Content uploaded by Laura Alba-Juez
Author content
All content in this area was uploaded by Laura Alba-Juez on Jun 04, 2014
Content may be subject to copyright.
Content uploaded by Laura Alba-Juez
Author content
All content in this area was uploaded by Laura Alba-Juez on May 22, 2014
Content may be subject to copyright.
e- Oral: A Hypermedia Web Application for English Language Oral
Exams within a Distance Education Environment
Mª Eugenia López, Laura Alba-Juez, Mónica Aragonés & Beatriz Molina
Universidad Nacional de Educación a Distancia (UNED), Madrid, SPAIN.
melopez@csi.uned.es, lalba@flog.uned.es, maragones@flog.uned.es,
beatriz_molina_garcia@hotmail.com
Abstract
The present paper constitutes an attempt to
describe, in a succinct manner, the objectives,
methodology and implementation of the e-oral
web application within the framework of the
CODE (Computerized Oral Distance Exam)
Project, carried out in the National Distance
Education University (UNED) in Madrid, Spain.
e-oral is initially a web application designed to
carry out distance oral exams in the first
English language course (Lengua Inglesa I) of
the new English Studies program at the UNED.
1. Introduction
The present paper constitutes an attempt to
describe, in a succinct manner, the objectives,
methodology and implementation of the e-oral
web application within the framework of the
CODE (Computerized Oral Distance Exam)
Project, carried out in the National Distance
Education University (UNED) in Madrid, Spain.
Starting with the 2002-2003 academic year, oral
exams were introduced as a mandatory requisite
within the Department of Foreign Languages (in
the English Studies Program) in four subjects,
namely Lengua Inglesa I, Lengua Inglesa II,
Comentario Lingüístico de Textos, and Fonética
Inglesa. Given the characteristics of the UNED,
a distance education university with students in
different parts of the world, the oral
examination process presented itself as a task
which was difficult to materialize, with different
obstacles emerging, such as the impossibility for
students who are in some of the UNED Centros
Asociados (different “branches” of the UNED
around Spain and the world) of taking such
exam, due to the fact that not all of these
Centros have tutors in charge of the subjects in
question.
Because of the above-mentioned problem,
the decision was made to set about working on
the CODE project, whose first stage is already
in an advanced phase of development, in a joint
effort between the Department of Foreign
Languages and the New Technologies Division
of the UNED. During this first stage, all the
efforts have been focused on the
instrumentation of e-oral in Lengua Inglesa I,
the English language course taken by the
students in the first year of the English Studies
program, where we have an average of 900
students enrolled each year. This first attempt is
intended to be taken as an example for other
subjects with a mandatory oral exam, as well as
for any other subject of the English Studies
program, or of the other language programs at
the UNED that could need such a tool.
Previous computerized oral language exams
exist in the market, such as the COPI
(Computerized Oral Proficiency Instrument)
from the Center for Applied Linguistics in
Washington D.C. (Malabonga 1998 [1], 2000
[2]; Malabonga & Kenyon 1999) [3], an
enterprise in which the researcher responsible
for this project participated during 1998-2000.
However, to our knowledge, neither COPI nor
any of other similar projects have the
technological characteristics of CODE’s e-oral,
which we describe in more detail in the sections
that follow. To be specific, in the first stage, we
are working to create a computer program by
means of which students will be able to carry
out oral activities and oral tests online,
equivalent to the traditional face-to-face oral
exams which test the knowledge and language
skills taught in the Lengua Inglesa I course, in
accordance with the course materials (Aragonés
& Medrano 2003 [4], Alba-Juez & Chacón
2003) [5]. This working system will allow
students to carry out self-evaluation oral
activities and final oral exams from their Centro
Asociado or from their personal computer,
through the software that will be published in
CD-ROM and which will also be possible to
download at the virtual course site. Tutors and
Professors will subsequently be able to access
the program from any UNED computer (both in
the Centros Asociados and in the Headquarters
(Sede Central) in Madrid) or from their own, in
order to listen to each student’s exam and grade
it.
The linguistic level required from students in
the final exam found in the web application is
that corresponding to level B2 (“Independent
user”), according to the parameters established
by the European Council in its European
Language Portfolio, which have become the
reference for the new European language
studies of the 21
st
century. Therefore, all the
learning material, from the textbooks and CD-
ROMS to the partial or final exams, has been
created with that level in mind.
2. Hypermedia as a teaching/learning
transformation tool
At the present time, the use of hypermedia in
distance education methodology is already
consolidated within the framework of
collaborative learning. The use of multimedia
material has spread within university studies,
which have profited from technological
advances in Internet communications through
the generalized use of learning-training tools.
To speak about the risks or drawbacks of this
learning method is no longer relevant, nor does
it stir up controversy as it used to during the
first years of this century. It is no longer
important or appropriate to debate whether
distance education is a method with deficiencies
as opposed to the face-to-face system of classes
in traditional universities. The experience
gained after several decades by the UNED and,
in particular, the last five year’s trajectory of the
Spanish university, prove that the face-to-face
and the distance systems can co-exist separately,
but can also complement each other, as occurs
in the blended learning system (López 2003 [6],
Anderson y Garrison 2005 [7]), consolidated in
the majority of European universities today.
In our country, great advances have been
made with the implementation of computer-
mediated communication in the postgraduate
and continuous education programs of many of
our universities (García 2003 [8]). From those
initial experiences onwards, computer-mediated
distance learning has been gradually gaining
terrain into the official programs of traditional
universities, and now they compete in means
and methods with the UNED, which has been
Spain’s one and only public distance education
university since the 1970s.
Multimedia technology is a powerful
resource for distance education if we consider
that, up to a very recent time, distance education
had strong limitations concerning the interaction
between teacher and student, especially within
language subjects, where the oral exchange
between teacher and student is fundamental.
In spite of the advance of computer-mediated
means, certain deficiencies still persist in the
teaching of subjects in which a high rate of
actual physical presence of both students and
teachers has been traditionally necessary, as is
the case with the teaching of languages.
Nevertheless, the viability of these cases within
the new virtual environment lies precisely in
accepting those limitations and establishing the
limits of each technological resource, because
only by acknowledging the current limitations
shall we be able to make improvements that will
eventually consolidate the virtual language class
as an authentic teaching/learning scenario.
It is already a well-known fact that distance
education has come to a point where it shares
certain characteristics with traditional, face-to-
face teaching/learning situations, and, at the
same time, traditional education has been able
to broaden its scope by integrating the
computer-mediated distance education model
(López 2003 [6]). As a result, a meeting point
arises between both teaching disciplines, giving
way to the current blended system, in which the
concept of traditional classes has also changed.
Simultaneous interaction in a video-conference,
a multimedia service that allows several users to
maintain a real-time distance conversation with
visual, audio and verbal interaction, is now a
normal practice within both distance and
traditional learning environments
.
3. The creation of multimedia
instruction content
The idea of multimedia learning implies the
conjunction of both didactic and instructional
criteria in the creation of course materials.
The architecture of hypermedia implicitly
carries along the integration of contents in
different formats: text, images (drawings,
photographs), audio (voice, music, sound
effects), moving images (animation, video) or
videoconferences, with the aim of generating
educational products in which teaching is
organized into a web-like structure (Coorough
2005 [9]).
In order to produce educational multimedia,
regardless of its final medium, it is necessary to
previously define the educational objectives,
create an index of contents, and outline
adequate strategies (López, 2004a [10]) for the
specific teaching area and academic level to
which it will be destined.
In such a task, it is also necessary to take user
criteria into account, and thus trace paths within
the material that will favor a global view of the
program as well as a physical or mental diagram
with the description of its different modules.
Hypermedia navigation (linear, hierarchical or
free) must be guided by the principles
established for its educational purpose in each
particular case, considering that all its
components (menus, icons, buttons and
hypertext) must help to provide a
comprehensive navigation. The same occurs
with help systems, designed so that the student
can carry out tasks without spending too much
time in finding out the interactional logic behind
each system. As a result, it becomes necessary
to describe, in a synthetic but sufficient manner,
the procedures that must be followed in each
case. That way, the student will be able to
receive final feedback that will facilitate her
understanding of the performed task (López
2004b).
These series of educational criteria have been
applied to the design and development of the e-
oral environment, the multimedia software that
is described in the following section, which
stands as a necessary contribution, as well as an
advance, in the study of second and/or foreign
languages at the UNED.
4. Didactic and functional description
of the application
e-oral is initially a web application
designed to carry out distance oral exams in the
first English language course (Lengua Inglesa I)
of the new English Studies program at the
UNED. A multidisciplinary team specialized in
the areas of Applied Linguistics, Computational
Linguistics and Educational Software
Development has participated in the design and
development of the software.
The creation of such a tool implies the
elimination of the impossibility of carrying out
long-distance oral evaluations, one of the main
drawbacks that have affected the teaching of
foreign languages at the UNED. As stated
above, in its first phase or Version 1, the
application is oriented towards the Lengua
Inglesa I student-users, who will be able to take
the oral exams at the computer room of their
Centro Asociado. This will be possible either
within national territory or at any of the UNED
foreign centers, as long as these have the
necessary computer equipment (multimedia and
a broadband connection). Furthermore, students
may also use this application and take the oral
exams using their personal computer with an
internet connection, a possibility that grants
students a great autonomy, besides providing
them with an ample calendar to carry out these
tests. They will only be restricted by the number
of times allowed by the system for each
authorized user to conduct them.
The use of this application only requires
basic hardware: an integrated sound card (a
basic component of any CPU), input and output
audio devices (speakers and microphone), as
well as a broadband internet connection. In
terms of software requirements, the user only
needs to have the latest web browser or, if this
isn’t the case, one that has been updated with
the program plug-ins. In this case, the only
plug-in necessary is Flash Player —created and
distributed by Macromedia—, which allows the
reproduction of vector graphics, sound, program
codes, stream video and bidirectional audio, etc.
Flash Player is the program used to execute
Flash files, already present in about 98% of
personal computers. Since it is not associated
with any system or software requirements, the
use of this application does not carry any
additional costs for the student, nor does it
require anything else that could become an
obstacle for any user with an ordinary personal
computer (PC or Mac). In addition, the use of
the oral exam application is very intuitive, apart
from the fact that a help system is included, and
that internal navigation takes place in a
graphically-designed user-friendly learning
environment.
As can be seen, the design of the interface of
a multimedia application and the architecture of
information are two dependent and inseparable
planes. Both the categorization of information
and the usability of the application are factors
that determine whether an interface is intuitive
and friendly or not. In the design and
organization of the interface, the interactivity
standards to which Internet users are used
(hyperlinks, action buttons, data registers,
password authentication, etc.), necessarily have
to be considered.
In the design of e-oral, standard navigation
conventions predominate, combined with a
certain degree of innovation in the physical
design and distribution of the elements within
the different evaluation scenarios (screens).
When non-intuitive elements are incorporated,
the student is provided with help support
containing the necessary information, following
the conversational maxims of quantity, quality,
relevance and manner appropriate for each case.
Even though e-oral is a multimedia system,
the first stage of the project (which involves the
development, testing and implementation of the
application for the 2006-2007 academic year)
only allows the oral exams to be carried out by
means of audio listening and recording. Later
on, in the second phase, video will be
implemented for both exercise presentations and
tests. The application is functionally prepared
for such types of evaluation, but it is still in the
process of developing the necessary audiovisual
repertoire (video script elaboration, recording,
editing and multimedia creation, etc.). As of
today, the available learning material is the
audio and text reference needed for carrying out
final exams (Figure 1).
Figure 1
Another of the objectives of the first phase is
to integrate e-oral into a Learning Management
System (LMS) which is compatible with the
technology used to develop the application. In
order to do that, proprietary (commercial
software) and open source solutions are being
analyzed, especially the learning platforms with
which the UNED currently works (WebCT y
Alf).
5. Technology used in the first stage
of the project
The oral exam application has been
developed using MySQL and PHP, and it
incorporates Macromedia Flash Server
Communicator 1.5 technology within the exam
recording tool.
e-oral runs on Windows XP
with an Apache server, and can also be
integrated within web and data services.
Because of its extensibility, it allows for the
administration of XML metadata, the access to
local system resources and the control of the
access and reports for a better and greater
integration within existing data management
systems.
PHP (PHP Hypertext Processor) is one of
most widespread server languages, which has
the advantage of being able to be used in
different platforms (Unix, Linux, Windows,
Mac). It is a versatile server-side language (all
the actions it conducts have their effect on the
server’s side) for the creation of multimedia
projects. Besides, it is a very advisable code for
developing applications and dynamic websites.
The integration of ActionScript (Macromedia
language) and PHP is an option that provides
great advantages in the connection of
Macromedia Flash MX interfaces with MySQL
databases (De la Cruz & Zumbado 2005 [12]).
PHP and MySQL are very much related and
many PHP extensions to support MySQL exist.
In addition, PHP is able to compile its code in a
multitude of platforms and operational systems,
something highly useful for programming web
applications. Regarding MySQL, it must be
remarked that it consists of a very fast multi-
user relational database server (Figure 2), robust
and adequate for the use of web applications
and PHP. Both MySQL and PHP are part of the
open source software category, which implies
lower costs and higher possibilities of
development to meet the demand.
Figure 2
Macromedia Flash Communication Server
MX is a type of software that allows
communication to take place, combining the
reproduction of audio and video with moving
Flash graphics in player customizable
environments.
Flash Communication Server is a server that
supports real-time applications. It allows for the
use of audio and video multi-channels, as well
as for data retrieval and remote functions. The
application automatically recognizes installed
standard USB microphones and webcams, and it
also authorizes the user’s access (Lesser
2005
[13]). This platform operates in different
environments such as the Internet, PDA’s or
interactive TV, allowing for communication
between Flash applications that can alternate
text, audio or streaming video, which favors
access to audiovisual data even if the user does
not have a broadband internet connection.
The architecture of
e-oral is, broadly
speaking, that of a standard web application,
distributed throughout three levels or layers.
The application gathers the user’s data (first
level navigator) and sends it to the server that
executes the oral exam program, linked to a
database and to documents marked with
metadata (second and third levels). On the third
level, the data that manages the data base and
XML document application can be found.
Finally, the system returns to the user the result
of his interaction by means of the navigator
(client). The presentation layer (first level)
receives the data and formats them so that they
can be displayed adequately (Figure 3).
Figure 3
6. Technology to be used for the
execution of the second phase of e-
oral
The second stage of the CODE e-oral project
(2007-08) will focus on the generation of
natural language processing computer tools that
will be incorporated in the current application
(Moreno 1998 [14], Rodríguez, Martí y
Castellón 1999 [15]). In particular, our research
will focus on technologies having to do with
speech and information extraction. The aim of
this second phase is the research and
development of the necessary technology to
implement a voice recognition and verification
system (Rodríguez 1997 [16], Furui 2001 [17])
by means of speech patterns that will allow for
student identification when distance oral exams
are carried out.
The speech recognition system will focus on
the use of isolated words. The treatment of
continuous speech (Martí y Llisterri 2004 [18])
is not, in principle, a necessity for the type of
software required by e-oral. Consequently, a
previously defined and restricted vocabulary
will be used for the recognition software
(Berstein y Franco 1996 [19]). The system will
not be limited to a restricted number of users; it
will be independent of the speaker. Thus, a
multi-user voice recognition and verification
software will be generated. In that manner, the
tool will be available for all students registered
within the authentication system provided by
e-
oral.
Another objective for this stage is to develop
data extraction software for the stored and
labeled tests in the application. These systems
are designed and constructed specifically to
carry out a certain task, and, as a result, a
different system will be employed depending on
the type of information that we wish to extract.
In this case, the information has to do with test
results. Having that in mind, the templates
within the necessary fields will be defined, in
order to gather information uniformly. The basic
criteria that will be applied are 1) precision,
which is equivalent to the quality of the
extracted information, and 2) coverage, which
refers to the real amount of extracted
information, regardless of previous predictions
(Gonzalo y Verdejo 2003 [20]).
The goal of this system is to detect, extract
and present information in a format that can be
processed automatically later on. The
architecture for the design of the information-
extraction tool will be standard, consisting of
the two phases typical of these systems: 1)
analysis of instructions and 2) extraction of
information. Three modules will be included in
each phase, as shown in figure 4
1
. In the
filtering module, the filtering of text takes place,
which consists in identifying the instructions
and the words that characterize them. In the
lexical and syntactic analysis, as well as in the
extraction modules, sentence analysis of the
texts resulting from the tests will take place. The
two last modules are oriented towards the
processing of instructions, and they therefore
possess a more discursive nature.
The morphological and lexical analysis
component is the one that carries out the
syntactic labeling and that also assigns semantic
features to some open category words (type of
user, exam model, date, question number). This
module includes the recognition of proper
names, identification documents, grades, etc.
The syntactic analysis is partial; basic
components are singled out and identified, such
as noun phrases, verb phrases or relational
elements (prepositions, conjunctions and
pronouns).
Figure 4
In the extraction module, relevant elements that
have previously been isolated will be identified.
In the following modules that make up this
extraction phase (in which results will be
1
Modified from Gonzalo & Verdejo, 2003: 186.
combined), the co-reference of terms will be
solved, establishing relationships between the
vocabulary. The last procedure consists in the
recombination of the partial descriptions
obtained in the extraction process (Gonzalo &
Verdejo 2003 [20]). The field descriptors that
will make up the templates will be filled.
As can be seen, these systems need to
apply complex natural language processing
(NLP) techniques due to the great amount of
precision that is required in the processes of
extraction and detection of the relevant type of
information (Cinchor 1998 [21]). In this sense,
the final goal of the project is to develop solid
and comprehensive software that will eliminate
the obstacles resulting from the absence of such
a tool in the field of oral evaluation within
language studies. At the same time, a linked
information management system will be
created, that we expect to be of great use for
future teaching and investigation in language
learning.
In conclusion, we can say that the
implementation of the previously-described
application will soon put an end to the difficult
task of conducting oral language exams within a
distance education environment. The possibility
of carrying out oral exams by means of a web
application with the characteristics described
herein presents interesting innovations that shed
light on a new dimension of communication,
unknown in the language teaching field to date.
References
[1] Malabonga, V. (1998). “Computers and
language testing: The Computerized Oral
Proficiency Interview”. Language Testing
Update, Issue 24, p. 29.
[2] Malabonga, V. (2000). “Trends in foreign
language assessment: The Computerized Oral
Proficiency Instrument”. NCLRC Newsletter
(http://www.cal.org/nclrc
).
[3] Malabonga, V. & D. Kenyon (1999).
“Multimedia computer technology and
performance- based language testing: A
demonstration of the computerized oral
proficiency instrument”. In M. B. Olsen (ed.),
Computer Mediated Language Assessment and
Evaluation in Natural Language Processing.
New Brunswick, NJ: Association for
Computational Linguistics.
[4] Aragonés, M. & I. Medrano (2003). Lengua
Inglesa I. Ejercicios y CD Roms. Madrid:
UNED.
[5] Alba-Juez, L. & R. Chacón Beltrán (2003).
A Grammar Companion to Lengua Inglesa I.
Madrid: UNED.
[6] López, E. (2003). “Producción de contenidos
para la Red y la tele-formación”, II Congreso de
Educación en Internet e Internet en la
Educación. Madrid, CNICE (MEC).
[7] Anderson, T. & G. R. Garrison (2005). El e-
learning en el siglo XXI: Investigación y
práctica. Madrid: Ediciones Octaedro.
[8] García, J. (2003). El e-learning en España:
modelos actuales y tendencias de actualización.
Madrid: Fundación Escuela de Organización
Industrial (EOI).
[9] Coorough, C. et al (2005). Multimedia para
la web (diseño y creatividad). Madrid: Anaya
multimedia-Anaya interactiva.
[10] López, E. (2004a). “El modelo docente
telemático”. Educación e Internet, Monográfico
87. Educaweb, ISSN: 1578-5793:
http://www.uned.es/euva/euvaflash/index.html
[11] López, E. (2004b). “Diseño de materiales
multimedia para cursos virtuales”. IV
Encuentro Internacional sobre educación,
capacitación profesional y tecnologías de la
información. Actas del Encuentro Internacional
VIRTUAL EDUCA,
http://www.virtualeduca.org/2004/es/vebarcelon
a.htm
[12] De La Cruz Heras, D. y C. Zumbado
(2005). Flash, PHP y MySQL: Contenidos
dinámicos. Madrid: Anaya multimedia-Anaya
interactiva.
[13] Lesser, B. et al
(2005). Programming Flash
Communication Server. Farnham: O'Reilly &
Associates.
[14] Moreno, A. (1998). Lingüística
computacional: introducción a los modelos
simbólicos, estadísticos y biológicos. Madrid:
Síntesis.
[15] Rodríguez, H., Martí, M. A. & I. Castellón
(1999). Formalismes lògics per al tractament
del llenguatge natural. Barcelona, UPC.
[16] Rodríguez M. A. (1997). “Introducción a la
conversión texto-voz”. Philologia Hispalensis,
11, Sevilla, 177-192.
[17]Furui, S. (2001). Digital speech processing,
synthesis and recognition. New York: Marcel
Dekker.
[18] Martí, M. A. & J. Llisterri (eds.) (2004).
Tecnologías del texto y del habla. Barcelona:
Universitat de Barcelona.
[19] Berstein, J. & H. Franco (1996). “Speech
recognition by computer”. En N. J. Lass (ed.),
Principles of Experimental Phonetics. St. Louis:
Mosby, 408-434.
[20] Gonzalo, J. & F. Verdejo (2003).
“Recuperación y extracción de información”. In
Martí, M. A. (coord.), Tecnologías del lenguaje.
Barcelona: UOC, 157-192.
http://congresos.cnice.mec.es/ceiie2/
[21] Cinchor, N. (1998). “Overview of
MUC/MET-2”. En Seventh Message
Understanding Conference. Washington D.C.,
USA.
