AN INTERFACE ENVIRONMENT FOR LEARNING OBJECT

SEARCH AND PRE-VISUALISATION

Laura Sánchez García, Rodrigo Octávio de Oliveira Mello

Universidade Federal do Paraná, UFPR University, Av. Francisco H. dos Santos, Curitiba, Brazil

Alexandre Ibrahim Direne, Marcos Sfair Sunye

Universidade Federal do Paraná, UFPR University, Av. Francisco H. dos Santos, Curitiba, Brazil

Keywords: Learning objects, LOR, LCMS, search, pre-visualization, HCI, semiotic engineering.

Abstract: Learning Objects – LOs – were devised in order to cut down on production costs and time, as well as to

facilitate the distribution and reuse of didactic contents by means of a series of functions, such as

reutilization, traceability, interoperability, durability and easy editing. Our main objective in the present

paper is to propose an interface environment for LO search and pre-visualization. The distinctive feature of

such environment is both the easy access to all search refinement functions available, thus allowing users to

fulfil their search objectives without requiring much cognitive effort, and the well-structured LO pre-

visualization. We started the project by defining a set of criteria to assess LO research and solutions, and

one of such works was used as our starting point for the interface environment. Our theoretical foundation

relies on HCI (Human-Computer Interaction), and the major outcome of the present project is a

nonfunctional prototype (in storyboard) of the interface environment proposed, which in turn solves the

majority of the problems we came across when revising the pertinent literature.

1 INTRODUCTION

The concept of LO (Learning Object) was devised in

order to grant digital didactic units the following

properties: reutilization, traceability,

interoperability, durability and easy editing.

Our main objective in the present paper is to

propose an interface environment for LO search and

visualisation whose main purpose is (i) to allow

users to take advantage of the entire search

refinement potential available with the minimum

cognitive effort, and (ii) to provide strong support in

object pre-visualisation.

The scope of the present work does not

comprehend an operational prototype, the definition

of working directives or the necessary architecture

of the search engine of the search interface

environment proposed. Traceability is reached

through two basic factors, i.e. a comprehensive

metadata standard and a search tool equally

comprehensive.

Great effort has been put into defining a

metadata index pattern. Nevertheless, studies on LO

visualisation and localisation simply complement

studies on metadata indexing, or rely on empirical

research.

2 METHODOLOGY

The first methodological step we took consisted of a

revision of the relevant HCI (Human-Computer

Interaction) literature, as well as literature

concerning the production of digital didactic content,

ontologies and teaching processes.

Once the literature had been revised, we carried

out an in-depth analysis of works both on LO

searching and pre-visualisation and on other

interface solutions that might contribute to the

development of the environment we propose. We

used works concerning search tools in various

multimedia collections, search interfaces by direct

manipulation, by result filtering, among others.

Based upon this analysis we devised a set of

assessment criteria as the first step towards the

solution of the problem proposed. Each of the

240

Sánchez García L., Octávio de Oliveira Mello R., Ibrahim Direne A. and Sfair Sunye M. (2008).

AN INTERFACE ENVIRONMENT FOR LEARNING OBJECT SEARCH AND PRE-VISUALISATION.

In Proceedings of the Tenth International Conference on Enterprise Information Systems - HCI, pages 240-247

DOI: 10.5220/0001701702400247

 SciTePress

solutions we studied was confronted with this set of

criteria so as to determine which one would be the

most appropriate and effective to work as the

foundation for our solution.

Once all solutions had been analysed – taken

in association with the literature and in the light of

current HCI premises – we came up with a new

solution for LO search and pre-visualisation.

Finally, we also devised Corel Draw® 12 and

Adobe Photoshop® CS static images simulating

different uses of the environment (storyboards) in

order to complement the description of the solution

proposed.

3 CONTEXT

Digital contents are one of the outcomes of

technological evolution, particularly of the Internet.

At first there were Content Management Systems

(CMSs), and then as Instructional Design evolved

and as the concept of learning object was introduced,

the very first systems oriented towards the

development and management of digital didactic

objects came out, as well as other systems oriented

towards learning management.

The main purpose of LOs is to contribute to

didactic content access and sharing. LO

conceptualisation has caused heated debate in the

scientific community, and it has many different

definitions.

(WILEY), with his pragmatic approach, defines

LOs in a pragmatic way as “any digital resource that

may be reused in order to support learning”.

An LO repository (LOR) is a collection of

learning objects (or of metadata describing leaning

objects) managed through a technology which in

turn allows users to request, find, post and submit

these objects to the network (JONES). Normally one

LOR is part of a Learning Content Management

System (LCMS).

One LCMS is a system that supports generating,

assembling, storing and posting tailor-made didactic

contents. Its main purpose is to cut down on

development time through its special tools, and to

make LO reuse easier. An LCMS allows students to

get the required content only, in a tailor-made way

and in the appropriate amount and time.

ALOCoM (Verbert et al 2004), i.e. Abstract

Learning Object Content Model, is an ontology that

defines LO through its components. The ontology

sorts out such components in three categories, as

follows: Content Fragments (CF), Content Objects

(CO) and Learning Object (LO). CFs are basic

elements, such as videos, sounds, texts, among

others, which in turn cannot be subdivided. COs are

the combination of two or more CFs, also including

a navigation path. It is the navigation elements that

are responsible for structuring CFs in COs. Unlike

CFs, COs may contain other COs in their structure

and thus may be of abstract constitution. Finally,

LOs add a didactic purpose to one or more COs.

Figure 1 portrays the ALOCoM model.

Figure 1: ALOCoM (Verbert et al 2004).

4 RELATED WORKS

So far, there have not been many efforts towards the

improvement of the quality of LO search and pre-

visualisation interfaces, which leads to rather

complex LOR and LCMS search interfaces. The

works presented in the following paragraphs have

contributed greatly to this debate, helping us to come

up with the solution presented here.

In this article (Klerkx et al 2004), they defend the

use of Information Visualisation techniques, which

basically refer to computer-supported visual

representations of abstract data that cut down on

users’ cognitive efforts (CARD et al 1999), aiming

at turning LO search into a more flexible, interactive

process.

The following article (Klerkx et al 2005)

complements the studies carried out in (Klerkx et al

2004) and brings forth a rather flexible environment

for interface development using Information

Visualisation techniques, particularly tree diagram

representation. Through this environment the

authors elaborated study cases with Ariadne LOR, as

well as with the EdMedia LOR (see Figure 2),

aiming at making LO search and localisation in base

repositories easier.

AN INTERFACE ENVIRONMENT FOR LEARNING OBJECT SEARCH AND PRE-VISUALISATION

241

Figure 2: Prototype interface (Klerkx et al 2005).

The following article (Klerkx et al 2006)

complements the abovementioned studies. The

authors present an interactive application for the

visualisation of large repositories with small LO

components (COs and CFs), which in turn were

obtained by means of the disaggregation of full

didactic contents. The study case they present refers

to the ALOCoM repository, taking this very same

ontology as its basis.

The main objective of the (Najjar 2005) research

was to improve the usability of the Ariadne LOR

search tool. Its authors came to the following

conclusions: a) Simple and efficient search functions

are essential for improving users’ motivation and

trust; b) Information organisation and structure must

be oriented towards users’ needs; c) Terminology

must be accessible for users; d) Help tools must be

improved so as to increase users’ participation; e)

The more refined search tools are, the better their

performance.

In (Yee et al 2003), the authors bring forth a

complete and creative solution for searching within

an image collection. Their approach allows users to

navigate through the conceptual dimensions that

describe an image making use of hierarchic metadata

and generating search samples dynamically. This

way, not only does the interface provide successive

search through clicks, avoiding thus complex syntax

search, but it also allows users to get to know the

structure of the image repository.

The (Wiza et al 2004) article presents a system

for 3D-visualisation of Internet search results

entitled Periscope. Three different interface

visualisation levels are possible, as follows:

synthetic interface, for displaying aggregate data;

analytic interface, for displaying details of

documents found; and hybrid interface, for

displaying both aggregate data and details of

documents.

5 OUR SOLUTION

After thoroughly analysing both the abovementioned

related works and the existing tools, we chose the

(Klerkx et al 2004) “tree diagram” as our starting

point for devising an integrated solution for the

problem of LO search and pre-visualisation.

Therefore, our solution follows the tree diagram

principle; nevertheless, its main distinctive feature is

the possibility of LO disaggregation into COs and

CFs based upon the ALOCoM ontology (Verbert et

al 2004). Other differences inherent to our solution

shall be presented below.

5.1 Interface Environment: Visual

Resources

Before describing the actual interface environment,

we would like to present some visual resources of

which we make use.

Colours: Colours play an essential role for our

solution’s communication coherence.

The colours

and are used in LO search or

classification. When the two blue symbols might

merge, the two shades of blue are juxtaposed. The

colours

, and stand for the LO rights.

refers to LOs whose use and visualisation are

unlimited;

stands for those whose use and

visualisation are limited (such as paid LOs); and

refers to those whose visualisation is unlimited, but

whose use is limited – such those under copyleft

(GNU) or creative commons (CC). The choice of the

colours

and , and follows (Klerkx et al

2004). The colour refers to LO listing, and since

its message is rather close to that of the colours

and

, we opted for because it is the colour

right next to blue in the light spectrum. The colour

stands for LO disaggregation, and we chose it

because it is a warm colour that contrasts with the

other colours. The colours

, and stand for

LOs, COs and CFs, respectively. We opted for these

colours because their message is close to the one of

the colour

; indeed, they are variants of for

their superior and inferior luminosity, representing

thus different completeness levels. Finally, the

colour

stands for unavailable symbols.

Icons: As we show in Figure 3, the interface

icons may be classified as action (ação) and

identification (identificação) icons.

ICEIS 2008 - International Conference on Enterprise Information Systems

242

Figure 3: Interface icons.

The action icons are colourful, and their colours

refer to the action they perform. Since stands for

disaggregation and

stands for listing, the icons

conform to the same pattern. The action icon

“Close/Remove” (

) conforms to the operational

system of the interface users.

The identification icons are not colourful and are

used in the disaggregation interface for making CF

identification easier. Each CF type has a different

iconic representation. Such icons may be displayed

alone or together with another icon that represents a

CF pre-visualisation in a smaller scale, in the case of

images, or a frame, in the case of videos. The

division between discrete and continuous CFs

respects the classification presented by (Verbert et al

2004). The icons for discrete CFs stand for static

objects, such as a text or a graph, among others. The

continuous ones stand for moving objects.

5.2 User Help

The main purpose of user help is to provide strategic

information so that the entire potential of the

environment can be explored. This sort of help is

particularly relevant for intellectual tools such as this

one. Such help consists of two complementary parts,

which provide insights on the application from two

different points of view.

Structural help, “What’s the potential of the

environment?”, describes the use of the application

from the point of view of the benefits it brings to the

actual activity of didactic content elaboration and

manipulation. Procedural help, “How...?”,

describes the procedure for achieving the objectives

of real life activity through the functionality

provided by the application.

5.3 Initial Search Interface

The initial interface display is a preview of the full

repository and follows the work of (Klerkx et al

2004), as shown in Figure 4. The blue portions stand

for LO areas and subareas of knowledge, whereas

the green, red and yellow portions stand for the

respective LO right limitation level. The numbers at

the top of the screen are the actual number of LOs

included in each one of them. The area taken up by

these portions is actually proportional to the number

of objects it comprises. By “clicking” one of the

blue, green, yellow or red portions, users are taken

to one level “above” or “below” the current level.

The initial display also includes a window for

information filter selection, as well as one for the

selection of the field which will be used as basis for

interface result mapping/display.

Figure 4: Initial interface display.

5.4 Viewing LOs in List

Listing was always an available option in (Klerkx et

al 2004). According to both (Klerkx et al 2004) and

(Yee et al 2003), list search is far more time

consuming than visual search. This is why in our

solution we include listing as an optional feature

which should be the last search option users resort to

(Yee et al 2003), and the lists should always have a

minimal number of options (Klerkx et al 2004).

Figure 5: LO listing within searched elements.

AN INTERFACE ENVIRONMENT FOR LEARNING OBJECT SEARCH AND PRE-VISUALISATION

243

If there are no other classifications, or if users

click

(see Figure 4), a list within the

quadrangular area is displayed (see Figure 5). In

(Klerkx et al 2004), the rectangular areas containing

a large number of LOs can get illegible. Therefore,

we opted for including the scrolling function in our

solution so as to avoid this problem. When

displaying LO listings, the frame of the rectangular

area is coloured

to represent the list.

5.5 Filter

When a form comprises many filtering fields it may

become voluminous and take up unnecessary

interface space. One solution for this problem is a

filter with a few entry possibilities, making it

possible to narrow or broaden the search. The filter

tool consists of three or four fields to be filled out, as

follows: search field, operation field, value field and

the Boolean operators AND and OR. By using the

filter one can establish conditions that, when applied

to the search, lead to immediate results.

Figure 6: Filter tool.

As shown in Figure 6-1, the first selection box is

responsible for indicating the field in which the

filtering shall take place. We have derived the first

few options from (Najjar 2005), and they correspond

to the most frequently used metadata (title – “titulo”

–, author – “autor” – and concept – “conceito”).

These options are entitled “Main” (“Principais”),

whereas the advanced search fields (language –

“linguagem” –, publication date – “data de

publicação” –, among others) are entitled “More

fields” (“Mais campos”).

The next step is then to select the operation

(“operação”) (“containing” – “contenha” –, or

ending in – “termine com” –, “the same as” – “seja

igual a”) that shall be applied to the search (as

shown in Figure 6-2). We derived these operations

from the “PHP My Admin®” system

(PHPMYADMIN).

Once these two first options have been selected,

users must then fill out the third field, i.e. the

information they wish to be filtered. The value of

this entry is contextualised within the field of the

metadata to be filtered and can be of limited entry or

textual. Figure 6-3 brings an example where the

entry is textual (field filled out with the “HCI” value

– “IHC” –, field “Title” – “Título”), while Figures 6-

4 and 6-5 bring examples where the entry is limited

through selection boxes (the fields “Language” -

“Linguagem” – and “Publication Date” - “Data de

Publicação” -, filled out with “Portuguese” –

“Português” – and “15/08/2007”, respectively).

Once these three first fields have been filled out

and users have clicked the INCLUDE (“Incluir”)

button, the first clause is then inserted in the filter,

and the search interface displays a new field, i.e.

AND/OR (“E/OU”), as shown in Figure 6-3.

Finally, this AND/OR field must be filled out.

By choosing the option “OR”, a new line with the

chosen clause is added to the filter preceded by a

line containing the word “OR” – “OU” – (see Figure

6-6). If the chosen option is “AND” – “E” –, this

clause is added right next to the last clause of the

filter, preceded by the word “AND” – “E” – (see

Figure 6-4).

This structure here described reproduces the

sentence building process of a natural language

(Portuguese), from right to left, originating

alternatives in columns. One such model allows for

clauses to be removed from the search filter. In order

to do so, users must “click”

right before the

clause in question. Once a clause is removed, it is

taken away from the filter interface and the search

interface is then affected by the remaining filter. We

derived these options of clause addition and removal

from (Yee et al 2003).

5.6 LO, CO and CF Search

We derived the idea of visualising disaggregated

LOs from (Klerkx et al 2006), whereas the concepts

concerning disaggregated LOs used in the present

solution come from the ALOCoM ontology (Verbert

et al 2004).

The icon

(present in Figures 4 and 5) stands

for the disaggregation of LOs. We have chosen the

explosion icon to represent this exploded view, a

metaphor we propose for the disaggregation. During

disaggregation, a new secondary tab coloured orange

is displayed and labelled either with the name of the

level or with the name of the LO in question.

ICEIS 2008 - International Conference on Enterprise Information Systems

244

When visualising disaggregated LOs there is a

filtering tool analogous to the one presented for LO

search – only now with metadata information on

COs and CFs.

Results may be grouped by LO, CO or CF. Each

result grouping has its own specific display, as we

shall describe in the following sections.

5.6.1 Disaggregated LOs grouped by LO

Figure 7: Disaggregated LO grouped by LO.

In the interface where disaggregated LOs are

grouped by LO are displayed the main metadata on

the LOs and COs it comprises. COs are displayed as

CF groupings ordered by inherent navigation and

distributed through broken lines which in turn

represent the boundaries of the CFs. CO ordering is

also done by navigation in their respective LOs. One

example of this is shown in Figure 7.

5.6.2 Disaggregated LOs grouped by CO

Figure 8: CO list grouped by type.

In CO grouping, on the other hand, there are only

COs and lists of CO types (see figure 8). Results are

grouped according to these types (Verbert et al

2004), which display, additionally, the number of

results in each of the groupings. CO display follows

the pattern shown in Figure 7, but it also has the

additional assistance of the filter component “Show”

(“Mostrar”), through which one can select only two

CO types so as to make CO localisation easier

within the repository.

5.6.3 Disaggregated LOs grouped by CF

Finally, in CF grouping (see Figure 9), besides the

listing of CF types and the number of elements

comprised by each of these sets, the pre-

visualisation of these fragments – together with their

titles and key words – is displayed as well.

Figure 9: Disaggregated LO grouped by CF.

The interface takes advantage of the

communicative potential of such fragments so as to

make the search process easier on users. For

instance, a picture is displayed by means of a

smaller-scale version of the original including an

informative icon explaining that that image is

actually a picture (rather than a video, for example).

At this point, neither the metadata nor the structure

of both LOs and COs are displayed. Similarly to CO

grouping, only the CF types chosen by users are

displayed.

5.7 LO, CO e CF Pre-visualisation

In the solution we propose, one can visualise both

LOs or COs and CFs the moment one interacts with

the icons and links displayed on the disaggregation

screen or on the LO listing screen. The screens

corresponding to LO, CO or CF pre-visualisation are

divided into two frames. The one on the left-hand

side comprises the metadata of the visualised

element (LO, CO or CF). Together with the

metadata is the information concerning element

reuse, indicating how many times it has been used in

the repository. By “clicking” this information, a list

AN INTERFACE ENVIRONMENT FOR LEARNING OBJECT SEARCH AND PRE-VISUALISATION

245

of the LOs that use this element is displayed. In the

right-hand frame is the pre-visualisation of the

element selected.

5.7.1 LO Pre-visualisation

Figure 10: Overview of LO content and metadata.

The LO pre-visualisation screens (see Figure 10)

become accessible when users “click” the title of an

LO (Figure 7), for instance. To the left are the LO

metadata and the number of times it has been used.

To the right is the full LO, containing all its

components (COs, CFs) ordered by the navigation

defined for such elements. Users may also

disaggregate this LO by clicking the

icon, which

will then lead them to a new tab.

5.7.2 CO Pre- visualisation

Figure 11: Overview of a CO and its metadata.

CO visualisation (see Figure 11) becomes available

when users “click” the rectangular area that delimits

the CO representation area (except for the area of the

icons representing CFs) in the screens shown in

Figures 7 and 9. In the left-hand area are the CO

metadata and their reuse. In the right-hand area are

the metadata on each of the CFs that make up the

CO. There is also where the pre-visualisation of the

CFs that make up the CO takes place.

5.7.3 CF Pre- visualisation

Figure 12: Overview of a CF and its metadata.

Similarly to the other cases, on the left-hand side are

the metadata indexing the fragment, whereas on the

right-hand side is the pre-visualisation of the

fragment. Users may access such CF pre-

visualisation by “clicking” one of the icons that

represent CFs (see Figures 7 and 9). If the fragment

is an image, graph or photo, the fragment is

displayed together with an icon representing the type

of fragment in question. If it is a text or link, the

very content of the text or link is displayed. If it is an

animation, video, sound or song, the interface

displays the necessary commands to reproduce or

execute them.

5.8 Pre-assessment

Due to the non-operational nature of the tool we

proposed, and aiming at testing this same tool by

getting potential users to work with it, we elaborated

an experiment based on the following basis.

As a first task, we will ask students to read the

technical description of the environment proposed.

After that, we will ask them to carry out two tasks

by using the technical description they will have just

read.

The first of such tasks involve the entire

potentiality of the tool, but specified in such way as

to have the students carry out the search steps in a

different order from the one presented in the

technical description of the application.

The second task required great skill and

comprehensive knowledge of the main potentiality

of the tool. We intend to ask students to searching

for all materials available on a certain topic as if

there were going to prepare a new course on that

ICEIS 2008 - International Conference on Enterprise Information Systems

246

topic. Nevertheless, we did not specify the necessary

steps, making it a rather generic, free activity.

Both tasks involve hypothetical data, different

from those used in the examples presented in the

technical description of the application.

We will also ask the participants to take note of

all and every difficulty they come across while

carrying out the tasks, as well as of the sequence of

operations performed and the expected solutions

within the application.

When analysing the results, we will take into

consideration whether both tasks have been

successfully fulfilled or not. For the second, generic

task, we will also take into consideration the use of

the entire potentiality of the tool.

We will now apply this experiment to a sample

of 7 Masters Students (potential users of this kind of

application) as a test to be applied to a larger

universe in a near future. Our hypothesis are that the

tool is self-explanatory when it comes to the

accomplishment of each individual task. However,

its potentiality may be underexploited, because of

the lack of strategic help, which in turn should

describe - in a more precise way - the distinctive

aspects of the environment when used in the real

world, i.e. in the search for materials oriented

towards the elaboration of didactic content.

6 CONCLUSIONS AND FUTURE

WORKS

In the present work we listed and looked into a

number of solutions available for building a proposal

of an environment (on storyboard) based upon

current HCI premises and concepts. The results we

achieved solve most of the problems we came across

whilst revising the literature on this sort of

environment. The main advances and achievements

of the environment we propose are the following:

the pre-visualisation of the semantic structure of the

content by taking advantage of the communicative

potential of LO content and its components; the

possibility of synthetic visualisation (at LO level)

and analytical visualisation (at CO and CF levels) of

search results; and the visualisation of results

according to different attributes/mappings.

When giving continuity to the present work, we

will applied an experiment to a sample of 7 Masters

Students as a first test. Our hypothesis are that the

tool is self-explanatory when it comes to the

accomplishment of each individual task. However,

its potentiality may be underexploited, because of

the lack of strategic help.

After the concept of the environment an the

experiment, the next step is to apply the

improvements needed to the conceptual environment

and apply again the pre-assessment to a larger

universe of potential users.

Other pertinent future works may approach the

incorporation of new search capacities to the

information filtering tool (Cardoso 2000); the

improvement of result ordering, such as through the

ranking of the results displayed (Ochoa et al 2006);

the help during the process of cooperative search

among repository users, including cooperative and

dynamic indexing as well as the possibility of taking

notes about the use of reused objects as inputs to the

new decisions and their possible reutilisation; among

others.

REFERENCES

Card, S., Mackinlay, J. D., Shneiderman, B., 1999.

Readings in Information Visualization, using vision to

think, Morgan Kaufmann Publishers, Inc.

Cardoso, J.C., 2000. iLIB, Uma Proposta de Interface de

Consulta Personalizável Para Bibliotecas Digitais.

Klerkx, Joris et al, 2004. Using Information Visualization

for Accessing Learning Object Repositories, Computer

Science Department, K. U. Leuven.

Klerkx, Joris et al, 2005. An Information Visualization

Framework for Accessing Learning Objects

Repositories, Dept. Computerwetenschappen,

Katholieke Universiteit Leuven.

Klerkx, Joris et al. Visualizing Reuse: More than Meets

the Eye, 2006. Katholieke Universiteit Leuven,

Belgium.

Najjar, Jehad et al, 2005. Finding Appropriate Learning

Objects: An Empirical Evaluation, Computer Science

Department, K. U. Leuven.

Ochoa, Xavier et al., 2006. Use of Contextualized

Attention Metadata for Ranking and Recommending

Learning Objects, ACM.

Verbert, K et al., 2004. Toward a Global Architecture for

Learning Objects: A Comparative Analysis of

Learning Object Content Models.

Wiza, Wojciech et al., 2004. Periscope – A System dor

Adaptative 3D Visualization of Search Results, ACM.

Yee, Ka-Ping; Swearingen, Kirsten; Li, Kevin; Hearst,

Marti, 2003. Faceted Metadata for Image Search and

Browsing.

CC, Creative Commons. http://creativecommons.org/

GNU, Copyleft. http://www.gnu.org/copyleft/

PHPMYADMIN. http://www.phpmyadmin.org/

AN INTERFACE ENVIRONMENT FOR LEARNING OBJECT SEARCH AND PRE-VISUALISATION

247