, mathematics, physics, language,
medicine, and other courses in many schools.
In the field of databases, KERMIT (Suraweera &
Mitrovic,
2002) teaches the conceptual modelling of
databases using the entity - relation data model; and
SQL-Tutor (Mitrovic, Martin & Mayo, 2002)
teaches databases SQL language. In both
approaches, a model based on restrictions has been
used. In the field of physics, Andes (Gertner &
VanLehn, 2000) allows the students to solve
problems of classic physics in an environment that
offers visualization, immediate feedback, and
procedural and conceptual help; and Why2-Atlas
system (VanLehn et al., 2002) teaches qualitative
physics by having students write paragraph-long
explanations of simple mechanical phenomena.
In the field of programming languages,
ELMART
(Weber & Brusilovsky, 2001) teaches
programming in LIPS, while JITS (Sykes & Franek,
2003) teaches JAVA, and Bits (Butz, Hua &
Maguire, 2004) teaches C++. In the field of
mathematics, PAT (Koedinger & Anderson, 1997),
Ms. Lindquist (Heffernan & Koedinger, 2002), and
Aplusix-Editor (Nicaud, Bouhineau & Huguet,
2002) help the students to learn algebra.
ITS also exist in many other newer but relevant
fields. For
example, SlideTutor (Crowley &
Medvedeva, 2003) is an ITS in dermatopathology,
and Design Pattern (Jeremic, Devedzic & Gasevic,
2004) is used to learn design patterns.
On the other hand, there has been a great
researc
h effort in learning strategies to be
incorporated into ITS (Boulay & Luckin, 2001). As
an example, Meyer has used the analogy (Meyer,
2002) to teach a less known domain from a more
familiar one. The case based reasoning paradigm has
148
Manuel Gascueña J., Fernández-Caballero A., Lazcorreta E. and Botella F. (2005).
DISTANCE LEARNING BY INTELLIGENT TUTORING SYSTEM - Part II: Student/teacher adaptivity in an Engineering Course.
In Proceedings of the Seventh International Conference on Enterprise Information Systems, pages 148-153
DOI: 10.5220/0002519601480153
Copyright
c
SciTePress
also been an inspiration to help in obtaining new
incrementing knowledge (Martens, 2004). Even,
reinforcement learning has been used (Bennane,
2002). When various strategies are together
implemented in an ITS, as for instance in (Prentzas,
Hatzilygeroudis & Garofalakis, 2002), the system
selects the most appropriate one for the activity that
the student is performing.
The structure of the paper is as follows. Section 2
shows the adaptivity mechanisms provided for the
student based on the overall e-learning capabilities
offered. Section 3 is devoted to the adaptivity to the
teacher. Finally some conclusions are provided at
the end of the article.
2 STUDENT ADAPTIVITY
2.1 E-learning Capabilities
First of all let us focus on the functionality that the
ITS offers to the student (see figure 1).
Of course, the student must register in the course
(use case “Register for the course”) as stated in
sequence diagram of figure 2. The registered
students can change their passwords (through use
case “Change password”) each time they enter the
course to begin a new study session.
Once a new study session has been started
through use case “Enter the course”, the student
reads pages of theory (”Read theory”), answers
exercises (“Solve exercise”) or test questionnaires
(“Solve test”) depending on the task that a
pedagogic module proposes through time. While
completing an exercise, the student can consult the
theory (use case “Consult theory”) closely related to
the exercise.
During a study session the student can also
change the style of presentation of the matter (that is
to say the visual preferences – “Change
preferences”). The student may also consult at any
time his state after performing any task.
2.2 Student Learning Mechanism
Figure 3 shows the steps followed by the pupil when
studying each topic of the course (“Matter learning”).
(1) Firstly, the student has to read the whole
theory for the current topic.
(2) Afterwards, the student has to solve the
exercises proposed. If the student is a level-1 (low
level) student, firstly he has to solve the basic
exercises and then the complex ones. O the other
side, if the student is a level-2 (high level) student,
he will only have to solve the complex exercises.
The basic exercises are all shown in a sequential
way, and then the ITS evaluates if the student has
reached a minimum score associated to the topic. On
the contrary, the complex exercises are shown in
blocks (composed of a pre-determined quantity of
exercises), and, after showing each block, there is an
evaluation to ask for a minimum mark before
composing the next block. After correctly fulfilling a
number of complex exercises, the system goes on to
the test questionnaires.
(3) Lastly, the student has to solve the test
questionnaire offered.
(4) If there are more topics in the course, the
system goes back to step (1). Otherwise, the student
has finished studying the matter.
Figure 1: The student’s requirements
Figure 2: Sequence diagram for use case
“Register for the course”
DISTANCE LEARNING BY INTELLIGENT TUTORING SYSTEM. Part II: Student/teacher adaptivity in an Engineering
Course
149
Figure 3: Activity diagram for “Matter learning”
During steps (2) and (3), if the student does not
obtain the minimum scores fixed for the topic, he
gets reinforcement in order to reach the objectives
for the course.
2.2.1 Reinforcement for Basic Exercises
Figure 4 shows how the alumni are reinforced
during their activity of solving basic exercises. The
system selects one of the exercises previously
proposed and not well solved from the set of basic
exercises and gets the reinforcement material (based
on previous topics studied). This way the system
helps the student to correctly solve the exercise.
After proposing the reinforcement material, and
before the student has to solve again the basic
exercise, the ITS shows the bad response that the
student gave previously. When the student passes
the minimum score, the system does not go on
providing reinforcement. But, and this is the worst
situation, if the system has provided reinforcement
to all badly answered exercises, and even so the
student has not been able to solve them, the ITS tells
the student to consult the tutor personally. After
having his meeting with the teacher, the student is
permitted to advance in the study of the course.
2.2.2 Reinforcement for Complex Exercises
The strategy for providing reinforcement to the
student in complex exercises is very similar to the
strategy followed to give reinforcement in basic
exercises. The only difference is that the ITS firstly
tries to reinforce with material of the current topic;
and, if the student is still not able to solve the
complex exercise, he is reinforced by material from
previous topics of the course.
If the student does not have seen all selected
complex exercises, he will only get reinforcement
for those exercises offered to the user in the last
block of complex exercises. But, if he already has
been offered all the complex exercises blocks, he
will be reinforced for all complex exercises
incorrectly solved and not yet reinforced previously.
ICEIS 2005 - HUMAN-COMPUTER INTERACTION
150
Figure 4: Activity diagram for “Reinforcement for basic
exercises”
2.2.3 Reinforcement for Tests
If the student does not get a minimum mark in the test
questionnaire proposed for the current topic, the ITS
builds a new test questionnaire, offers it to the student, and,
if the student does not perform well, the professor
personally must reinforce in order to proceed with the
learning activity – activity “Go to the teacher” - (see figure
5).
Figure 5: Activity diagram for “Reinforcement for test”
3 TEACHER ADAPTIVITY
3.1 E-teaching Capabilities
Let us also talk about what the teacher can do with
the ITS (see figure 6).
Let us firstly focus on the general requirements
of the professor. Evidently, the teacher must be
authenticated successfully to accede to the ITS
functionality (use case “Authenticate”). Once the
teacher has been authenticated, he can consult all the
didactic material (theory –“Consult theory”-,
exercises –“Consult exercises”-, and test
questionnaires –“Consult test battery questions”) of
each of the topics of the matter and obtain quiet
statistics fruit of the interaction of the students with
the system. It can give reinforcement to the students
who need his help, by means of use case “Reinforce
the student”. This is because the student has not
managed to advance in the study of the subject
because the material that has provided to him the
pedagogic module is not sufficient to overcome the
goals of the topic that he is studying. The teacher
may change the style of presentation (colours,
margins, interlineate, size and type of source) of the
interface.
But, the biggest benefit is that the teacher may
consult statistics fruit of the interaction of the
students with the system. For every topic of the
matter, the teacher also obtains the number of times
there has been a need to reinforce the students to be
able to advance in the study.
The information provided to the teacher is
gathered during interaction of the students with the
ITS. Respect to the theory read by the students, the
teacher can know the number of times that the
students have acceded to every page of theory and
the average time that they have spent in every visit.
The system also records when students have done
scroll when they visit a page of theory and it may
reproduce all movements performed. In the same
sense, and in accordance with the exercises proposed
to the students, the teacher is able to know the
average time that students have spent in performing
them, the percentage of pupils that have not been
able to perform them correctly, and how many
students have answered well or badly. The teacher
can also know the number of times that every
exercise has had to be explained again by means of
theory pages. Lastly, for the case of test
questionnaires, the teacher can know the number of
times that every test question has been shown to the
students, the percentage of tests answered well or
badly, and so on.
DISTANCE LEARNING BY INTELLIGENT TUTORING SYSTEM. Part II: Student/teacher adaptivity in an Engineering
Course
151
Figure 6: The professor’s capabilities
3.2 Teacher Learning Mechanism
The teacher may learn how to enhance teaching of
the course from the statistics gotten from the
interaction of the students with the ITS.
The teacher will be able to know the efficiency
of the implemented mechanism to reinforce the
students. For each topic of the course, the teacher
may consult the information on the number of times
that the students have needed reinforcement and how
many times the students have personally been
reinforced by the professor. Moreover, the teacher
may consult statistics of information gathered during
theory, exercises and test phases, respectively.
The teacher may know the number of times that
the students have accessed each theory page and the
mean time the students have been on each visited
page. The statistics are classified as (a) pages read
during the theory phase, (b) pages consulted during
the exercises phase, either solicited by the student or
due to the reinforcement mechanism. The ITS also
offers the possibility to know which students have
performed scrolling when visiting theory pages and
to reproduce the scroll movements as performed.
In relation to the exercises statistics, the
professor may consult to how many students an
exercise has been shown, the mean time the students
have spent to solve the exercise and the percentage
of blank, correct and incorrect answers to the
exercise. There is a classification in exercises
presented as reinforcement and normal exercises.
The information on how many times an exercise has
been explained personally by the teacher is also
provided.
In relation to test questionnaires, the teacher may
look for the number of times that each test question
has been presented to the students and the number of
times that the students have left the test blank, have
answered correctly and have answered incorrectly. It
is possible to know if the test question was presented
as reinforcement to an exercise, or if it was part of a
test questionnaire. Furthermore, the teacher may
know the number of times that he had to personally
explain the test question.
4 CONCLUSIONS
The ITS have turned into a technology of increasing
interest to complement traditional education so
much from the perspective of the students as from
that of the teachers. In this article the application of
an ITS architecture to an engineering course has
been introduced. The aim of the ITS is that the
students can learn more and better, and on the other
hand that the teachers can extract conclusions that
help them to improve their teaching activities.
In this paper, we have introduced an explanation
of how the course adapts to the students as well as to
the teachers. User adaptation is provided by means
of the so called pedagogical strategies, which among
others specify how to proceed in showing the
contents of the matter for a better assimilation of the
knowledge by the student.
ICEIS 2005 - HUMAN-COMPUTER INTERACTION
152
ACKNOWLEDGEMENTS
This work is supported in part by the Spanish Junta
de Comunidades de Castilla-La Mancha PBC-03-
003 and the Spanish CICYT TIN2004-08000-C03-
01 grants.
REFERENCES
Anderson, J.R., Corbett, A.T , Koedinger, K.R., Pelletier,
R., 1995. Cognitive tutors: Lessons learned. The
Journal of the Learning Sciences, 4: 167-207.
Bennane, A., 2002. An approach of reinforcement learning
use in tutoring systems. International Conference on
Intelligent Tutoring System, ITS 2002. Lecture Notes
in Computer Science, 2363: 993.
Boulay, B., Luckin, R., 2001. Modelling human teaching
tactics and strategies for tutoring systems.
International Journal of Artificial Intelligence in
Education, 12:235-256.
Butz, C.J., Hua, S., Maguire, R.B., 2004. Bits: A Bayesian
Intelligent Tutoring System for computer
programming. Western Canadian Conference on
Computing Education.
Crowley, R.S., and Medvedeva, O., 2003. A general
architecture for intelligent tutoring of diagnostic
classification problem solving. Proceedings of the
Annual AMIA Syposium, pages 185-189.
Gertner, A., VanLehn, K. 2000. Andes: A coached
problem solving environment for physics. Intelligent
Turing Systems: 5th International Conference, ITS
2000, Lecture Notes in Computer Science, 1839:133-
142.
Heffernan, N.T., Koedinger, K.R., 2002. An Intelligent
Tutoring System incorporating a model of an
experienced human tutor. International Conference on
Intelligent Tutoring System, ITS 2002. Lecture Notes
in Computer Science, 2363:596-608.
Jeremic, Z., Devedzic, V., Gasevic, D., 2004. An
Intelligent Tutoring System for learning design
patterns. International Workshop on Adaptative
Hypermedia and Collaborative Web-based Systems,
AHCW’ 04.
Koedinger, K.R., Anderson, J.R., 1997. Intelligent tutoring
goes to school in the big city. International Journal of
Artificial Intelligence in Education, 8:30-43.
Martens, A., 2004. Case-based training with Intelligent
Tutoring Systems. Proceedings of the The 4th IEEE
International Conference on Advanced Learning
Technologies.
Meyer, C., 2002. Hypermedia environment for learning
concepts based on inter-domain analogies as an
educational strategy. International Conference on
Intelligent Tutoring System, ITS 2002. Lecture Notes
in Computer Science, 2363:281-290.
Millán, E., Agosta, J.M, Pérez J.L., 1999. Application of
bayesian networks to student modelling. Proceedings
of PEG'99: Intelligent Computer and Communications
Technology: Teaching & Learning for the 21st
Century.
Mitrovic, A., Martin, B., Mayo, M., 2002. Using
evaluation to shape ITS design: Results and
experiences with SQL-Tutor. International Journal of
User Modeling and User-Adapted Interaction, 12(2-
3):243–279.
Murray, T., 1999. Authoring Intelligent Tutoring Systems:
An analysis of the state of the art. International Journal
of Artificial Intelligence in Education, 10:98-129.
Nicaud, J.F, Bouhineau, D., Huguet, T., 2002. The
Aplusix-Editor: A new kind of software for the
learning of algebra. Intelligent Tutoring Systems, ITS
2002. Lectures Notes in Computer Science, 2363:178-
187.
Prentzas, J., Hatzilygeroudis, I., Garofalakis J., 2002.
Web-based Intelligent Tutoring Systems using hybrid
rules as its representational basis. Intelligent Tutoring
Systems, ITS 2002. Lectures Notes in Computer
Science, 2363:119-128.
Suraweera, P., Mitrovic, A., 2002. KERMIT: A constraint-
based tutor for database modeling. Intelligent Tutoring
Systems, ITS 2002. Lectures Notes in Computer
Science, 2363:337-387.
Sykes, E.R., and Franek, F., 2003. A prototype for an
Intelligent Tutoring System for students learning to
program in Java. Proceedings of IASTED
International Conference on Computers and Advanced
Technology in Education, CATE2003, pages. 78-83 .
VanLehn, K., Jordan P.W., Rosé, C.P., Bhembe, D.,
Böttner, M., Gaydos, A., Makatchev, M.,
Pappuswamy, U., Ringenberg, M, Roque, A., Siler, S.,
Srivastava, R., 2002. The architecture of Why2-Atlas:
A coach for qualitative physics essay writing.
Intelligent Tutoring Systems, ITS 2002. Lectures
Notes in Computer Science, 2363:158-167.
Weber, G., Brusilovsky, P., 2001. ELM-ART: An adaptive
versatile system for Web-based instruction.
International Journal of Artificial Intelligence in
Education, 12:351-384
Woolf, B.P., Beck, J., Eliot, C., Stern, M., 2001. Growth
and maturity of intelligent tutoring systems: A status
report. In Smart Machines in Education, AAAI/MIT
Press, pages 99-144.
DISTANCE LEARNING BY INTELLIGENT TUTORING SYSTEM. Part II: Student/teacher adaptivity in an Engineering
Course
153