|
|
Learning Technology publication
of IEEE Computer Society's |
|
|
Volume 7 Issue 2 |
ISSN 1438-0625 |
April 2005 |
CASLO: Collaborative Annotation Service for Learning
Objects
Computational Sciences in Higher
Education: "Have You Created Your Model Yet"
Modeling Learning in simSchool
Testing an interactive lecturing technique for teaching
ubiNEXT: Exhibition Guide System
that Supports Visitor’s Learning Experience in Museums
Intelligent Agents: a New Paradigm
to Support Collaborative Learning in Distance Education Systems
The Civic Nexus Project: Supporting
Sustainable Technology Practices in Geographic Communities
TutorBot: an Intelligent System to support Tutoring on line in an
e-learning platform
Awareness : Framework for e-learning
Using a Knowledge Management System
for E-Learning
Mobile Learning with Intelligent Download Suggestions
SMACKDOWN in the Telecommunications
Course, an Application of the ARCS Instructional Design Model
Breaking communication bottleneck
for student group through online learning system
Towards Semantic-Aware Learning
Object Retrieval – An Ontological
Approach
CASLO: Collaborative Annotation Service for Learning Objects
[Awarded 2004 IEEE TCLT
Small-Scale Research Project award]
ABSTRACT
CASLO consists of a web service-based infrastructure and a client tool used
to collaborate in making annotations to a learning object. Collaboration is
provided through pluggable coordination protocols and evaluation strategies, as
well as auxiliary services like version control, notification and tracing.
CASLO is immersed into a more general collaborative authoring environment of learning
objects.
1. Introduction
From the classic
constructivist perspective, educational models should be focused on the active
and participative role of learners for the construction of knowledge 4. Learning Management Systems (LMS) are key tools to
support educational models, to encourage the active participation of learners,
and to foment their responsibility in the achievement of competencies, whilst
improving relationships between instructors and learners. Current LMS
successfully implement learning object interoperability standards, but the
integration of authoring and collaboration facilities are not tight enough to
support the participative and creative roles that are required by a deeper
constructivist view. Any constructivist approach implies furthering learners’ involvement
in the instructional process. In particular, different stakeholders can play
their role during the creation of any learning material, including
instructional designers, pedagogues, instructors, media designers, and even
students 6. For this reason, support for collaboration in
accomplishing the tasks involved in the learning object creation is welcome.
Authoring of learning
material is not a matter of copy-and-paste chunks of digital media with added media
design, but it is intended as the semi-automated composition of simpler
learning objects in order to provide extended learning conveniences that do not
exist originally. If automation is considered as a process in which the
computer automatically and dynamically composes lessons, and whose result must
have a pedagogical purpose, it becomes clear that the computer should have
access to information regarding the design of instructional material 9. A set of formally modeled metadata allows
incorporating instructional design rationales, among other relevant information,
in a learning object that is being composed 7. Metadata can be used to describe, certificate,
annotate, retrieve, extend or keep an updated history of a learning object.
They can also provide an interpretation of the learning object resources for a
machine-understandable layer (e.g. software agents, sophisticated search
engines or web services) that facilitates their automated processing.
Therefore, metadata are essential, not only to improve reusability of learning
objects, but also as a key authoring task.
But metadata annotation is
usually an arduous and not often attained goal, though metadata specifications
are mostly mature, and a number of annotation tools are readily available. From
our point of view, a major issue is how to provide learning objects with
appropriate metadata annotations in order to increase the automation level for
the composition of learning objects and services.
We took into consideration
the previous rationales to design and develop CASLO, which consists of a web
service-based infrastructure and a tool for the collaborative annotation of
learning objects. CASLO can help a distributed community of experts, media and
instructional designers, pedagogical advisors, and learners, to share the task
of annotating a learning object. This tool facilitates collaboration on
Learning Object Metadata (LOM)-compliant annotation by alleviating its centralization
in a specific role.
The rest of the paper is
structured as follows: Section 2 describes CASLO system architecture; Section 3
shows the required interactions to annotate a learning object, and illustrates
it with a simple annotation case; finally, Section 4 outlines some conclusions
and future works.
2. Overview of CASLO architecture
CASLO consists of two main
elements, depicted in Figure 1. These are a back-end annotation server that behaves
as the collaboration provider and a front-end gateway to the annotation
services. Front and back-end communication is carried through the exchange of
SOAP messages. The front-end part translates clients´ annotation proposals into
the collaboration primitives of the back-end server, according to their WSDL
published descriptions. The front-end client can locate the server and get an
appropriate WSDL specification by searching in an UDDI directory.
Figure 1. CASLO service-oriented
architecture
2.1.
Collaboration service provider
All annotations performed
in CASLO are carried through the collaboration service provider. Therein, the main
collaboration web service accepts annotation proposals before submitting them
to the collaboration protocol. This service also allows defining projects,
participants, discussions, assessment criteria, and timing parameters of the
protocol. The current implementation uses a two-phase consolidation protocol, in
which proposals are subject to peer evaluation before being consolidated into
the manifest file 2. Nevertheless, the system design allows replacing the
collaboration protocol with another desired implementation if required. Once proposals have
passed successfully through the collaboration protocol, they are automatically translated
into changes to the learning object manifest file.
A second web service
monitors all pending tasks that clients must perform (for instance, to assess
annotation proposals coming from other users). The reason to define such service
is a lack in the current versions of UDDI protocol, which cannot track clients’
operations in order to respond asynchronously from the server, nor initiate on
the server an action over the client. Since this service is implemented apart
from the main collaboration service, it can be substituted later, when UDDI
versions evolve.
Finally, every annotation (either trial or consolidated) is automatically logged and version-controlled through its submission to the collaboration server. This way, an updated history of each learning object annotation lifecycle is kept, so it can be recovered to analyze the rationale or authorship of any annotation.
2.2.
Access to collaborative
annotation services
The access to collaborative
annotation services is made through Vizzini, a thin client application that is used to propose,
negotiate, and assess annotations on IMS-packaged learning objects. Since CASLO
has been conceived to provide a general-purpose environment for collaborative
annotation of XML files, the client has been easily adapted to work on learning
object manifest files. Nevertheless, in order to author full-fledged IMS
content packages, a more complete edition tool should be used. Currently we are
providing this by integrating access to CASLO services into Reload editor 5. Nevertheless, all collaborative annotation
functionalities can be readily tested through Vizzini.
3. Interaction with CASLO
The kind of information
that can be managed with CASLO includes any XML file that can be validated
against a given XMLSchema.
In order to structure interactions and negotiations during a collaboration
project, the following elements have been defined:
Ø
System: the set of collaboration
projects, registered users, and users’ participation in projects—. Common users
can try their own annotations, as well as to send assessments for annotations
coming from other users. Users are associated to projects by the system
administrator according to their roles. The project management is delegated to
special users that act as project administrators.
Ø
Project: the set of information
artifacts (i.e. files and resources) that can be subject to collaborative
annotation—. Each project is described by a main XML document file that
provides access to its files and resources (e.g. in the case of learning
objects, the main document is the manifest file).
Ø
Discussion: the frame in which a
concrete annotation decision must be taken—. Each discussion is related to an
item within the main project file. All common users that are registered in the
project can participate in any discussion.
Ø
Item: any XML element that can be
identified within the schema—. Such identification is currently done through XPath, although
other mechanisms are also possible (e.g. id
attributes).
Once the collaboration project is created, it can be parameterized with a given protocol and assessment strategy. Then, interaction between users can start, always according and restricted to the protocol rules. The collaboration proceeds until the protocol rules decide it has to (e.g. a certain degree of consensus was attained; all users achieved a certain level of participation, etc.)
3.1.
User roles
Not all users have the same responsibilities in a project. Three main roles have been defined for project management purposes: system administrator, project administrator and common users. In turn, additional roles have been defined for user interaction purposes. These have been extracted from 2.3.1 attribute role of LOM base schema 3 ―i.e. author, publisher, editor, graphical designer, technical implementer/validator, content provider, educational validator, instructional designer, subject matter expert, etc. Nevertheless, the set of roles is easily extendable to incorporate other specific roles.
3.2.
A Case of collaborative
annotation
There are a number of steps
to be performed before starting a collaborative annotation. At least a
collaboration project must be created, and a set of users must be registered in
the system and the project. As well, the system administrator has to define a
project administrator. The, the project administrator can choose the
collaboration protocol, the timing parameters (e.g. typically hourly, daily, or
weekly), and the evaluation strategy.
Considering that a valid
manifest document is available to all the project users, the collaborative
annotation can proceed as follows:
1.
The initiator user locates the
item she wants to annotate and starts a discussion. The system checks if the
discussion can occur by assuring that there is no other active discussion on
the same item.
2.
If the discussion can be
created, a configurable stand-by period is defined to allow users to propose
alternative annotations. Although the initiator is usually the first in
proposing annotations, she is not obliged to. After that period, if no
proposals are received, the discussion is discarded without reaching an
agreement.
3.
Every proposal sent in a
discussion is managed through the collaboration protocol. In the current
version, every annotation proposal is passed through two phases: broadcast and
consolidation. During each phase, the project users can evaluate the proposal
or send alternative annotations. If there is no alternative proposal, a
configurable timeout sets up the move from broadcasting to consolidation phase.
But if there is any alternative, they are compared to each other, and
eventually one becomes consolidated.
In any moment, each
proposal can be in one of three states: consolidated, in-evaluation, or
abandoned. Figure 2 shows an example resulting from the collaborative
annotation of a learning object, which is an IMS-LD Level B example retrieved
from OUNL DSpace repository 8. The upper right panel shows the content of the manifest file after a number of annotations that
are made to select the correct answer to a question, through editing the
corresponding property and condition. Such annotations were issued by different
users, as shown in the lower left panel, which depicts the collaboration log.
The proposal with id 3 is eventually consolidated, after submitting it to the
protocol, and according to a given assessment strategy —in this simple case,
the higher aggregated score of an annotation, the better evaluation it receives—.
A more complete description of this and other examples can be found in Sourceforge 1.
Figure 2. Example
of a collaborative annotation with Vizzini, one of CASLO front-ends
4. Conclusions
In this work we presented a
web service-based infrastructure and a tool for the collaborative annotation of
learning objects. The aim of the collaborative approach is to facilitate the
authoring and composition of learning objects, which is also a major objective
of the MD2 frame project, in which CASLO is included.
The future lines of work begins by providing access to CASLO services from Reload, a major learning object authoring tool, in order to provide an authoring environment that is more complete than Vizzini. A further step is to integrate the CASLO collaboration facilities inside a model-driven development environment that can increase the automation level required to author and compose new learning objects.
5. Acknowledgements
This work is partly supported
by MD2 project (TIC2003-03654), funded by the Ministry of Science and
References
1. CARLOS Project, available at http://sourceforge.net/projects/carlos/
2.
Dodero, J. M.; Aedo,
5.
Milligan, C. Reload “A recipe
for interoperability in practice”
6. Polsani, P. “Use and Abuse or Reusable Learning Objects”, Journal of Digital Information, 3, 4, 2003, Article No. 164.
7. Sánchez-Alonso, S., and Sicilia, M. A. (2004), Design contract-based selection and composition of learning objects, in Proceedings of the 6th International Symposium on Computers in Education (SIIE), Cáceres, Spain, 2004.
8.
Tattersall, C., and O’Neill, O. “Programmed Instruction”, IMS-LD Level B example,
Retrieved from http://hdl.handle.net/1820/314
9.
Wiley, D. A., “Connecting
Learning Objects to Instructional Design Theory: A Definition, a Metaphor, and
a Taxonomy”, in D. A. Wiley (ed.), The
Instructional Use of Learning Objects, Agency for Instructional Technology
and Association for Educational Communications and Technology, 3-23, 2002.
|
Carmen L. Padrón Departamento de Informática Universidad Carlos III de Madrid Juan Manuel Dodero Departamento de Informática Universidad Carlos III de Madrid Jesús Lanchas Departamento de Informática Universidad Carlos III de Madrid chechu@dei.inf.uc3m.es |
Computational
Sciences in Higher Education: "Have You Created Your Model Yet"
ABSTRACT:
Computational Science software is available
that supports an understanding by students of computer modeling, visualization,
and computational sciences at all levels.
The presentation will demonstrate computational science models that
assist students in building understanding in their classrooms, develop systems thinking
and computational concepts, and use visualization techniques.
With the publication of The Fifth Discipline in 1990, learning organizations and systems thinking have become buzz words in our society. BUT, these are not just buzz words anymore. They represent concepts in which people continually expand their capacity to create results that are truly desirable, in which new and expanding patterns of thinking are nurtured, collective aspirations are set free, and people learn how to learn together. In short, collaboration is used to help students learn how to think and build understanding.
According to the author Peter Senge, the five disciplines are personal mastery, mental models, shared vision, team learning, and the fifth discipline, systems thinking. Systems thinking makes the results of the other disciplines work together for the benefit of the organization (Senge, 1990).
Systems thinking is nothing new. The framework
of the concept has been developed over fifty years, and its essence lies in a
shifting of the mind from linear cause and effect chains to visualizing
interrelationships, and from observing snapshots of change to viewing change as
a process. Barry Richmond, while an
Engineering Professor at Dartmouth College, founded High Performance Systems,
Inc.(now referred to as isee systems), in 1984. His mission was to
make the world a better place by increasing people's capacity for thinking,
learning, communicating, and acting systemically.
What the software also does is eliminate the need for students to create the mathematical formulae to solve the linear equations that create the computer models – the software does it for them, in the background, and may be viewed by selecting the appropriate link at any time. This has lead to a controversy regarding the appropriateness of such a method that purports to teach students how to think. Does removing the need to create the correct mathematical construct also inhibit the process of understanding?
Since computers enable human beings to perform calculations without actually performing calculations, there has been a debate regarding the erosion of that capability. Doctoral dissertation candidates routinely run complicated statistical tests without the need to use a calculator or the proper formulae to arrive at accurate results. When the United States Army Field Artillery School began teaching students to use a computer to solve the “firing problem,” or trajectory algorithm based on ballistic firing tables, instead of teaching the manual method, a hue and cry erupted from the purists who insisted that students continue to be taught the old manual methods in order to understand the process in case the computers did not work.
As we move into the 21st Century, the argument may be made that it is just as important to teach modeling and visualization of concepts as it is to pour mathematical knowledge into the heads of students. “Modern science and mathematics are more concerned with pattern recognition and characterization (http://www.shodor.org/interactivate/activities/rabbits/index.html) than with mere symbol manipulation (http://www.shodor.org/talks/modeling/reality.html). Teachers and students have found that computational science enables them to understand a broad range of phenomena that are otherwise inaccessible. This includes phenomena that are too small (atoms and molecules), too large (galaxies and the universe), too fast (photosynthesis), too slow (geological processes), too complex (automobile engines), or too dangerous (toxic materials) to be studied in the undergraduate laboratory.” (Panoff, 2003).
A new generation of students lead by forward-thinking faculty are now learning how the tools of computational science- interactive numerical simulations, scientific visualization, and computer assisted algebra systems and data mining- may be incorporated into the classroom. The computational sciences education movement is evidence of this, as grant funding from the National Science Foundation supports the endeavor. Thus, the systems thinking software demonstrated in this presentation enables students to achieve the goals of developing understanding and pattern recognition as opposed to the rote memorization of mathematical symbols so soon forgotten after the Final Exam.
A team of faculty and technology support
professionals from the Dunwoody Campus of
The College is now entering into a national partnership with NCSI and six other institutions in order to get a deeper understanding of the important issues that surround undergraduate computational science education. The only way to get inside the issues is to see first hand how computational science is being incorporated at specific institutions. This can include a wide spectrum of activities from using pre-constructed models in a course to illustrate concepts (the most common type of computational science education activity we have seen from our data), redesigned courses with new curriculum components or labs, new computational science courses, new computational science programs (concentrations/majors/minors), educational partnerships with industry/organizations/research groups, outreach endeavors, etc. (Searcy, 2004).
What the GPC/NCSI research partnership
hopes to accomplish is to develop a national community of undergraduate faculty
interested and involved in computational science incorporation in their
classrooms, and have NCSI workshop participants incorporate computational
tools, techniques, and technologies into their teaching. Of utmost importance is the opinion of
students. The research will consider
their perceptions of the computational tools, techniques and technologies they
have used in class (even if they may not have any idea about
"computational science"), and their feedback about using models or
technologies to understand concepts in their courses. It is important to
know what they are thinking about the world in which they will soon be employed
and whether or not they think the skills they are learning now will be useful
later. (Searcy, 2004).
References:
Panoff, Robert, Ph.D. 2003. http://www.shodor.org/talks/modeling/.
Richmond, Barry. 2001. An
Introduction to Systems Thinking.
Searcy, Mary Elizabeth, Ph.D. 2004. A
Partnership to Explore Computational Science Education Innovations
Implementation.
Senge, Peter. 1990. The Fifth Discipline: The Art and Practice of
the Learning Organization.
|
Alan
N. Clark, Ed.D. Director of Technology and Media Group Dunwoody Campus, Tel: +1 770-551-3053 |
Modeling Learning in simSchool
Introduction
Interest in games and simulations in teaching (Prensky, 2002; Aldrich, 2004), and the potential for network-based assessments (Stevens, 1991; Stevens, Lopo & Wang, 1996; Dexter, 2003; Gibson, 2003) has led to developmental research on “simSchool,” a simulation for teacher education. This paper outlines how theoretical frameworks for leadership, learning theory, interpersonal psychology and behaviorist teaching models are being combined into a new synthesis model of classroom learning in simSchool. The model controls how the simulated students learn and behave in response to task, teacher and peer interactions and provides a complex multicausal network of relationships for experimenting with teaching approaches.
A “flight simulator” for teaching
The goal of SimSchool is to provide a realistic and safe practice environment – like a flight simulator for a pilot - for users to develop skills in differentiating instruction, managing a classroom, and adapting teaching to multiple cognitive abilities and special education needs. Future teachers and anyone interested in improving teaching see thesimSchool classroom from the “first person” view of a teacher when experimenting and playing. The simulator works by linking many variables into three broad concepts: the student’s personality, teacher moves, and academic and social behaviors of students.
The personality frameworks, which include the Five Factor Model or “Big Five” (McCrae & Costa, 1996) and the Myers Briggs Type Indicator model (Myers, 1998), define the emotional and behavioral dimensions of the students. Teacher moves include planning and choosing tasks and making “in flight” decisions while class is in session. The model for student academic behaviors (e.g. performance on tasks) has drawn from theories of intelligence and learning theories informed by cognitive science (e.g. (Festinger, 1957; Bloom, Mesia & Krathwohl, 1964; Gardner, 1983; Bloom, 1984; Piaget, 1985; Lave & Wenger, 1991; Perkins, 1995; Bransford, Brown & Cocking, 2000). The social behavior dimension of the students is defined by the Interpersonal Circumplex (Leary, 1957; Kiesler, 1983) mediated by a model of conversational leadership (Lambert, 1995) and social expectations states theory (Cohen & Lotan, 1997). The simulation proceeds from state to state through interactions within and between the components of the student personalities, the task design and teacher interactions, and each student’s social responses within bounds suggested by and determined by these theoretical frameworks.
simStudent Personality
Underlying the simStudent personality, a computational model determines how each student independently behaves in the classroom. Each student has a unique profile that includes their history, academic performance, their social and emotional characteristics. In addition, as the user works with a student over time, the history of the profile changes. The student’s personality is made up of three components, Traits and Needs, Learning Preferences, and Social Expectations. Students have any number (in the trillions) of combinations of these factors and these determine what a student does during teaching (e.g. how they act and what they say in class) and how they perform when prompted (e.g. on performance-based and limited response assessments).
Social expectations settings, for example, determine how the SimStudent reacts when the teacher interacts with other students in the class, and how the SimStudent acts when in a small or large group. Planned additions in this area will take advantage of (Cohen & Lotan, 1997) observations about the state to which a student is driven by expectations of a group about potential contribution to the group’s success. To indicate student responses, students change their posture at their desk (e.g., falling asleep or raising their hand), respond verbally (“I am trying really hard to complete my work”), and display results on performance assessments.
Teacher Moves: Task Design & Verbal Interactions
Task Design processes in SimSchool adhere to standards-based teaching practices and are based on theories such as ”Teaching for Understanding” (Wiggins & McTighe, 1998) and assessment approaches that measure alignment of performance with the demands of a task or an assessment prompt (Shannon, 1999). The tasks design features include cognitive load (Sweller, 1988) of the content as well as loadings concerning the social, emotional and physical requirements of a task.
Task design elements may have direct or indirect and complex relationships to a student’s personality settings on the Big Five (e.g., extroversion) and to the Interpersonal Circumplex (e.g. dominant; see Figure 1). If the instructional level of the task is below the student’s ability, the student gets bored and performance suffers; if above their ability the student gets frustrated and performance suffers and if the task falls within their zone of proximal development (ZPD) (Vygotsky, 1962; Vygotsky, 1978) the student's performance is enhanced.
Figure 1. The Interpersonal Circumplex
Teacher-to-student (and vice versa) verbal interactions take place within the Interpersonal Circumplex (Figure 1), which uses 16 categories arranged along two axes: dominant – submissive and hostile - friendly. Teachers choose statements that represent positions on the circle, and students respond based on their current states, immediate past history and the theory of interaction expressed in the Circumplex Theory. Teachers, through their choices, will decide whether to reinforce the student’s actions positively or negatively.
Other teacher moves include “pre-flight” and “in-flight” decisions that a user can make, such as arrangement of students in the classroom or whether to lecture or do hands-on activities. In-flight decisions to change tasks, groupings, and to interact personally with a student or group, intersect with the individual personality and academic characteristics. Teachers can learn about students from “student files” as well as by watching classroom behavior and considering the measured academic performance of students in response to the planned tasks and classroom actions. Replay and display screens are planned to enable reflection, fine tuning of actions and to document the growth of teaching skills over time.
Figure 2. Complex behaviors emerge as
variables are added.
The combination of these elements creates complexity through the structure of tasks, the student’s personality, and neighbor interactions. As a result, complex behaviors begin to emerge (Figure 2). Multiplying the individual complex behaviors over time by the number of students in a classroom produces a “game” in which the only choice is to develop skills of differentiating and individualizing instruction in order to maximize outcomes for all students.
Conclusion
SimSchool is designed to be a small window into the complexities of teaching. It is an environment for aspiring teachers to practice making decisions about planning, task design, and responding to students with complex personalities and cognitive profiles. Initial research and testing shows promise for creating a “flight simulator” for teachers that models some of the complexity of real teaching.
References
Aldrich, C. (2004). Simulations and the
future of learning: an innovative (and perhaps revolutionary) approach to
e-learning.
Bloom, B. (1984). "The 2 sigma problem: The search for methods of group instruction as effective as one-to-one tutoring." Educational Researcher 13(6): 4-16.
Bloom, B., B. Mesia, et al. (1964). Taxonomy of educational objectives.
Bransford, J., A. Brown, et al., Eds. (2000). How people learn: Brain, mind, experience and school.
Cohen, E. and R. Lotan (1997). Working for equity in heterogeneous classrooms: Sociological theory
in practice.
Dexter, S. (2003). The
Promise of Network-based Assessment for Supporting the Development of Teachers’
Technology Integration and Implementation Skills. Society
for Information Technology and Teacher Education 14th International Conference,
Festinger, L. (1957). A theory of cognitive dissonance.
Gardner, H. (1983). Frames of mind : the theory of multiple intelligences.
Gibson, D. (2003). "Network-based assessment in education." Contemporary Issues in Technology and Teacher Education 3 (3).
Kiesler, D. (1983). "The 1982 interpersonal circle: A taxonomy for complementarity in human transactions." Psychological Review 90: 185-214.
Lambert, L., Walker, D, Zimmerman, D P, Cooper, J E, Lambert, M D, Gardner, ME and Slack, PJF (1995). The constructivist leader.
Lave, J. and E. Wenger (1991). Situated learning:
Legitimate peripheral participation.
Leary, T. (1957). Interpersonal diagnosis of personality.
McCrae, R. and P. Costa
(1996). Toward a new generation of personality
theories: Theoretical contexts for the five-factor model. The five-factor model
of personality: Theoretical perspectives. J. S. Wiggins.
Myers,
Perkins, D. (1995). Outsmarting IQ: The emerging science of learnable intelligence.
Piaget, J. (1985). The equilibration of cognitive structures: The central problem of intellectual development.
Prensky, M. (2002). "What kids learn that's positive from playing video games."
Shannon, A. (1999). Keeping
score: Assessment in practice.
Stevens, R. (1991). "Search path mapping: A versatile approach for visualizing problem-solving behavior." Academic Medicine 66(9):S72-S75.
Stevens, R., A. Lopo, et al. (1996). "Artificial neural networks can distinguish novice and expert strategies during complex problem solving." Journal of the American Medical Informatics Association 3: 131-138.
Sweller, J. (1988). "Cognitive load during problem solving: Effects on learning." Cognitive Science 112: 257-285.
Vygotsky, L. S. (1962). Thought and Language.
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes.
Wiggins, G. and J. McTighe (1998). Understanding by design.
|
David
Gibson |
Testing an interactive lecturing technique for teaching
An
interactive method of teaching and learning was tested in the preclinical
course of basic pharmacology. The method was based on an Interactive Presenter
multimedia presentation system combined with a soft ware for interactive
audience feedback and analysis. One of its main features is an audience polling
system. This gives a possibility for the teacher to ask students questions and
receive their answers during teaching. The response can be received and
analysed immediately. Furthermore, the answers can also be analysed and shown
immediately.
Therefore,
the teacher would have a possibility to evaluate students learning already
during teaching. This gives also a possibility to change the content and
methods of teaching in order to affect on the learning results. E.g. based on
the immediate response, the teacher may repeat something very important,
concentrate on something essential, and perhaps decrease or increase the
content or level of teaching. This system also makes it possible to follow that
the students are paying attention on the teaching. In general, the teacher can
use this method to activate and motivate students and learning, too.
In
addition, the Interactive Presenter system was tested in the multiple choice
examinations. In practice, the questions are first given to the students and
they have some time to prepare for the answers. Next the answers are collected
using Interactive Presenter Audience Polling System. The students are making their
choices, and the answers are analysed. Then, the results are shown after the
examination. Now, the teacher can go through the answers, explain the right
answers and background for the questions. With this feedback, the teacher may
change the examination towards a real learning situation more than having just
a specific examination. This practice may have several advances for teaching,
e.g., by promoting cognitive and constructive active and interactive views of
teaching and learning.
This paper
describes some plans and experiences about the use of the interactive method
for the contact teaching and for the multiple choice examinations. Our aim is
to develop a novel Interactive Presenter-based ‘On site’-model for teaching and
learning. The use of the interactive teaching and learning method will be
tested and studied. An example of using this method for teaching molecular
biology to students of pharmacology is described. The results are discussed
especially in relation to tutoring, orientation, evaluation and response of
teaching and learning.
The
Interactive Presenter can be used for the immediate analysis of students and
audience in different ways. Some examples of the questions are enclosed here.
I. Analysis of background:
where are the students coming from?
I am participating the present course as
1) a 1st year student of medicine
2) a 2nd year student of medicine
3) a student of another field
4) nothing of the possibilities
II. Analysis of activity:
do the students participate in optional examinations?
Do you participate
in the optional examinations?
1) allways
2) sometimes
3) never
4) not decided
III. Analysis of basic
knowledge: how much do the students already know?
Have you
already studied pharmacology earlier?
1) I have a
basic examination in pharmacology
2) I have
studied some pharmacology by my self
3) I have
never studied any pharmacology
IV. Analysis of specific
knowledge: what is the starting level of students to study biotechnology?
What is not
true for biotechnology?
1) biotechnology is already a very old field of human activity
and history
2) biotechnology is actually another name and synonym for gene
technology
2) biotechnologically produced recombinant proteins are used to
treat diseases
3) biotechnological drugs should also undergo pharmacological
investigations
Here, the
Interactive Presenter system can now be used to answer the questions and
analyze the answer immediately. Then, the analyzed data can also be presented
for specific questions and students either in groups or even individually. The
results can be shown as personal results or anonymously. The software calculates
the results per cent of each answer and shows these results as a summary for
each question. The results can be first shown as such without indicating the
right or wrong answers. Then, the right answers will be highlighted. Therefore,
after presenting all the answers, it is possible to continue by immediately
explaining the background for right answers. The information from students and
audience can be used as an orientation to find out the correct level of the
course for further teaching and learning. In this way, it is also possible to
use the IP system during the teaching to follow how well the learning has
progressed, and therefore to repeat something important or to move forward to
more difficult or new issues. Moreover, the IP-system could be used to change
the common lecture to an interactive method of teaching and learning. An additional
advantage of IP system is that the data from the teaching and learning can be
saved in the computer for further pedagogic analysis. This gives the
possibility to effectively use the learning results for the development of
teaching in the future, too.
|
Atso Raasmaja Division
of Pharmacology & Toxicology Faculty
of Pharmacy, FIN-00014
|
ubiNEXT: Exhibition Guide System
that Supports Visitor’s Learning Experience in Museums
1. Introduction
Museums, including art, history, natural
history, science, zoos, aquariums, historic sites and arboretums, strive to
facilitate the learning experience of ever-increasing diverse audience. The number
of museum visitors has increased in the past twenty to thirty years (Falk and Dierking 2000, p. 219). In order to meet the diverse visitor’s needs,
some museums are applying advanced information technologies to museum
multimedia guides and developing the prototypes (Wilson,
G. 2004). We designed an educational museum guide called ubiNEXT
(ubiquitous Networked Exhibition eXplore
Technology) to enhance visitor’s learning experience in museums.
2. ubiNEXT
ubiNEXT provides visitors with Internet and on-site services. The Internet service
has two major features, a pre-visit planning and visit review. Pre-visit
planning is designed for school teachers and visitors, who would like to
personalize their tours before entering a museum. Visit review offers visitors
to continue their museum experience at home or in the classroom. At the museum,
the PDA presents object information and educational activities, with
interactive text, audio and video. It also presents object locations. When the
visitor scans the barcode of an object with the PDA, the related interactive
content is accessed from the server through the wireless network and displayed on the
screen.
3. The Feature of ubiNEXT
As a step further from audio guides and
other exhibition guides available at museums, the PDA provides visitors a
number of object recommendations in which visitors may be interested in seeing
subsequently.
Recommendation is determined by two major factors. One important factor is the allocation of keywords to each object. The system assigns several keywords, such as sculpture, Impressionism, and mythology to each object. If, for example, a visitor was looking at Vincent van Gogh’s “Sunflowers” which includes a keyword “Impressionism,” ubiNEXT will automatically suggests Claude Monet’s “Water Lilies” which shares the same keyword. Such keywords virtually connect different objects exhibited in multiple galleries. The information content presented on the PDA changes dramatically according to each keyword. Thus, it allows visitors to encounter various stories or find different perspectives based on individual keyword.
Another significant factor that contributes
to the recommendation is visitor’s personal preferences. After learning about
objects shown on the PDA, visitors have an opportunity to express their
evaluations on each object. Then, ubiNEXT will give
the recommendation to visitors by calculating visitors interests based on their
personal evaluation, previous browsing of objects and the physical distance
between objects displayed. For instance, if a visitor gives a high rating on
objects that have a keyword, “Impressionism,” ubiNEXT
will provide the visitor a tour that includes objects containing the keyword, “Impressionism.” ubiNEXT
makes it possible for visitors to experience a tour that matches their
interests.
4. Pilot Study
We conducted a pilot study at the
We asked visitors to rate ubiNEXT by answering several questions on a questionnaire.
Among 328 visitors of individual and groups who used ubiNEXT,
142 visitors returned the questionnaire. Among those who answered the
questionnaire, 114 visitors were over 19 years old. The rest, 28 visitors, were
children and youth from
Q1: Did you discover anything new at the
exhibition with ubiNEXT?
Q2: Did your interest level on the
exhibition increase with ubiNEXT?
Q3: Did ubiNEXT
allow you to look at the objects in depth?
Figure
1: The average rating of adults over 19 years old
We divided the questionnaire results according to the age group. Figure 1 shows the average rating of adults over 19 years old and Figure 2 shows the average rating of children and youth 7-18 years old.
Figure
2: The average rating of children and youth 7-18 years old
There was a significant difference between the age group of visitors over 19 years old and visitors 7 to 18 years old. Overall, younger visitors 7 to 18 years old gave a more positive evaluation on ubiNEXT than adult visitors over 19 years old. ubiNEXT provided a better learning experience for the younger visitors. Some of the factors that may have influenced the difference were adult visitor’s previous knowledge about the exhibition, the content materials targeted for children, and young visitor’s unfamiliarity with the exhibition content.
5. Conclusions
We introduced the museum exhibition guide
system that was particularly designed to support visitor’s learning experience.
We also discussed the pilot study at the
References
Falk, J.H., and L.D. Dierking. (2000). Learning from Museums: Visitor Experiences
and the Making of Meaning.
Wilson, G. (2004). Multimedia
Tour Programme at Tate Modern. Museums and the Web 2004.
|
Aya Masuoka,
Toru Takahashi, Takugo Fukaya Advanced
Telecommunication Research Institute (ATR) Media
Information Science Laboratories Machi
Takahashi Advanced
Telecommunication Research Institute (ATR) Intelligent
Robotics and Communication Laboratories |
Intelligent
Agents: a New Paradigm to Support Collaborative Learning in Distance Education
Systems
Introduction
Current distance and open learning devices attempt to reduce the difficulties encountered by learners when they follow a distance course. In particular, laboratory experimentation/practice is inconvenient because the students have to be physically present in the university laboratories. In local laboratory experiments, students usually work together in groups of two or more. This learning paradigm is often called collaborative learning. One solution for this problem is usage of virtual collaborative environments which bring together users geographically dispersed but connected via a network.
Another inconvenience in distance education is tutor’s difficulties when he follows a distance collaborative learning process, and in particular those participants who cannot keep up progress with their group mates. A solution to overcome this inconvenient can be smart agents which can partially replace a human-tutor in the collaborative learning process of distributed student-teams.
Implementation
A synchronous groupware named PASSENGER was developed at University of Duisburg-Essen throughout the last years. This groupware is a distributed application for participants in partially not foreseeable places. The system was designed to support collaborative learning activities and especially for the usage in Software Engineering – Education.
The Passenger-Client (see bellow) consists of a communication component, a cooperation component and several shared tools and resources to carry out Software Engineering tasks. It contains video screens of each member and a whiteboard that is divided in a public window for common process on the outline documents and a private window for individual process on the documents.
Each member has the same view of the public window according to the What You See Is What I See (WYSIWIS) principle, but only one of them can alter the document at a certain time. Each member is also equipped with a private working window to try out own ideas and to work simultaneous on an individual solution. The client interface has also a function which permits a member to transfer the content of the public window into its own private window.
Passenger Floor Control (PFC) handles the access to shared resources and coordinates the course of communication via the administration of different kinds of permissions, e.g. permissions to speak, permissions to alter the documents etc.
The creative tasks within the collaborative group processes, respectively the modification of the common artefacts take place in the common work area. The Public-Window area of the Whiteboard is managed by the PFC.
A Telepointer serves to elucidate and present the facts. The Telepointer is implemented as a collaborative service which can be used by the Floor-Holder. This user can lead the attention of the other participants to some objects or screen areas of the public window during a discussion.
For the aspects of group awareness, each participant is always placed into the same video screen. Each participant appears always in the left corner of its own screen. The Tutor (agent or human) is always in the right corner. All other participants are associated to a fixed position with the help of an ID-number assigned by the Passenger Server. If a participant is missing, the associated position is empty.
Group awareness functions are implemented by means of providing all needed information for a late coming in participant to discover the actual conference state. This is implemented by highlighting the video screen of the person who has access to the shared resources.
We integrated animated-agents in Passenger in order to overcome the lack of tutor when the students meet outside tutor working’s hours. We consider agents in a virtual society of learners as entities that occupy a social position and perform several roles. To exemplify the concept in our environment the tutor-agent should have besides the tutoring-roles like:
Ø
Interrogator – poses questions and the students of a collaborative group then
provide answers. The questions should
provide help for the students to reach a common learning goal.
Ø
Reviewer – analyzes the students’ answers, including whether it is correct or
not.
Ø Monitor – records the answers from all the students and the communications among students during the collaborative learning process.
Ø Instructor – gives individualized instructions and helps those students who cannot keep up with the progress of their group-mates.
also:
Ø Group Manager – controls the coherence of the group. For this role agent’s requirements are: to monitor the owners of the PFC, to control the PFC access grant/take the PFC to/(from) the inactive users. Also it must assure a fair distribution of PFC among participants in a learning session. This role tries to solve one of the open problems in the collaborative virtual environments: communication issues among participants.
Results
The evaluation study concerns two levels:
Ø Usefulness level: the usefulness of the agent facilities within Passenger groupware needs to be evaluated by human teachers.
Ø User friendliness level: highlights how the agent was accepted by students.
Several experiments concerning the second part of the evaluation study were conducted among 25 first year Master-students. Each session consisted of three students and one tutor (human or agent). The students experienced the traditional lab with the human tutor and also with the agent feature. After these experiments, students had to answer to questionnaires files. A sample of questions is the following:
Ø
Do you consider the application
attractive? If yes, what did you like about it?
Ø
Do you think that the “agent”
features prevented you from understanding the educational process better?
Ø Do you prefer the agent tutor instead of the human tutor? Please justify your answer.
Although the number of participants in the
evaluation test was rather small for a quantitative evaluation, the trends seem
to be unambiguous. Based on these questionnaires several
statistics could be made. Some results concerning the agent integration and acceptance
are shown bellow.
Conclusions
Within this paper it was shown how agent technology’s presence in a synchronous virtual environment can overcome several major inconveniences of distance education systems like: awareness and communication.
|
Bogdan-Florin Marin, Axel Hunger |
The Civic Nexus Project: Supporting Sustainable Technology Practices in Geographic Communities
Civic Nexus (http://cscl.ist.psu.edu/public/projects/nexus/)
is a three-year participatory design project with the goal of working with
geographically-based community groups to understand how they learn about
technology and develop the capacity to implement technology projects. In the
Civic Nexus project, we study technology use and work practices occurring in
nonprofit organizations as they work to achieve social and communitarian goals.
Nonprofit community organizations are important to study because they employ a
significant number of workers and they address social issues that impact the
well being of most communities. In the
The Civic Nexus project builds on previous
work that takes a long-term participatory design approach that combines
ethnographic fieldwork with participatory design to create information systems
that address local needs (Carroll, Chin, Rosson,
& Neale, 2000). We are now adapting this
participatory design model to the study of nonprofit community organizations.
For the last 18 months, we have conducted fieldwork and have identified (and in
some cases carried out) technology projects with community partners located in
Our research focus in the Civic Nexus project is on sustainability: finding ways of working with groups to help them develop the capacity to solve technical problems and to implement technology projects that meet their needs (Merkel, et al., 2004, 2005). Technology can play an important role in helping community organizations to achieve their mission. Websites and newsletters can be used to advertise services, recruit volunteers, and shape public discourse around a cause. Databases can play important roles in record keeping and in assessing a group’s effectiveness. Calendars can be used to organize work and to promote an organization’s activities. While there are significant opportunities, many community organizations face challenges when implementing IT in their organization because they often lack the financial and human resources and experience with technology needed to complete a project.
Because of our concern with sustainability, one of our research directions has been to consider the role that designers can play in facilitating community control over technology. Our research and design efforts involve finding ways for our community partners to take control of the design process itself by directing what should be done and by maintaining the technology infrastructure and the outcome that is produced. In our work, we have considered a number of different roles that emphasize the less directive role that we wish to play including that of a lurker, facilitator, consultant, and bard (Carroll, 2004).
Tied to taking more of a facilitative role in the design process, we have worked to identify technical and conceptual tools that promote long-term learning and technology planning. We have found that conceptual tools such as scenarios can be used by community groups to help facilitate the design process (Lee, et al., 2005; Farooq, et al., 2005). For example, in an environmental group that we worked with, we introduced the use of scenarios as a way to think about how an audience might use their website and to elicit design requirements. We also found that designers can play an important envisionment role, helping community groups see new ways that technology might be used in their organization.
Another central theme in our work is studying how groups manage technology learning in their organization. This goes beyond working with community groups to help them learn a set of discrete set of technical skills (e.g. skills required to update a website). It also includes helping groups make choices about how to implement technology in ways that connects to their work practices, history, and organizational roles. For example, we worked with a group of high school students to develop an online health course for their school (Xiao, et al., 2005). This required the teacher who directed the project to take more of a consulting role because she did not have the technical knowledge that her students possessed to complete the project. The project also served as a proof of concept that the technology could be used causing the teachers to reflect on (and in some cases resist) the use of online material in the classroom. We have found that there is no one size fits all strategy for working with community groups and that part of our role as designers involves provoking reflection on what counts as a “good” use of technology for a group (Lee, et al., 2005; Merkel et al., 2005). This has also led us to reflect further on the ways that designers can support the informal learning practices taking place in community organizations (Rosson & Carroll, 2005).
Finally, we have considered some of the unique structural features of nonprofit groups that impact design requirements for community-oriented information systems. For example, community groups often rely on volunteers to perform tasks central to their mission. This raises issues for community-oriented systems design such as group access control (e.g., specifying which volunteers can update or modify parts of a website) and knowledge management (e.g., as volunteers leave how can the group preserve and develop their understanding and use of technology).
As we move forward with the project, we will look for reusable strategies such as patterns that can be used to address sustainability in community computing settings (Carroll & Farooq, 2005). This includes the strategies that community groups have developed to manage technology in their organization. It also includes strategies that can be used by designers to encourage community groups to participate more fully (and even direct) the design and implementation process.
Extensions
The Civic Nexus project is partially supported by the National Science Foundation under award IIS 03-42547.
References
Carroll, J. M. (2004). Participatory design
of community information systems: The designer as bard. In F. Darses, R. Dieng, C. Simone,
& M. Zacklad (Eds.), Cooperative Systems
Design (pp. 1-6).
Carroll, J. M., Chin, G.,
Rosson, M. B., & Neale,
D. C. (2000). The development of cooperation:
Five years of participatory design in the virtual school. In D. Boyarski and W. Kellogg (Eds), DIS
2000: Designing Interactive Systems (pp. 239-251).
Carroll, J.M. & Farooq, U. (2005).
Community-based learning: Design patterns and frameworks. Paper
to be presented at the 9th European Conference on Computer-Supported
Cooperative Work (
Farooq, U., Merkel, C., Nash, H., Rosson,
M. B., Carroll, J. M. & Xiao, M. (2005).
Participatory design as apprenticeship: Sustainable watershed management as a
community computing application. In Proceedings of the
Lee, R., Merkel, C. B., Ganoe, C., H. Schafer, W., & James, K., Rosson, M. B., & Carroll, J. M. (2005). Making use of scenarios for achieving effective
use in community computing contexts. Paper to be
presented at HCI International,
Merkel, C., Clitherow,
M., Farooq, U., Xiao, L., Ganoe,
C., Carroll, J., & Rosson, M. (2005, January 31).
Sustaining computer use and learning in community computing contexts: Making
technology part of “who they are and what they do”. The
Journal of Community Informatics [Online], 1(2). Retrieved
Merkel, C. B.,
Xiao, L. Farooq, U., Ganoe, C. H., Lee, R., Carroll, J. M. & Rosson, M. B.
(2004). Participatory design in community computing
contexts: Tales from the field. In Proceedings of the
Participatory Design Conference, Vol 1, (pp
1-10).
Rosson, M. B.
& Carroll, J. M. (2005). Minimalist
design for informal learning in community context. In Proceedings of the Conference on
Communities and Technologies
Salamon L.M., Sokolowki, S.W.,
& List, R. (2003). John hopkins comparative nonprofit
sector project: Global civil society at-a-glance.
Xiao, L., Merkel, C. B.,
Nash, H., Ganoe, C., Rosson,
M. B., Carroll, J. M., Shon, E., Lee, R., & Farooq, U. (2005). Students as
teachers and teachers as learners. In Proceedings
of the
|
Cecelia Merkel (cmerkel@ist.psu.edu) Lu Xiao (lxiao@ist.psu.edu) Umer Farooq (ufarooq@ist.psu.edu) Craig Ganoe (cganoe@ist.psu.edu) Mary Beth Rosson (mrosson@ist.psu.edu) Jack Carroll (jcarroll@ist.psu.edu) Computer Supported Collaboration and Learning Laboratory Center for Human Computer Interaction The |
TutorBot: an Intelligent System to support
Tutoring on line in an e-learning platform
ABSTRACT
In the
present work, an intelligent system called TutorBot, used as a support for
tutoring activities on-line in a web-learning environment and based on the AIML
language will be described.
TutorBot allows to automatize
the interaction processes between tutor-learners
guaranteeing forms of flexible learning in terms of space-time, thus minimizing
the use of human resources, often lacking for the management of activities that
involve a large target of users in a web-learning environment.
Introduction
From the
analysis of e-learning systems, particularly dwelling upon the activities of
tutoring on-line, some strategies and methodologies for the planning and
implementation of intelligent systems have
been individualized, in order to optimize the learning processes on the web.
The
studying models have been analyzed inside the GRIAD (GRIAD is our Group of research (Department Economics
and Statistics, University of Calabria)) and are
founded on two essential technical-scientific
areas: the Artificial Intelligence (AI)
and the Web-based Informative Systems (WIS),
from the integration of which up-to date systems can come out for the information retrieving, which can be
implemented in an e-learning platform for the optimisation of the contents
management [1].
The study
in this direction has led us to the realisation of a software system called TutorBot, which
founds its principles on the AIML (Artificial
Intelligence Markup Language) technology applied
and experienced in an e-learning environment [2].
TutorBot presents itself as an intelligent agent able to interpret the
requests made by the learners in natural language and to give the answers
through a friendly interface, with
the advantage of emulating in an efficient way the human figure, which in the
specific case is represented by the one of the Tutor during the activities of tutoring on line.
Later on, a
description of TutorBot
will be given first to understand better its running, the structure of its
knowledge base, and finally the main expected benefits will be shown.
TutorBot: an AIML based application
The studies
on the applicability of the AIML language in e-learning contexts and the need
of optimizing the processes for the recognition of natural language, in order
to support the tutoring activities in
a web-based learning environment, has
led us to the realisation of TutorBot. The system knowledge base founds its main
characteristics on a series of AIML data objects, that are
associated to appropriate meta-tag according to particular rules that
substantially derive from the XML language. The informative base is structured
in particular AIML files, which following the aiml
standard can be easily produced from external sources as well and imported
inside the database of the LMS to
enrich the TutorBot
knowledge base [5],[6].
We want to
emphasize this last aspect, since according to this scheme, it encourages the
fulfilment of high levels of interoperability and sharing of resources, so
auspicate in the e-learning field and object of international research.
The main AIML data objects used for the management of the TutorBot Knowledge Base (KB) are listed as follows:
Ø
<topic>: is the meta-tag that defines the
content topic object of treatment; this is very important above all in a
context characterized by a high complexity of
knowledge, in which the same concept can refer to different topics;
Ø
<pattern>: in this object, the information
that will be useful for the matching
with the requests in the dialogue phase between TutorBot-learners are structured;
Ø
<template>: this object contains the relative
answers to the patterns;
Ø
<category>: includes in it the elements pattern and template.
An example
of AIML code that shows how the same request can have different answers
according to the reference topic, is represented as
follows:
<aiml>
<topic name="history">
<category>
<pattern> What
do you think about
</pattern>
<template>
</template>
</category>
</topic>
<topic name="geography">
<category>
<pattern> What
do you think about
</pattern>
<template>
</template>
</category>
</topic>
</aiml>
The
mechanism at the basis of the AIML engine is a stimulus-answer type: to each user’s Query, an appropriate answer
that finds the best correspondence between the possible ones is given and takes
the reference topic into account, after the intervention of an appropriate
algorithm which takes care of the Input redefinition, that takes into account
the rules concerning the symbolic reduction of the text, insertion and scission
of sentences, the synonyms management, the grammatical corrections etc. For an
exhaustive explanation, please consult [2].
Interaction between TutorBot and learners
TutorBot
as thus projected can be integrated inside an LMS as a further means of
communication/interaction to guarantee a form of tutoring on-line usable at any
moment and anywhere [3],[4]. The learner has a friendly web interface at
disposal, inside of which he can directly build his questions (Query) in
natural language, as if he was asking them to a human tutor. The Query in
object is intercepted by the artificial tutor and is simplified through an
appropriate algorithm in its grammatical form, also individualizing in this
phase the reference topic. At this point, a matching operation happens inside
the database, and in the case in which TutorBot
manages to find the answer, he directly gives it to the learner through the
same interface, leading to the dialogue in real time.
When he
will obtain the relative answer to the proposed query, the learner will be able
to continue the dialogue with TutorBot, thus
establishing a continuous interaction man-machine that emulates as far as
possible the one between man-man. The running scheme is represented in figure
1.
Figure 1: Interaction TutorBot-learner
The
mechanism we have just described has been subsequently extended to give a
solution even in the circumstances in which
TutorBot would not be able to find an answer to
the formulated Query, since absent in its database. In this case, a software
module has been integrated for the support of TutorBot, that allows to send the “unsatisfied” Query automatically to a research
engine in order to fill up the verified informative lack. The realised research
Output will be proposed as template
to the learner, using the same interaction interface [3]. A part of the AIML
code that allows this operation is illustrated as follows:
<category>
<pattern>_
<template> The
capital of
can find other information about
resource:
<search query="http://160.97.33.198/cgibin/search/search.pl?">
<topicstar/>
<star/>
</search>
</template>
</category>
Conclusion
The
described system represents in conclusion an innovative tool for the processes
management in an e-learning environment that, integrated in an LMS, is able to optimize either the back-office activities or the front-office ones. Thanks to this tool,
a constant activity of tutoring on-line is guaranteed, in which the learner can
interact at any moment and from anywhere with the artificial tutor in order to
use the contents and have deepenings on these latter.
Another important aspect is represented by the possibility of making the system
able to manage possible verified “informative lack”, having recourse to the
integration of the research engine that is integrated in an automatic way and
in background with TutorBot. Finally, the system
being implemented in AIML language, thanks to this standard, it is easily
possible to share the resources between heterogeneous platforms as well [5].
References
[1] De Pietro
O., Appratto F. (Oct. 15/19/2002), Weblearning: aspects of a new paradigm, E-Learn 2002,World
conference organized by AACE, Proceedings CD Rom,
[2] De Pietro
O., Frontera G., TutorBot:
an application AIML based for Web-Learning. 7° IASTED International Conference on "Computers and Advanced Technology in Education" (CATE 2004), Proceedings conference, pp. 284-290,
[3] De Pietro
O., Frontera G., De Rose M., Automatic update of AIML
Knowledge Base in e-Learning environment. 8°
IASTED International
Conference on "Computers and Advanced
Technology in Education" (CATE 2005),
[4] H. Shi, S. Revithis,
[5] Artificial Intelligence Markup Language (AIML) Reference Manual. Available at: http://www.alicebot.org/documentation/aimlreference.
html.
[6] Symbolic Deduction in Artificial
Intelligence Markup Language (AIML). Available at: http://www.alicebot.org/documentation/srai.html
|
De Pietro Orlando Department
of Economics and Statistics, De Rose Maurizio Department
of Economics and Statistics, Frontera Giovanni Department
of Economics and Statistics, |
Awareness : Framework for e-learning
ABSTRACT
E-learning technology provides the tools necessary for collaborators (e.g., instructor and students) to achieve work (i.e., tasks and assignments) interactively. "Awareness" is one of the important features that should be carefully designed in an e-learning environment. While working virtually, collaborators will need to know necessary information (no more, no less) about the content of the event, the process, and the people who are involved. This article explores further the issue of awareness, and highlights the necessary awareness elements in a proposed theoretical framework.
What is “awareness”?
"aware-aware·ness (noun) 1: archaic: watchful, wary, 2: having or showing realization, perception, or knowledge." Excerpt from WWWebster's Dictionary. Awareness support in CSCW systems’ design is crucial, since "Knowledge of group and individual activity, and co-ordination are central to successful co-operation. These factors are clearly critical concerns in the design of computer systems […]." (Dourish and Bellotti, 1992).
In general, when people use technology to
communicate, various types of information are communicated through many
channels, both implicitly and explicitly. Types of awareness information vary
from awareness of documents, projects, and tasks i.e., workspace, to awareness
of the location and activities of co-workers i.e. social. Tele-pointers, office
snapshots, video glances, document/project tracking, and background noise are
some of the various forms of providing awareness which have been used up to
date in latest e-learning technologies. In creating support for awareness,
considerations include: what information to provide, how to provide it, how to
give users control of the information, reciprocity, privacy, and interruptions.
Awareness and interaction can be both direct and indirect. Direct awareness and interaction are represented by for example, eye-contact, facial expressions, and gestures, while indirect awareness and interaction are represented by for instance, gaze, peripheral awareness, and synchronisation. The following are the types of awareness support found in the literature with brief explanations and examples of systems.
Research in e-learning applications e.g., conducting e-seminar, e-lecture, and e-workshop has resulted in highlighting what needs to be in awareness from the stakeholders’ points of view. Typical stakeholders (roles) in an e-learning session are the presenter, the facilitator, the system administrator, and the participant. What needs to be “in awareness” is clarified in the following section, which raises the proposed framework (Figure 1).
Figure 1 Triadic framework for “Awareness” support (Al-Matrouk, 2003)
Components of the framework are the awareness issues:
(1) Terms of Reference (TOR)
The following distinct items of information
need to be shared and collectively understood:
Ø
The tasks/assignments
specification i.e., objectives and scope.
Ø
Roles played in the event.
Ø
Subject, topic, or problem
clarification.
TOR is the agenda for the event, which includes information concerning the topic, brief information of what it is about, its scope, a list of its objectives and goals, the presenter and brief background, people who are going to manage it, and the structure of how it is going to operate. The structure lists and clarifies the stages of the event, time schedules and corresponding activities.
(2) Virtual Session Management (VSM)
This is mainly related to providing
information about how the process is managed. Awareness of the roles, their
responsibilities, and tasks (i.e., what they are supposed to do.), and the
certain pre-requisites for managing the virtual event become crucial. “Virtual
Session Management” denotes all the tasks that are involved in assuring the
event runs properly and effectively.
(3) Substantive Knowledge (SK)
SK is concerned with the content of the virtual event (i.e., the e-learning session), specifically with information relating to the topic of the event, e.g., from the presentation itself, related questions and answers (QA), and discussion.
(4) Achievement of Understanding (AU)
Understanding in the process of conducting a virtual event is a cognitive human issue that is concerned with how well knowledge is delivered, perceived, and comprehended between the stakeholders (e.g., instructors and students) during the process. AU is associated with how well the process is structured, conducted, and maintained according to a well-conceived and designed agenda, which is also closely related to virtual session management (VSM).
(5) Biographical Work and History Information (BWHI)
The provision of presenters’ and participants’ biographical information is extremely useful in creating a social and friendly working environment. It is even more useful to support this with information about their work information (job titles, affiliations, work positions, etc.), and work experiences to establish the background work knowledge.
(6) Presence
Presence is an awareness issue, which indicates to the participants of a particular virtual session who is present, engaged in the process, and available for contact.
(7) Involvement and Contribution
IC generally refers to participants’
interactions and activities with facilitators, the speaker, and each other in
relation to the process at all phases i.e., prior to, during, and after the
presentation. Participants’ IC is specifically represented in answers to
questions like ‘Who said?’ and ‘In response to who and what?’
with respect to the talk, QA, and discussion. Participants’ IC also includes
the acknowledgement of one’s own questions and contributions during
sessions.
Conclusion
“Awareness
is still about the observability of relevant
information and the media used to render observable entities and activities” (Berard and Coutaz,
1997). Working contexts and users’ requirements vary, and so do awareness
requirements. Understanding awareness needs in a working context requires a
closer investigation and research, and there is not a generic guidance to
follow in the design of awareness in an electronic system. The work setting
defines what needs to be ‘in awareness’. The framework highlights the
high-level awareness requirements which are specific to the ‘delivering an online electronic presentation’
context.
References
Al-Matrouk, H. S.
(2003), “Awareness Support for Electronic Presentations: Guidelines for User
Interface Design”, PhD thesis submitted to the
Berard, F. and Coutaz, J.
(1997). “Awareness as an
Extended Notion of Observability”. ACM CHI’97,
CLIPS-IMAG, BP 53, 38041, Gernoble,
Dourish, P. and Bellotti, V., Awareness and Coordination in Shared Workspaces. In Proceedings of CSCW'92 (Toronto Ontario, Nov 1992), ACM Press.
|
Dr.
Hasan Saleh Al-Matrouk Kuwait Institute for Scientific Research |
Using a Knowledge Management System for E-Learning
ABSTRACT
This paper relates our experience with using a knowledge-sharing approach as adopted in a post-graduate diploma course to part-time students employed in an ICT-intense work environment. The need was for an extremely flexible internet-based knowledge management system to allow the quick and efficient inter-organizational distribution and management of course information including group work, assignments, readings and multi-media documents. This was achieved by using a state-of-the-art proprietary knowledge management system (“Archi”) instead of the more traditional, less flexible e-learning systems such as WebCT. It was found that this approach is particularly useful for managing complex, dynamic and information-intensive part-time courses.
Description of the Distributed Knowledge Management System
Archi
Archi is a
proprietary web-based knowledge management tool that integrates an
object-oriented knowledge repository, a customizable meta-model and a variety
of user interfaces. It is based on a multi-user, three-tier model of browser
client, web server and database server. It can manage a wide variety of
knowledge asset types such as word processing documents, web-documents,
presentations, models, plans, and multi-media files (http://www.inspired.org).
Figure 1: Screenshot of Archi's GUI
Its primary goal is to provide a platform for distributed knowledge management within information-intensive, professional group work environments such as business analysts, IS project managers and other professionals. Archi is driven by a customizable meta-model of the information which can distinguish various conceptual types of information held, along with rich attribute sets. The information base is navigated and queried using a standard web browser by means of a number of user views, such as the unified, item, cross reference, calendar, worksheet, tree and spatial views.
Using Archi for a part-time
post-graduate course
Archi has been used to support two part-time post-graduate Information Systems courses since 2001: the post-graduate diploma and honours course. These are geared primarily towards IT professionals wishing to obtain a quality post-graduate qualification. The courses are characterized by small student numbers (20 to 40 pa). Most hold a demanding IT job and spend relatively little time at the University of Cape Town (UCT) outside the weekly contact sessions. All students have access to excellent computer facilities and the Internet at work, if not at home. They are generally highly IT literate. The courses have a high student-participation input, consisting of a large number of student group presentations in addition to the lecturer-driven presentations and work shops. Several individual student deliverables are also required.
In 2001, the course convenor Graham McLeod decided to use the courses as a beta-site for Archi. The paradigm of a distributed knowledge-base was considered to be an ideal match with the course attributes. The conceptual course meta-model was set up in a matter of hours and has required only slight refinements over the years. A typical snippet of the model is whereby the course object (e.g. “PG Diploma 2004”) has various attributes such as a title, duration, description, etc. It is linked to a number of other objects in the database, including persons (examiner, coordinator, students), documents (books, FAQs, articles etc.) who can all have different relationships or roles. For example, a document can constitute a course prescription, required reading for a lecture session or other course-related activity (e.g. group meeting or examination session), a specification for an essay etc. (see figure 1).
Student Experiences with Archi
Two different cohorts of students were surveyed, one group in 2001 and another 2004. Students use the system a lot, both from work and from home. The most common uses were the downloading and uploading of course materials, including class presentations and readings (95%-100%). The calendar feature was also used by many (73%) but only a minority followed the hyperlinks to non-essential references available inside Archi (41%) or the world-wide web (only 27%).
Figure 2: Student Feedback on Archi
As shown in figure 2, the most common problems experienced were bandwidth (which is not related to Archi but an infrastructural problem) and the complexity user interface. The latter has been modified and upgraded significantly during the last few years, with the addition of a student dashboard and a number of customized views, but the problem still persists. Strangely, although the user interface is fairly complex – possibly partly due to the conceptual course model – the actual learning curve appears to be quite short. On the other hand, students like the functionality of Archi, and most would prefer continued use.
Staff Experiences with Archi
Archi is a very stable tool. Software bugs are rare and its robustness is a pleasant contrast to some other web-based tools. Once set up, Archi requires very little maintenance. It currently runs on a “junked” webserver and, true to its claims runs with different clients (user tier), different web servers (middle tier) and relational databases (back-end tier).
Perhaps the biggest challenge lies in its flexibility. Unlike WebCT, which has basically a fill-in-the-blanks configuration, Archi requires the convenor to set up (or buy into a pre-installed) conceptual course model. Academics not used to working with multi-tier course administration tools, lament the latencies and user interface constraints which are inherent with a web-based platform.
An important practical constraint is the fact that Archi has a relatively small installed user base, which makes system support an important issue. Due to the university regulations, this also means that it has to be managed and maintained by the department itself rather than forming part of the university-wide technical IT support structure. Finally, it is also difficult to see Archi being adopted in its current incarnation in course environments where the students are less IT proficient, or do not have access to the same rich IT infrastructure at home or in their work environment.
Summary
The IS Department at UCT uses a number of different web-based paradigms to enhance the quality and efficiency of the student learning experience. These range from standard course web pages to WebCT. However, it was found that the use of a distributed web-based knowledge management system could fill a useful niche for very knowledge intensive courses with lots of cooperative learning.
References
Inspired 2004 Archi:
Knowledge Management for Professionals and Consultants (Release 2.3.6). Available http://www.inspired.org/html/archi.htm
(Accessed
Konrad, J (2003) Review of Educational Research on Virtual Learning
Environments European Conference on Educational Research,
Van Dam, N. (2004) The e-Learning Field
Book.
|
Jean-Paul
Van Belle Private Bag, Rondebosch,
7700, Tel: +27-21-6504256 Fax:
+27-21-6502280 |
Introduction
Learning
object is any entity, digital or non-digital, that may be used for learning,
education or training [1]. Learning object metadata describes any kinds of
learning objects, such as study courses, the pedagogical features of the
course, the contents, special target groups, and the technical requirements of
the study course.
Learning
object metadata is used for many purposes:
Ø
It
facilitates the learners’ decision-making process and aids the educational
institutions to provide suitable information about their course supply.
Ø
It
is also needed for improving the retrieval of learning objects, for supporting
the management of collections of learning objects.
Ø
It
is also used for guiding non-experienced users through a large collection of
learning resources
In addition
to these features we are extending the ONES-system [2, 3, 4]
in away that we can use learning object metadata for automatic composition of
learning objects.
Learning Object Content
Model
The
composed learning object (in our learning object content model) is a course composed of lessons. Further each lesson is comprised of a video, slides and textual presentation (Figure 1).
We have
attached metadata to each lesson (for now lessons are seminar presentations
given by the students).
The
composition of a course is based on the metadata item subject. To standardize the values of this metadata item we use
taxonomies. In Figure 2 is an example of taxonomy.
Each subject item is specified by picking
appropriate nodes of the used taxonomy and giving weights on the picked nodes
such that the sum of the weights equals to one. The weigh of the nodes that are
not picked are zero. So mathematically, subject
item is a vector in the vector space which dimensions are the nodes of the
taxonomy.
A nice
feature of this approach is that by varying the weights we can express the subject
of the lesson as well as the level of the lesson. To illustrate this, let us consider the
lesson “The future of educational metadata”.
So it is obvious that the weights should be given at least on the nodes
“Educational metadata” and “Metadata”.
Further, if the lecture is strongly focused on educational metadata then
it should have rather high weight (e.g., 0.8). So the more specific the lesson
is the higher the weights of the lower level nodes have in its subject item.
Specifying Composed
Learning Objects
The
composed learning object is comprised of a set of lessons. Each lesson is
specified by a profile. The profile has same structure as the subject item of a lesson, i.e., a set of
picked nodes of the taxonomy such that the sum of their weights equal to one.
For
example, assume that we want to specify a course “Metadata and ontologies in the semantic web”. Its first lesson should be an introduction to
the topic, and so its profile should have high weights on the nodes near the
root of the taxonomy, e.g., weight 0.3 for the node Web and weight 0.7 for the
node Semantic web. If we want that the
next lesson focus only on metadata, we should set weight 1 to the node
Metadata. If we want that the third lesson deals different classes on metadata
we could give for example weight 0.4 for the node Educational metadata and
weights 0.3 for the node Descriptive metadata and Semantic metadata.
Computing Composed
Learning Objects
Composed
learning object are produced by an Object composer (Figure 3.).
The object
composer has three modules: User interface, Composer and the Metadata and
taxonomy database. The database includes lessons metadata (including subject
items)and taxonomies. When the Composer receives a request
(a set of lesson profiles), it performs a distance query for each lesson
profile and returns the lesson having the best match.
The used
matching algorithm [5] calculates the cosine measure between the query (a
vector) and the lessons profiles (vectors). As a matter of fact the algorithm
does not compute distance measures but rather approximates distance measures by
computing the angles of the query vector and the vectors representing learning objects.
Conclusions
Automatic
composition of learning object requires that we must have a model of learning
objects. For now, as the only form of learning objects that we have are lessons
also our content model is rather simple. However, in proportion as we store
other kind of learning objects on the database we will extend our content
model.
Automatic
composition of learning objects also requires that we must have some way to
describe the subject of learning objects as well as the subject of composed
learning object. We argue that by using the vector model instead of only using
keywords we can achieve much higher accuracy with respect to the content of
composed learning objects.
References
[1] LOM – Learning Object Metadata. (2002).
http://ltsc.ieee.org/doc/wg12/LOM_1484_12_1_v1_Final_Draft.pdf.
[2] J. Puustjärvi, and P. Pöyry. Searching learning objects from virtual universities. In the Proc.of the Internationational Workshop on Multimedia Technologies in E-Learning and Collaboration. (http://research.microsoft.com/workshops/WOMTEC/WOMTEC_papers/WOMTEC_128.pdf). 2003.
[3] J. Puustjärvi. Integrating e-Learning Systems. I)n the Proc. of the International Conference on Web_Based Education (WBE2004). 2004
[4]J. Puustjärvi.
Using one-stop portal in integrating eLearning
systems. Advanced Technology for Learning.
1(2): 99-106. 2004
[5] Baeza-Yates,
R.& Ribeiro-Neto, B. 1999. Modern Information Retrieval.
Addison Wesley.
|
Juha Puustjärvi Laboratory
of Communication Engineering |
ABSTRACT
In this paper, we propose a scheme to provide intelligent
download suggestions for mobile learning contents using
artificial
intelligence. To deliver learning content via mobile devices, including PDA’s, cell phones, pocket PCs, or
other handheld devices,
the file size must be considered. The success of downloading a
large file depends on the Internet traffic at the moment.
According
to the induced decision tree, we can predict whether the downloading would be
successful. If not, the server would suggest the user not to download the large
file at the moment or to download an alternative file.
1. Introduction
The Internet-based environment has been rapidly replacing the traditional lecture-based one due to due to the rapid growth of information and communication technologies. More and more people try to apply the artificial intelligence techniques, such as the agent technology [1][2], to the application of distance learning. In 2004, Büchner and Patterson suggested platforms to keep track of learners’ activities including content viewed, time spent and quiz results [3]. In [4], the authors presented a PDA-based interface for a computer supported system. Some researchers use mobile videophone and e-learning to help a student with severe physical impairments [5]. In [6] and [7], the authors developed a museum tour guide system on PDAs. Many studies have been developing software tools that can in used in various mobile devices to personalize the learning context. In this paper, we propose a scheme to provide intelligent download suggestions for mobile learning contents using artificial intelligence.
The remainder of the paper is organized as
follows. In Section 2, different mobile learning devices are introduced.
In Section 3, we present
the scheme to provide intelligent download suggestions for mobile learners using artificial
intelligence. The mobile learning platform is described in Section 4. Finally, the conclusions are made in Section 5.
2. Mobile Learning
Mobile learning has been defined as learning that takes place via
wireless devices as mobile phones, personal digital assistants (PDAs), or laptop computers.
To deliver learning content via mobile devices, including PDA’s, cell phones, pocket PCs, or other handheld devices, the file size must be considered. Large files, such as video clips, are not recommended. Some alternatives like audio files might be considered.
Table 1 illustrates users might download mobile learning contents at anytime, anywhere. The success of downloading a large file depends on the Internet traffic at the moment. In this study, we propose a scheme to provide intelligent download suggestions for mobile learning contents using artificial intelligence.
Table 1. An example of users’ profile
when downloading a large file during a time slot
|
|
Attributes |
|
|||||
|
User |
Location |
Time Slot |
|
Thruput |
Largest File Size |
Wireless Network |
Successful |
|
No.1 |
163.26.228.12 |
|
No |
500Kbps |
800K~ |
|
Yes |
|
No.2 |
203.67.153.76 |
|
Yes |
250Kbps |
800K~ |
802.11b |
No |
|
No.3 |
60.248.40.208 |
|
Yes |
400Kbps |
500K~800K |
802.11b |
Yes |
|
No.4 |
210.243.17.97 |
|
Yes |
450Kbps |
> |
802.11b |
No |
|
No.5 |
N/A |
|
No |
30Kbps |
< 50K |
GPRS |
Yes |
|
No.6 |
222.156.11.196 |
|
No |
200Kbps |
50K~500K |
802.11b |
Yes |
|
No.7 |
220.22.71.103 |
|
Yes |
300Kbps |
500K~800K |
802.11b |
No |
|
No.8 |
N/A |
|
No |
20Kbps |
50K~500K |
GPRS |
No |
|
No.9 |
61.63.100.137 |
|
Yes |
600Kbps |
> |
|
Yes |
|
No.10 |
203.71.160.118 |
|
No |
400Kbps |
500K~800K |
802.11b |
Yes |
|
. . . |
. . . |
. . . |
. . . |
. . . |
. . . |
. . . |
. . . |
Figure 1. The decision tree to predict whether a large file would be downloaded successfully
3. Artificial Intelligence
In Figure 1, we see a tree
used to determine whether the downloading is successful or not. This
kind of tree is called a decision tree
[6][7]. A decision tree takes in the attribute values
and outputs a Boolean decision. Once a decision tree is constructed, we start
at the root node to check the attribute value. Based on the attribute value, the
branch labeled with the corresponding value is chosen. Continue checking the next
attribute value until a leaf node is reached. According to the induced decision tree, we
can predict whether the downloading would be successful. If not, the server
would suggest the user not to download the large file at the moment or to
download an alternative file.
3.1. Entropy and Information Gain
Entropy measures the impurity of a data set
S and is defined as:
Entropy (S) = 
(1)
where P+ is the proportion of positive examples while P-
is the proportion of negative ones. The entropy reaches its maximum value of 1
when half of examples are positive and half are negative.
3.2. Decision Tree
Decision tree induction involves a set of
training data to generate a decision tree that can classify the training data
correctly. If the training data represent the entire space of possible data
adequately, the decision tree will then correctly classify new input data as
well. The best-known decision tree induction algorithm is ID3, which was
proposed by Quinlan in 1980s. The algorithm builds a decision tree from top
down and finds the shortest possible decision tree to classify the training
data correctly. The method used by the algorithm to determine which attribute
to be chosen for each node is to select the attribute that provides the
greatest information gain at each
stage [6][7]. A perfect attribute divides the examples into sets that are all
positive or all negative. The information gain of an attribute tells us how
close to perfect the attribute is. Information gain is defined as the reduction
in entropy and calculated by:
(2)
where Entropy(S)
denotes the entropy of a data set S and can be calculated by equation 1. 
is the expected
reduction in entropy due to sorting S on attribute A.
4. Mobile Learning Platform
Figure 2. The user interface of
playing mobile learning contents
The user interface of playing and downloading mobile learning contents are shown in Figure 2 and Figure 3, respectively. According to the above induced decision tree, we can predict whether the downloading would be successful. If not, the server would suggest the user not to download the large file at the moment or to download an alternative file.
Figure 3. The user interface of downloading mobile learning contents
5. Conclusions
In this paper, we propose a scheme to provide intelligent download
suggestions for mobile learning contents using artificial intelligence. The decision tree tells the importance of each attribute. After
the decision tree is constructed, we can see that not all attributes are needed for the prediction; in
other words, the download suggestions might be made by only 2 or 3 attributes. Unfortunately, it is quite
possible that even when vital information is missing, the decision tree
learning algorithm will find a decision tree that is consistent with all the
examples. This is because the algorithm uses the irrelevant attributes. In the
future, some technique to eliminate the dangers of over-fitting must be
developed in order to select a tree with good prediction performance.
6. References
[1] Y.M. Huang,
J.N. Chen, K.T. Wang, C.H. Fu, “Agent-Based Web
Learning System,” IEEE Learning
Technology Newsletter, Vol. 6, Issue 2, pp. 38 – 41, April 2004.
[2] Y.M. Huang,
J.N. Chen, S.C. Cheng, W.C. Chu,
“Agent-Based Web Learning System Applying Dynamic Fuzzy Petri Net,” Proceeding of International Conference on
Web-Based Learning, pp. 338 – 345, 2004.
[3] Alex G. Büchner, David Patterson, “Personalised E-Learning Opportunities”, Proceedings of 15th
International Workshop on Database and Expert Systems Applications, pp.410 –
414, 2004.
[4] A. Riera, D.L.
Moreno, E. Sánchez, M. Lama and X. Vila, “A PDA-based
Interface for a Computer Supported Educational
System”, Proceedings of the 3rd IEEE International Conference on
Advanced Learning Technologies, pp. 12 - 16 , 2003.
[5] K. Kanamori, I. Kobayashi,
"Mobile Videophone and E-learning for Students with Physical Impairment", Proceedings of the 2nd
IEEE International Workshop on Wireless and Mobile Technologies in Education, pp. 203 -
204, 2004.
[6] Ching-Chao Lee, Chun-Yen Chang, Li-Der Chou and
Ming-Yu Lee, “A Tour Guide System for
[7] Chia-Hsieh Wu, Chen-Chow Lee, Jui-Ming Chen, Li-Der Chou,
Shih-Pang Ho, "Requirement Analysis and Implementation of Palm-based Multimedia Museum Guide Systems", AINA 2004. 18th International Conference on Advanced
Information Networking and Applications, vol.1, pp. 352 - 357,
2004.
[8] Stuart Russell, Peter Norvig, “Artificial
Intelligence: A Modern Approach”, Prentice
Hall, pp.531-544, 1995.
[9] Ben Coppin,
“Artificial Intelligence Illuminated”, Jones
and Bartlett, pp.276-286, 2004.
|
Shu-Chen Cheng Department of Computer Science and Information Engineering, Chih-Wei Su Department of Computer Science and Information Engineering, Yen-Ting Lin Department of Computer Science and Information Engineering, |
ABSTRACT
3D
modelling of human organs is a project that combines multimedia technology and
3D animation development. The aim of this project is to develop an interactive
learning material for early stage medical students and individuals who are
interested to get knowledge about parts of the human body. The 3D images for
the selected human organs are created for this project so that the user can
view the models and use all multimedia applications such as viewing the movie
of 3D models, narration for explaining the functions of the modelled organs,
and also to do a quick indexed search of scientific terms for medical uses all
by navigating a user friendly interface. In short, the main features of this
project are to provide information about human organs in a 3D environment.
Introduction
3D modeling
of human organs and skeleton is not a new technique in Information Technology
since their function is to reinvent medical purposes especially in education.
In the area of learning the medical clinical anatomy, there are some
exceptional programs that can run and combine an established multimedia
computer-based education. The programs that help students to learn about
anatomy, pathology and radiology but programs about learning an interactive
multimedia of 3D human organs and skeletons are very insufficient in the market
now with several weakness where their interfaces are still somewhat confusing
for the user and the information are presented using text to deliver the
details about the the models.
3D Human Organs is a
project that combines multimedia technology and 3D animation development. The
3D images for the selected human’s organ are created for the user to view the
models with use of multimedia applications such as viewing the movie of the 3D
models, narratory explaination
of the functions ,and quick search indexing of
scientific medical terms all by navigating through a user friendly
interface. The target is to reduce costs
in medical learning industries and increase efficiency of the usage of 3D
technology especially while doing basic medical learning process.
Materials and Methods
This project is
implemented using a software called 3D Studio Max for designing the 3D model of
human organs; Adobe Photoshop for editing and createing
suitable graphics designs for the page’s linking button and the interface. The
Sound Forge software is used for recording and editing the sounds. The Video
shows are exported from the source to this combination. Then all the multimedia
elements are combined together using macromedia software called Flash MX. Most of the objects are created using
polygons then applying extrude to the polygons and shaping them according to
the actual objects. For objects that are relatively complicated, surface
combination and modifier are used.
Figure 1 shows the main interface of the software.
Basically,
the main features of this project are to provide information about human organs
with 3D view of the organ in concern. The User can view the rotation of the 3D
models while listening to the explanation from the narrator.
To select a new 3D model image the user has to move the mouse over the image. An eye icon will appear along with the name of the item next to the cursor or pointer. Selecting one of the buttons will display a new page of the human’s organ 3D model. Selecting the text icon will provide information on the image the user is viewing. Figure 2 shows a 3D model of the human skull, Figure 3 shows a 3D modeling of a human backbone, and Figure 4 shows one complete backbone 3D model with textual information alongside the image.
Figure 1 shows the main interface of the software.
Figure 2 shows modeling of the
human skull
Figure 3 shows the modeling
of the backbone
Figure 4 shows the complete backbone model
Results and discussion
Creating an
interactive learning will attract fullest attention from thousands of people to
put their effort to learn using maximum hours per day. User-friendly, 3D
models, colorful interface can get user attention if the
information is put on an interactive CD.
Therefore the information about the organs and skeleton selected ; 3D model, data, images, audio, plus a special
features such as quizzes and quick search are organized in a nice interactive
CD. The CD includes special features
such as try and error quizzes which the user can click on the button for
quizzes, then a short questions page appears about the organs or skeleton that
the user chooses. The second features is quick search whom the user can utilize
o find the meaning and description of difficult or unfamiliar scientific terms
referring to the alphabet indexing. The
interactive CD also includes a movie of injury for the models. The user can view the short movie of the 3D
animation while the organ model is making a rotation and the user can click to
see the model from a 3 dimensional view; top, front and perspective view.
Conclusions
3D Modelling of Human
Organs is a project that focuses on applying 3D modeling
and animation technology for the learning of medical student.
The project’s target
is to develop various new techniques in creating a product that will enhance
and simplify the clinical anatomy oriented learning. In this high- technology world hopefully this
project will increase the efficiency in using paperless learning while reducing the
costs and expenditures in the educational system and also attracting more
people for gaining medical knowledge.
Also, 3D Modeling of Human Organs is created
specifically for medical students. In addition, this is an opportunity to
develop further understanding and explore the 3D modeling
and animation world deeply. In short,
this project uses 3D modeling and animation
technology as a new step in Clinical Anatomy basic learning that will attract
more students and public to simplify their medical learning process.
References
[1] ICI Paints (1996): “Clinical Oriented Anatomy”, (
[2] SARGUR N.SRIHARI. “Representation of Three-Dimensional Digital Image” , December 1981.
[3]HUMAN ORGANS, Internet site address:http://www.arg.org/
[4]MODELING,
Internet site address:http://www.acm.org/
|
Majdi Al-qdah and Norkhairanizan
bt Che Mohd. Khan Faculty
of Information Technology Jalan
Multimedia, 63100 Cyberjaya Selangor, Malaysia |
Administrators at all levels of the educational hierarchy are attempting to survive a sociological phenomenon, the increased utilization of technology (Harder, 1997). Technology for both instructional and administrative uses has created a schizophrenic atmosphere for educational leaders (White, Ringstaff & Kelley, 2002). Improvements in administrative and instructional efficiency will not be significant unless a comprehensive and systematic plan for evaluating the technology and its application to the educational organization is both present and used.
Accountability
Accountability has always been a concern
for educational leaders (Kirst, 1990); however, only
recently has technology become a focus for accountability. The globalization of
the 21st century fuels the current accountability because the future of
The long-range purpose of technology accountability is to enhance productivity. A second powerful application of accountability is comparing productivity across organizational components. Also, technology accountability is necessary to measure quantitatively the investment in educational technology. There are three models used to measure accountability: the Efficiency/Effectiveness Model; the Cost-benefit Model; and the Utility Model.
Efficiency/Effectiveness Model
Efficiency and effectiveness coexist in complex educational organizations (Waite, Boone & McGhee, 2001). Educational organizations can temporarily survive without perfect efficiency, but they usually die if they are ineffective. The mutual goals of increased efficiency and improved effectiveness are vital to hardware and/or software purchases. All purchases and technology applications must reflect both increased efficiency and improved effectiveness.
Efficiency typically implies short-term accountability, while effectiveness connotes long-term accountability. According to Kohn (2000), “it is easier to measure efficiency than effectiveness, easier to rate how well we’re doing something than to ask whether what we’re doing makes sense” (pp. 3-4). Drucker (1974) stated: “Efficiency is concerned with doing the right things. Effectiveness is doing things right” (p. 45).
Cost-benefit Model
Educational leaders are forced to objectively measure and quantify persons, programs and processes. The discrepancy of quantity over quality creates an unhealthy ethos that threatens to destroy the very persons and programs that technology accountability measures are designed to quantify (Berliner & Biddle, 1995).
The threat of measuring and improving aspects of operation not essentially vital to overall organization productivity is called the threat of "suboptimization"--optimizing the performance of secondary, or even irrelevant, aspects of operation. The sub-optimization threat is especially valid in today's large and complex educational institutions where there are literally hundreds of unique units (Guile, 2001). The currently accepted model of technology accountability becomes one of political posturing, not one of sound educational practice. This means that technology purchases should take place when supporting data can predict the benefit of the cost incurred.
Utility Model
The concept of utility implies that administrators
must examine the overall percentage of the institution that utilizes the
technology.
Applications
Accountability measures offer administrators
data to substantiate changes to enhance productivity. The key concept is determining how to measure
productivity in service organizations like schools and colleges (
The
Change is difficult and often produces results
that are not intended, therefore, organizational capacity to meet technology accountability
demands must be critically examined (Olsen, 2000). Educational leaders need to understand
technology accountability, know their organizational capacity and be able to
clearly articulate the role technology accountability plays in their
organization. The size of the accountability movement indicates that the
survival of public education may very well hinge on the ability of educators to
demonstrate productivity and accountability in a chaotic
marketplace (Fusarelli, 2001) .Technology accountability
is here to stay, but it must be understood in relation to the contextual reality
of schools and colleges.
References
Berliner, D. C., & Biddle, J. B.
(1995). The manufactured crisis: Myths, fraud and the attack on
Drucker, P. F. (1974).
Management: Tasks,
responsibilities, practices.
Fusarelli, L. D. (2001). The political construction of accountability: When rhetoric meets reality. Education and Urban Society, 33(2), 157-169.
Ginsburg, R., &
Goodlad, J. L. (1979). An ecological version of accountability. Theory into practice, 18(5), 3083-315.
Guile, D. (2001). Education and the economy: Rethinking the question of learning for the ‘knowledge’ era. Futures, 33(6), 469-482.
Harder, C. S. (1997). Are we better off? Policy Review,(84), 57-61.
Houston, P. D. (1999). Skeptical about our nation’s accountability agenda. School Administrator, 56(10), 54
Johnson, W. R. (2001). Meeting accountability or evading responsibility. Theory Into Practice, 18(5), 372-378.
Kirst, M. (1990). Accountability: Implications for state and local
policymakers.
Kohn, A. (2000). The case against
standardized testing.
Newman, F. M., King, M. B., & Ridgon, M. (1997). Accountability and school performance—Implications from restructuring schools. Harvard Educational Review, 67(1), 41-74.
Olsen, F. (2000). New data on technology spending offer benchmarks for college administrators. Chronicle of Higher Education, 47(9), p A46.
Raywid, M. A. (2002). Accountability: What’s worth measuring? Phi Delta Kappan, 83(6), 433-437.
Waite, D., Boone, M., & McGhee, M. (2001). A critical sociocultural view of accountability. The Journal of School Leadership, 11(3), 182-203.
White, N., Rigstaff, C., & Kelley, L. (2002). The learning return on our educational technology
investment: A review of the findings
from research.
|
James
F. Burnham Michael
D. Richardson Howard
C. Woodard |
A collaborative approach for development of Arabic courses for e-learning, A case study of Tunisian-Saudi Arabian experience
ABSTRACT
In this
paper, we present a case study of collaborative experience to develop an Arabic
course for e-learning. This course is developed at distance jointly in
Introduction
The learning objects paradigm [3] is deeply rooted in the evolution towards a knowledge society, which entails the need for lifelong learning and for more flexible, adaptive learning systems, inside and outside the public education system [2]. With the spread of new technologies of Information and Communication, we are living a huge demand for Web-based resources for work and learning and thus for the reusability and interoperability of digitized information materials.
The use of learning objects is a relatively new emerged approach and technique for modeling and implementing learning contents. It enables reusability and sharability of learning material between several institutions around the world that offer courses on the Web.
This case study
starts from a common need to Arabic courses for initiation to basics of
computer science in particular courses such as words, excel,
…etc. Such courses will be used for professional training in
The initial French MS-word course is
developed in the research Unit of Technologies and Communication of the
Description of the experience
The
methodology used to translate the course consists in treating every learning
object separately, i.e. if the learning object is a text we have just to
translate the content. If it is a video sequence, the task consists on
recording a new sequence in Arabic corresponding to the initial French sequence, … etc.
All new
developed materials are exchanged between the above authors by e-mail in order
to validate them and eventually to correct and/or modify them.
Evaluation of the cost of development
As first
results, the first chapter has been ready in a very short period. This result
shows that the cost of development of a new course is extremely low
comparatively with the initial French course. For instance, there is no new
investment in the design of the structure for the course
(story boards) nor in the graphical charts and templates.
The new
investment in the content development is mainly devoted to translate texts.
However the new investments in the technical implementation consist mainly of
the new recording audio and video sequences after just translating their
associated texts. Modifying flash animation by replacing
French texts to Arabic format.
Moreover, images are modified only if they contain texts.
Regarding the flash animation, the programmer has to deal with the problem of compatibility between Arabic and French since the writing direction in Arabic is from right to left.
Concerning
now the learning object metadata, we had in fact two aspects to consider : metadata content and metadata content
manipulation. Two choices were possible : we rewrite
metadata in Arabic to permit their interrogation in Arabic. This alternative is
easy to implement but it needs that the author has to deal with metadata every
time he added new content in Arabic. However the second alternative, that we
adopted, consists in conserving initial metadata and adding information about
existence of an Arabic version. With this solution, the author is completely
discharged of metadata rewriting, however we
need a tool that translates Arabic requests to French. This new
investment is justified due to the importance of the amount of translated
metadata content.
Conclusion
We have
shown in this paper that it is possible and useful for Arabic educational
institutions, professors, and learning content providers to contribute to the
usage of learning objects.
First
results of this collaboration seem to be promising in both pedagogy and
research. In pedagogy, we plan to continue development of the whole of the
course. In research we are already working in designing automatic tool to
translate and/or analyse Arabic metadata based requests in learning object
paradigm.
References
[1] Henda CHORFI
& Mohamed JEMNI, "Evaluation
and Perspectives of an Innovative Tunisian e-Learning Experimentation",
International Conference Advances in Infrastructure for e-business,
e-education, e-science and e-medecine on the
Internet,
[2]
Paquette G.,(2003) Instructional Engineering for
Network-Based Learning. Pfeiffer/Wiley Publishing Co, 262 pages.
[3] Downes, S. (2001) Learning objects. International
Review of Research in Open and Distance Learning 2(1). Available : http ://www.irrodl.org/content/v2.1/downes.html
[4] Houcine CHEBLI & Mohamed JEMNI, The Virtual University
of Tunis, Realisations and Perspectives, The Third International Internet
Education Conference, October 11-13, 2004 in Cairo, Egypt.
|
Mohamed Jemni Ecole
Supérieure des Sciences et Techniques de Tunis Research Unit of Technologies of
Information and Communication (UTIC) 5, Av. Taha
Hussein, B.P. 56, Bab Mnara
1008, Abdullah Ali
Bahattab Computer Technology Department Jeddah, Saudi
Arabia |
SMACKDOWN in the Telecommunications Course, an Application of the ARCS Instructional Design Model
Today’s twitch-speed students, raised on Nintendo, MTV, cell phones and instant messaging, require new and innovative strategies for instructional design if the teacher is to capture and hold their interest (Tapscott 1997). In the October 2004 edition of the Learning Technology Newsletter, Tim Roberts (2004) applied John Keller’s (1983) ACRS (attention, relevance, confidence, and satisfaction) model of instructional design to curriculum design for distance learning. This article extends Roberts’ ideas into the realm of instructional design for an introductory course in data communications through the use of an unlikely metaphor; pay-per-view wrestling.
Keller’s (1983) instructional model consists of four phases; arousing interest, creating relevance, developing an expectancy of success, and producing satisfaction through intrinsic/ extrinsic rewards, and has become known as the ARCS (attention, relevance, confidence and satisfaction) model. Roberts (2004) established the link between learning, interest, and motivation and showed how the ARCS model improved learning in distance learning courses. This paper describes an extension of this linkage to the presentation of content in a data communications course, specifically network physical and logical topologies, channel usage alternatives, and Ethernet addressing via an ARCS format designed to increase student interest, motivation and thereby learning.
To reach the students, the teacher needs to raise the interest level of the students. The first two components of the ARCS model, arousing of interest and establishing of relevance, support just this objective. Therefore, using the metaphor proposed by this paper, we introduce the concept of pay-per-view wrestling as a surrogate vehicle to demonstrate the actual course concepts. Additionally, the introduction of such an unexpected metaphor raises the level of humor in the class which has been shown to be an additional tool for raising student interest Tomkovick 2004).
Almost all university students in the
The third stage of the ARCS model is “developing an expectation of success”. Since the student’s have personal experience with the pay-per-view cable technology and the discussion is somewhat light-hearted due to the objective of obtaining professional wrestling programming, the students accept that the underlying concepts are manageable and that they can master them. This conclusion has been confirmed through empirical observations and interactions with numerous classes and their performance on subsequent examinations that cover this subject area.
The final stage of the model, satisfaction, occurs as the explicit knowledge contained in both the textbook and lecture notes becomes internalized as tacit knowledge on how this pervasive technology functions.
The components of the class to present the desired learning concepts are introduced using the following scenario. First the general subject area is introduced through the use of an unexpected statement/question on the order of “Let’s talk about watching pay-for-view world wide wrestling on TV.” This opens the subject discussion by raising the first component of the ARCS model, attention. The class is queried about what are the requirements to watch a pay-per-view match on television. The discussion ultimately evolves to describe the solution as an addressable cable TV converter box. This answer leads to the question of how to identify one specific cable box from the other cable boxes in the neighborhood and how this relates to the physical topology of the network. The telephone system is used as an example of a star network to contrast with the bus topology of the cable TV system. The need for an addressing scheme that works at the instrument level in a bus network is explored. This discussion leads into the analogous function of the address in the cable TV converter and the MAC (media access control) address used by Ethernet for identification of sending/receiving nodes. Once the students grasp the concept of the addressable converter, there is little difficulty extending concepts used in the metaphor to the real subject, the Ethernet address.
The discussion then moves to how cable TV uses the coaxial and fiber optic cable to transmit the different programs. This becomes the basis for a discussion of broad-band vs. base-band use of the media and how the cable TV systems employs frequency division multiplexing to allow the simultaneous transmission of 100+ channels of programming.
Experience has shown that use of the ARCS
model’s structure applied to personally relevant activities leads to a highly
interactive and dynamic learning environment.
The addition of humor also helps to involve the students in the learning
activities. The metaphor of purchasing pay-per-view programming via cable
service is very appropriate for students in the
References
Keller, J.M. (1983) Motivational design of instruction, In C. Riegeluth (ed.), Instructional Design Theories and Models Hillsdale, NJ:Erlbaum.
McLaughlin, K. (2001) The Lighter Side of Learning, Training, Vol. 38, Iss. 2
Roberts, T. S. (2004) What Interest Rate
are You Getting?, Learning Technology Newsletter, http://lttf.ieee.org/learn_tech/issues/octover2004/learn_tech_october2004.pdf
Tapscott, D.
(1997) Growing Up Digital: The Rise of the Net
Generation, McGraw-Hill,
Tomkovick,
C. (2004) Ten Anchor Points for Teaching Principles of Marketing, Journal of Marketing Education, Vol. 26, Iss. 2
|
Randall
C. Reid Richard
G. Platt MIS and Management Department |
Breaking communication bottleneck for student group through online learning system
Effective communication stands at the nexus of the whole process in knowledge sharing. There are barriers in communication that require a system that can tackle holistically the identifiable issues [1]. Communication does not only involve sharing common languages but also the artifacts being used as the media in the knowledge transfer, common communication time for both parties and common social identities. Social identities are referred to the race, culture, sex and personality of the communicator. When the common language is shared by the communicators, the meaning, interpretation and understanding of the language are almost uniformed. Communication requires sufficient time to be allotted to both communicators for them to listen, think, analyze and respond. Artifacts are sometimes used as communication media for visible and audible illustrations in the form of video, sound and documents presentations. For example, the road map is a communication media for an effective communication in giving direction. Social identities may affect communication such that one is more comfortable to converse with someone s/he already knew. We investigate three of these factors by conducting experiment of group discussion in two modes, namely the traditional face-to-face and the online discussion.
The three factors that are hypothesized in this experiment are that
i) online discussion which provides comments that are associated with artifacts as communication media will make understanding about the contents in the artifacts easier. Artifacts and communicated messages that are stored in accessible form are preferable for future reference;
ii) online discussion provides ample time for the discussants to think and rationalize what to be said and consequently enable him/her to provide more ideas and personal thought;
iii) online discussion provides less stress, less inferiority complex in posting ideas and less worry with others knowing her identity.
The online discussion is performed using Intelligent Conversational Channel (ICC) that was developed to facilitate the communication among the students [2,3]. The system allows the students to discuss, argue and suggest within the group through the internet environment. The students are assigned to nine groups and each group consists of 5 – 6 members. Each student is given a login ID based on the student ID and these information is not displayed through out the entire knowledge sharing session. The students display themselves using nicknames for the purpose of referring to each other during the online discussion. Therefore, a student can appear as more than a single person if two nicknames are chosen. The students can upload articles (pdf, Microsoft word or ps), pictures (bitmap, tiff, pcx), application files (excel, powerpoint), video files as artifacts attached together with messages explaining their opinions, arguments or summaries.
The experiment is conducted on Management Information System class where there are two modes of group discussion. The first mode uses traditional face-to-face group discussion and the second mode is an online using ICC system. The students are asked to investigate several topics and discuss them. The group discussion runs for 10 weeks for both modes. The traditional face-to-face group discussion is conducted during the class period and extendable outside of the class. The online discussion is an outside class activity and the system is accessible at anytime. At the end of the 10th week, the students are required to fill in survey questionnaire.
The following table shows the question statements that investigate the three factors stated earlier. There are 40 students who participate in the survey. All of the students are involved in Project 1 and Project 2. Project 1 is the face-to-face group discussion while Project 2 is the online group discussion. In the question statement, the student can circle either P1 (Project 1) or P2 (Project 2) or both (if they feel the statement is true for both projects). For the question statement that requires False or True, the student is allowed only to choose one. The counts are the total number of circles marked by the students.
|
No |
Factors |
Question statements |
Counts |
|
1 |
Comments that are associated with artifacts as
communication media will make understanding about the contents in the
artifacts easier. Artifacts and communicated messages that are stored in
accessible form are preferable for future reference |
The gist from the articles can easily be
understood from the comments given by the group members |
P1 – 12 P2 – 29 |
|
The artifacts should be kept in a repository for
future reference by other group members who are interested in sharing similar
knowledge |
True – 36 False – 4 |
||
|
2 |
Online discussion provides ample time for the
discussants to think and rationalize what to be said and consequently enable
him/her to provide more ideas and personal thought |
I have more time to think and rationalize of
what being discussed by the group members |
P1 – 14 P2 – 29 |
|
I contribute more ideas and personal thought
without hesitation |
P1 – 19 P2 – 27 |
||
|
3 |
Online discussion provides less stress, less
inferiority complex in posting ideas and less worry with others knowing her
identity |
I have less stress in group discussion |
P1 – 17 P2 – 27 |
|
I find that less feeling of inferiority complex |
P1 – 20 P2 – 31 |
||
|
I do not have to worry about people knowing myself when ideas are expressed |
P1 – 12 P2 – 30 |
For Factor 1, there are two question statements that support the claim. There are 29 (P2) students feel that the articles are well-understood from the online discussion even though there 12 (P1) of them feel the opposite. One of the count claims that the question statement is true for both environment. The second question statement receives majority supports that knowledge repository is important in knowledge sharing.
For Factor 2, there are two question statements to confirm the claims. Both of the question statements receive more confirmation from the counts that online discussion (P2) provides more time for the communicator to think and rationalize as well as contributes more ideas and personal thought.
For Factor 3, the three question statements find that the online discussion (P2) provides less stress, less feeling of inferiority complex and less worry about people knowing the identity of a person who posted the message. Our result here may differ from [4] as the discussion period for this experiment is longer.
The students are a mixture of different races mainly Chinese, Indians, Malays and foreign students. Their social identities are not known among them such that they find the online discussion that does not display their actual identities make them more comfortable in posting their comments.
Note
The
research was previously conducted in Faculty of Computer Science and
Information Technology,
References
[1]
S.M.F.D Syed
Mustapha and T. Nishida. Communicative Social Intelligence: Technological
Concepts for Breaking Communication Impediments. International Journal of
Web-based Communities, Inderscience Publishers,
[2] S.M.F.D Syed
Mustapha. Intelligent Conversational Channel for Learning Social Knowledge
among Communities. 8th
International Conference on Knowledge-based Intelligent Information &
Engineering System, KES 04, New Zealand, Vol I,
pp343-349.
[3]
S.M.F.D Syed
Mustapha. Theoretical optic on enabling PBL activities for large group through
intelligent conversational channel. 5th Asia-Pacific Conference on
Problem-based Learning, 16 & 17 March, 2004. CD
publication.
[4]
Allan, J and Lawless, N. Stress
caused by online collaboration in e-learning. Learning Technology, Vol 5, Issue 4, October 2003. Available at
http://lttf.ieee.org/learn_tech/issues/october2003/#6
|
Syed Malek
Fakar Duani Syed Mustapha, PhD |
Towards Semantic-Aware
Learning Object Retrieval – An Ontological Approach
Introduction
In recent years, several international organizations have devoted themselves into building e-learning standards, such as SCORM [1][2] and IEEE LOM [3], which are proposed for describing the structures and contents of integrated teaching materials. Following the standards or no, there are currently many Learning Objects available for corporate and academics. Despite the available of such growing libraries, e-learning now faces a more pressing challenge: how to find the most appropriate contents for a learner, advanced or not. Some of the standards, such as SCORM LOM, facilitate the locating of learning objects from a repository by extended sharing and searching capabilities. For example, a learner can search via keywords, data, author, or any other metadata fields. However, because SCORM LOM defines over 60 fields, an average learner will find it difficult to completely specify the conditions in such a large number of attributes.
There are other significant problems in learning object retrieval:
(1) It is quite different between traditional document retrievals and e-learning objects retrievals. The major problem is that it’s almost impossible to analysis the contents of a learning object for specific search terms. This is because the contents of learning objects are various; it could be a text, a web page, a flash, an audio, a piece of video, or any kind of media.
(2) It is hard and unreasonable to ask a learner to be familiar with the LOM standard, in contrast to the pure text keywords in document retrievals.
(3) Providing a menu of fields for a user to fill in makes the system inflexible and a learner still has to familiar with various types of metadata. That is, although the titles and the descriptions of a field is well defined, the contents of learning objects still depend on the common usage of languages, glossaries, expert opinions, personal experiences, and so on.
(4)
Most systems just provide the
capability of keyword match, and are deficient in semantic search and query
reasoning.
One way to overcome the problems mentioned above is by using well-defined learning object metadata ontology and by indexing learning objects according their meaning rather than some keywords. An ontology [4] defines a set of common understood terms, called the concepts, and relationships among these terms. In a well defined ontology, the concept hierarchy and the relationships can also be used to evaluate the semantic distance and the relevant degree.
The Proposed Ontological Approach
We propose an architecture for Semantic-Aware Learning Object Retrieval; it is shown as in figure 1.
The main purpose of using ontological representation is to gain the ability of inferring a user’s intention, even though the user may not know what he/she wants exactly. In our model, the followings are the key works to reach the goal.
Ø Build a Semantic-Aware LOM ontology. The ontology is capable of addressing the relationship and evaluating the semantic distance between learning objects, and of inferring a user’s intension and disambiguating the query by evaluating the weight of matched instances.
Ø Expand Queries. Recommending high relevant objects can increase both the recall rate and the retrieval precision.
Ø Map learning object metadata instances into ontology (indexing documents). A LOM description will be distributed into the ontology as the instances of concepts. The mapping is automatically completed by the system, and the author of an LOM description need not understand the structure of the LOM ontology. Indexing the LOM description into the ontology is the key step for instantiating the relations between learning objects.
We have been using Protégé 2
Figure 1. Semantic-aware ontological approach for learning object retrieval
Conclusions and Future Work
In this article we propose a semantic-aware, ontological approach for learning objects retrieval. By applying the ontology technology to the e-learning environment, e-learning systems can be more intelligent, powerful, and adaptive. This ongoing project will also integrate a well-defined learner’s feedback mechanism to score learning objects and help the system to recommend high quality learning objects.
References
[1] IEEE LTSC Drafts & Documents, http://ltsc.ieee.org/doc/index.html
[2] IEEE LTSC (IEEE Learning Technology Standards Committee), http://ltsc.ieee.org
[3] SCORM (Sharable Courseware Object Reference Model), http://www.adlnet.org
[4] Gruber, T. R., “Toward principles for the design
of ontologies used for knowledge sharing”,
IJHCS, 1994, no. 43(5/6), pp. 907-928.
[5] Protégé 2
[6] Web Ontology Language (OWL), http://www.w3.org/2
|
Ming-Che Lee Department of Engineering Science, Department of Engineering Science, |