ACCESSIBILITY AND USABILITY

OF A COLLABORATIVE E-LEARNING APPLICATION

Andrea Bozza

, Marco Mesiti

, Stefano Valtolina

, Silvia Dini

and Marina Ribaudo

DiCo - University of Milano, Italy

Istituto David Chiossone, Italy

DISI - University of Genova, Italy

Keywords:

Wiki system, Design for all, Accessibility and usability evaluation.

Abstract:

VisualPedia is a collaborative environment proposed to facilitate the development of educational objects

thought for all students, including students with different forms of disability. In this paper we brieﬂy in-

troduce VisualPedia and then report on the experience in the evaluation of accessibility and usability of the

system prototype we have developed so far. We also discuss possible future improvements that have become

evident after the experimentation with end users.

1 INTRODUCTION

The knowledge life-cycle has radically changed with

the advent of Web 2.0 technologies: the user in fact

has become active (prosumer) in the process of build-

ing knowledge. The network is no longer a mere

tool for content distribution but it is rather a facil-

itator for the interaction among the participants in-

volved in the educational process, and the literature

clearly states that collaborative learning is superior

to individual learning. Wim Veen, for instance, de-

scribes adolescents (Veen, 2009) as the generation of

“Homo Zappiens (who) lives in human and technical

networks that provide new opportunities to act exper-

iment and learn. Homo Zappiens learns within social

networks where content is distributed and discontinu-

ous. As a consequence, future learning will be based

on concepts of aggregation, externalization, collec-

tive knowledge creation and immersion . . . ”

Active participation is one of the main ingredient

of our software prototype, named VisualPedia. The

second important ingredient is personalization since

we think that the delivery of content tailored to the

user is particularly relevant in the school educational

context where students with different needs coexist.

Personalization in this context does not mean that

the system automatically selects different learning

paths based on the learner’s proﬁle. This is typical

of Adaptive Hypermedia Systems which are gener-

ally more complex that our software prototype; for a

brief survey on different views of personalization see

for instance (Verpoorten et al., 2009). In VisualPe-

dia personalization means that the same content can

be delivered in different format depending on users’

visualization preferences. The prototype, in fact, is

a wiki-based application developed to offer an initial

response to some school needs towards the standard-

ization of e-learning content for students with differ-

ent forms of disability (visual, ranging from blind to

high levels of visual disorders, cognitive and verbal).

VisualPedia stakeholders are of two different

types: school teachers and students. Teachers from

different classes and, possibly, from different schools,

are encouraged to collaboratively produce learning

material that students can access in different formats,

depending on their proﬁles. The model is still tradi-

tional in some sense: teachers are the producers of the

educational content. However, they can perform this

task in collaboration with other colleagues (i.e. with

peers), and no longer in isolation.

The aim of this work is the evaluation of the ac-

cessibility and usability of the developed system. Key

point of the paper is the comparison of the evaluation

of usability performed by heuristic experts and end-

users with disabilities. We remark that this is particu-

larly important because the end-user perspectives can

be quite different from those of experts since they use

assistive tools to access the system. The evaluation

opens up discussions and proposals of enhancements

along several directions that will be discussed.

The paper is organized as follows. Section 2

brieﬂy introduces related work and Section 3 de-

scribes the main characteristics of VisualPedia. Sec-

tions 4 and 5 discuss the experience on the evaluation

102

Bozza A., Mesiti M., Valtolina S., Dini S. and Ribaudo M. (2010).

ACCESSIBILITY AND USABILITY OF A COLLABORATIVE E-LEARNING APPLICATION.

In Proceedings of the 2nd International Conference on Computer Supported Education, pages 102-109

DOI: 10.5220/0002798001020109

 SciTePress

of the accessibility and usability of such environment.

Future developments are reported in Section 6.

2 RELATED WORK

The literature on e-learning is huge, ranging from the

description of innovative learning management sys-

tems (proprietary or open source) to the proposal of

different pedagogical models, to the description of

case studies involving real users, and so on.

Here we concentrate on e-learning softwares de-

veloped having in mind Web 2.0 suggestions that call

for promoting online collaboration among teachers

and students, without any software barrier, in the pure

spirit of the design for all philosophy. Therefore, our

work takes advantage of experiences gained in the ar-

eas of accessibility and usability, software for people

with disabilities and wiki systems.

Accessibility. This is a big issue in the development

of online content and services. Quoting Tim Berners-

Lee: “The power of the Web is in its universality. Ac-

cess by everyone regardless of disability is an essen-

tial aspect” and several activities have been promoted

since 1997 by the W3C to deﬁne technical guidelines

to produce accessible digital resources.

Indeed, there are several ongoing discussions on

this topic. Some authors (Kelly et al., 2007) sug-

gest that the focus on technical guidelines can be

counter-productive and they elaborate on the term Ac-

cessibility 2.0 for the development of accessibile Web

sites and/or e-learning resources that address users

needs rather than compliance with technical guide-

lines. Moreover, since ﬁnished and perfect solutions

are generally not available, they suggest an incremen-

tal process with successive reﬁnement and develop-

ment involving a wider range of stakeholders (the

always-beta metaphor).

Accessibility experts are also suggesting to relax

some of the guidelines since it seems these are en-

couraging common mistakes instead of facilitating the

access. As an example, they claim the use of access

keys is contradictory since these can override key-

board shortcuts for screen readers (Moss, 2008).

Usability. Usability evaluation is one of the most

important cornerstones of the interaction systems de-

sign (Dix et al., 2003). Beyond its perceived useful-

ness, any software tool is really accepted by a large

community of users if it is also easy-to-use.

In order to assess and test a system several eval-

uation methods can be adopted to measure or pre-

dict how effective, efﬁcient and satisfying it would

be when the user interacts with it in her/his context of

use. These methods range from inspection techniques

to tests with end-users.

Results from some studies (Thompson and Kemp,

2009; Silva and Dix, 2007) have highlighted how

conventional usability evaluation methods like heuris-

tic evaluation (Nielsen, 1995) and even exploratory

methods like the cognitive walkthrough (Cathleen

et al., 1994), bare not appropriate to analyze Web

2.0 or social networking sites. These studies argue

that the adopted inspection methods do not reﬂect the

opinions of the users. For example in (Silva and Dix,

2007) the authors found that YouTube failed when

tested using heuristic evaluation although it is one

of the most popular Web applications. Moreover,

(Greenberg and Buxton, 2008) presents a thesis ac-

cording to which, in some cases, focusing on the us-

ability can be harmful.

This is because usability methods tend to put the

lens on the usability bugs and not on the whole useful-

ness of an application. Innovative ideas, such as those

at the base of the Web 2.0 on social networking sites,

could be discouraged by negative results and give up

on plans that might otherwise bear good fruits.

(Petrie and Kheir, 2007) presents the results of

a test with end-users (blind and sighted people) to

investigate the accessibility of two Web sites, dis-

cussing the issue of the relationship between acces-

sibility and usability. The authors show that usabil-

ity and accessibility problems do not belong to two

distinct, non-intersecting sets, but they can be seen

as two overlapping sets including “pure accessibility”

problems, “pure usability” problems, and “universal

usability” problems (Shneiderman, 2000) that affect

both disabled and not-disabled people.

Special Softwares. Several software tools for peo-

ple with disabilities have been proposed. As an exam-

ple, the software platform Edushare (Sauvain and Szi-

las, 2009) is a Web-based learning environment de-

signed for autistic children who are exposed to sim-

ple and repetitive tasks. Different users can access

Edushare covering ﬁve different roles. Autistic chil-

dren are the learners and they are helped by accompa-

niers during the execution of their learning sessions.

Developers create new exercises outside the platform,

the program director is responsible of the progress

of the children, and distant analysts evaluate the pro-

posed exercises, typically by analysing the log ﬁles

collected by the software platform.

The main difference between VisualPedia and

Edushare is that the latter is speciﬁcally built for autis-

tic children while we aim at a online virtual place

where digital resources should be inclusive of every-

one. Moreover, in the paper it is not explicit whether

developers can collaboratively build new exercises.

ACCESSIBILITY AND USABILITY OF A COLLABORATIVE E-LEARNING APPLICATION

103

Figure 1: XML representation of the Tower of Pisa.

Wiki Systems. Many wiki systems do ex-

ist, as witnessed by the long list displayed at

http://www.wikimatrix.org/ where different wikies

can be selected and compared. However, as discussed

in (Taras et al., 2008), not much research has been

devoted to the analysis of their accessibility. The

authors underline that accessibility is not an issue for

many wiki developers and suggest some improve-

ments that could be added to wiki systems to reﬁne

their level of accessibility. We mention here the

need of adequate presentations for different visual

disabilities since this is one of the issues considered

in VisualPedia, as discussed in next section.

3 VISUALPEDIA FACILITIES

The main goal of VisualPedia is that of offering a

repository of e-learning content, called educational

objects, that can be presented to users (students) ac-

cording to their proﬁles. Educational objects can be

described at different levels of complexity and images

can be delivered in their original format or in a sim-

pliﬁed version which is thought for users with visual

impairments.

The overall motivations, architecture and charac-

teristics of VisualPedia have been already described

in (Boccacci et al., 2009), and we brieﬂy present here

only its main features. VisualPedia is built on the top

of MediaWiki, which has been selected as our start-

ing point because of its popularity, facility of use,

and large community of developers. In the remain-

der of the section, we discuss how educational ob-

jects are handled, stored, and retrieved in the system.

Moreover, we describe the image simpliﬁcation pro-

cess and the modiﬁcation on the MediaWiki system

to make it accessible.

Educational Objects. Educational objects describe

notions represented through different media: text, im-

age, sound. Each object can be described at different

Figure 2: Simpliﬁcation of the Leaning Tower of Pisa.

levels of complexity: complete, summary and simpli-

ﬁed text. The last level preserves essential concepts

that are described with a simpliﬁed structure of the

sentences and a controlled vocabulary, and are de-

voted to students with cognitive disabilities.

Educational objects are uploaded into VisualPedia

thanks to an extended user interface: school teachers

can collaboratively upload new objects and describe

them at different levels of detail. Descriptions can be

initially incomplete and afterwards extended by col-

laborating peers.

Educational objects are internally stored in the

text attribute of the MediaWiki Page table as XML

documents with different tags, one for each descrip-

tion level plus the bibliography reference and optional

links to vocal descriptions. Figure 1 reports a frag-

ment of the XML document for the Tower of Pisa.

Images and sounds are stored in the MediaWiki

Image table. Tags are introduced in the XML doc-

ument to distinguish the role (an object or part of a

description) of the image/sound within the education-

al object. Moreover, in the Image table the values of

longDesc and alt attributes required for image ac-

cessibility are also included.

Image Simpliﬁcation. The images stored in Visu-

alPedia can be delivered in a simpliﬁed form trying

to mask many irrelevant details, which are discarded

without even noticing by normal user, while, on the

other hand, capturing such details wastes much effort

for visually impaired ones. Moreover, the use of many

bright colours intended to capture the attention, can be

perceived as confusing by disabled users, who would

most appreciate a line drawing with strong contrast-

ing background and no further details.

Therefore simpliﬁcation algorithms, based on the

Canny method (Canny, 1986), have been developed

so that any colour or grey scale image can be con-

verted, automatically and relying on the user proﬁle,

into another simpliﬁed one.

Figure 2 shows an image of the Leaning Tower

CSEDU 2010 - 2nd International Conference on Computer Supported Education

104

of Pisa and two simpliﬁed versions that can be ob-

tained by applying the algorithms. Only the original

image is stored, the simpliﬁed versions are computed

on the ﬂy. Simpliﬁed images can also be printed with

a Braille printer thus offering to blind users some in-

formation of the shape of the educational object itself.

MediaWiki Engine Modiﬁcations. The Medi-

aWiki PHP code has been modiﬁed in several aspects.

First, for creating, handling and visualizing educa-

tional objects. Moreover, in order to guarantee the

accessibility of pages, the generated HTML code has

been checked and parsers for checking and sanitizing

the content introduced by the users have been added.

Users proﬁles containing information about the

users’ visualization preferences as well as the level

of details of the descriptions of educational objects

have been integrated in the standard MediaWiki user

proﬁle. This information is used both for the visual-

ization and retrieval of educational objects.

Finally, retrieval facilities that take the user’s pro-

ﬁle into account to rank and return educational objects

have been introduced. For each educational object all

the available descriptions are returned, sorted accord-

ing to the resulting ranking. The snippet associated

with each link contains also the type of the description

(complete, summary, simpliﬁed), a qualitative assess-

ment of the result (good, discrete, sufﬁcient), and the

percentage rank value. As an example of the search

results see Figure 3.

4 ACCESSIBILITY EVALUATION

VisualPedia needs to deliver educational contents

compliant to the technical accessibility requirements

as speciﬁed in the Italian Law n. 4/2004 (Legge

Stanca, 2004). Therefore we have updated the Me-

diaWiki output functions to meet the Italian require-

ments and - when possibile - also WCAG 1.0 (W3C,

1999).

Technical accessibility tests have been per-

formed asking a group of experts to ana-

lyze VisualPedia. They used CSE Validator

(http://www.htmlvalidator.com/) to check the HTML

code and analysed the contrast between the back-

ground and the text by inspecting the CSS ﬁles.

A report has been written highlighting some tech-

nical mistakes that have been corrected in the current

version of the system. The lack of alternative text

(alt tag) for some images or the lack of back buttons

in some pages are examples of the identiﬁed mistakes.

In addition, two small parts of the interface did not

show enough contrast between the background and

the foreground.

Figure 3: Search results.

The general (and informal) comment was that,

apart from minor mistakes, the accessibility of the

pages is generally good at least from a technical point

of view, i.e. when considering the “quality” of the

output HTML code. However, they also observed that

the real problem will move to the production and up-

load of the educational objects. Indeed, even if the

framework is perfectly accessible, the contents within

stored could not be accessible. Therefore, data entry

is the critical phase that could possibly vanishing the

development of an accessible framework.

Unfortunately, as we know from the literature, the

conformance to the technical accessibility guidelines

does not necessary mean that a Web site is usable

or that ”websites that achieve higher conformance to

WCAG are also more usable by people with disabili-

ties. . . ” (Petrie and Kheir, 2007).

5 USABILITY EVALUATION

In this section we present an usability study obtained

from a combination of two methods: an evaluation of

the user interface according to general purpose heuris-

tic principles and a test with end-users. They will

lead to some reﬂections on the validity of VisualPe-

dia. The experimental test is designed in order to

involve different communities of users, in particular

disabled users, to check the software and the assump-

tions on image simpliﬁcation.

The ﬁnal goal of this usability study is to con-

tribute to the discussion on evaluation of Web appli-

cations to support communities of practice using Web

2.0 approach. The main aim is to determine whether

the results of the heuristic evaluation accurately re-

ﬂected the user’s expectations and needs and whether,

ACCESSIBILITY AND USABILITY OF A COLLABORATIVE E-LEARNING APPLICATION

105

Table 1: Heuristic principles applied in this ﬁrst test.

1 Visibility of system status

2 Match between the system and the real world

3 User control and freedom

4 Consistency and standards

5 Recognition rather than recall

6 Flexibility and efﬁciency of use

7 Display all but needed information

8 Error prevention

9 Help users recognize, diagnose, recover from errors

10 Help and documentation

in spite of poor usability, VisualPedia can be consid-

ered an innovative e-learning system for disseminat-

ing educational matters, which future releases might

bring good results.

5.1 Test with Experts

Participants. The heuristic evaluation was con-

ducted by ﬁve evaluators with experiences in HCI and

was designed according to the conventional practice

(Nielsen, 1995). This heuristic evaluation strategy

followed an approach without observers and adopted

the Nielsen heuristics as evaluation grid (Table 1).

Tasks. According to the Nielsen principles no spe-

ciﬁc task should be assigned to the evaluators. Each

one can choose a user proﬁle for which her analysis

would be performed. Then, she can move on the sys-

tem as she prefers in order to check the system.

Procedure. The ﬁve evaluators examined VisualPe-

dia according to a speciﬁc user proﬁle and each one

performed one analysis session and wrote a report

describing the identiﬁed usability problems. A de-

brieﬁng session among all the evaluators followed in

which the problems detected by each of the evaluators

have been collected and organized into categories.

Results. Table 2 was built relying on the results of

this ﬁrst test with experts. The table reports only

the main usability bugs discovered. For each bug a

ﬁrst column describes the problem, a second column

presents the heuristics that have been violated and ﬁ-

nally in the last column the number of evaluators that

have discovered that problem is reported.

Preliminary Considerations. According to the re-

sults of the heuristic evaluation, the diagram in Fig-

ure 4 reports the number of heuristics that have been

violated. This diagram highlights that seven out of

the ten heuristic principles were violated while the

other were supported in a satisfying degree but not

fully (except for principle number 9).

In addition to the previous result, the evaluators

carried out a qualitative analysis of the data collected

Figure 4: Distribution of the violated principles.

using a methodology that adopts six main categories

in evaluating usability of Web-based systems (Polillo,

2004). These categories are: accessibility, architec-

ture, content, communication, functionality, manage-

ment and usability. The methodology is based on the

attribution of a score (from 0 to 4) for each category

by each evaluator. By exploiting these scores it is pos-

sible to draw another diagram reporting the qualita-

tive analysis of VisualPedia (Figure 5). The diagram

points out a low evaluation along the different cate-

gories with the exception of the accessibility category.

By a discussion with them, we identiﬁed that the low

scores were due to the lack of content within Visual-

Pedia, that did not encourage their evaluation of the

characteristics of the system. In the concluding sec-

tion, we will propose countermeasures for this issue.

Figure 5: Qualitative analysis of VisualPedia.

5.2 Test with End-users

Participants. The test with end-users was conducted

with a group of 25 students with different forms of

disability, different ages (the majority being over 14),

different education (ranging from primary to high

schools), and different skills in the use of the com-

puter and/or assistive tools. The majority (22) hold

a computer at home and use it mainly for study-

ing and Web browsing. They use Internet to search

CSEDU 2010 - 2nd International Conference on Computer Supported Education

106

Table 2: Main result of the test where experts were asked to conduct a heuristic evaluation study.

Problems Heuristic violated N. of evaluators

Some links point to pages of Web site maintenance 2 2

Log page is too detailed. All users can see the other users’ activities 3 1

In some tables the formatting is not correct and some XML code appears 8 3

The search function ﬁnds only exact matches 6 3

Poor information on how to search 10 5

The home page has too many items and the list is too complex 5,7 3

In the ﬁrst page the educational object is called page 4 1

Poor information about VisualPedia functions 10 5

The menu appears even when not needed 7 3

Poor information about the menu containing the MediaWiki preferences 2,10 2

Image visualization and simpliﬁcation take too long 1 2

Some borders of the object inhibit the content understanding 1 2

How to come back is not clear 1,4,5, 6,7,10 5

It is very difﬁcult to understand where I am 5,6,10 1

Some technical terms are not clear 2 1

In some cases the menu on the left mysteriously changes 1,4,5,7 1

Some pages contain HTML code in clear 2 1

No help available 10 5

contents for their homeworks (14), exchange e-mails

(12), communicate via forums and/or social networks

(9). In the following we report for each category of

users the assistive technologies employed for the ex-

periments and/or the preferences set in their proﬁles.

• 7 users (3 male, 4 female) were totally blind: they

all used the screen reader software Jaws (a product

developed and distributed by Freedom Scientiﬁc

http://www.freedomscientiﬁc.com); in addition to

Jaws, two students also used the Braille display at

the same time.

• 10 users (5 male, 5 female) were visually im-

paired: ﬁve of them used magniﬁer softwares,

four used the enlarger tools provided by the oper-

ating system (Windows Vista), and one enlarged

font size using the VisualPedia personalization.

• Finally, 8 users (5 male, 3 female) were af-

fected by down syndrome with intelligence quo-

tient lower than 45. Two of the users were also

affected by visual impairment. They used Visual-

Pedia preferences: text was shown in upper-case

and surrounded by thick lines (see Figure 6) to fo-

cus their attention.

Tasks. Individual half-hour sessions have been orga-

nized in which every single user has been asked to

perform really simple tasks:

1. login to the system,

2. browsing, i.e. ﬁnding an educational object fol-

lowing the available links, and

3. searching, i.e. ﬁnding the same educational object

using the internal searching facility.

Short questionnaires have been delivered before

and after the execution of these tasks.

Procedure. All participants performed their tasks in

front of two evaluators (selected among the authors of

this paper) but they were assured that the evaluation

was on the VisualPedia software, not on their abilities

on the use of the browser and/or assistive tools. In

some cases, parents were also present during the test,

without interfering during the execution of tasks.

Results. All the users performed successfully their

tasks. Some needed more time and/or some help but

this was mainly due to their inexperience in the use

of the computer. Visually impaired users did not have

any problem, blind students had some troubles with

the screen reader, especially for those pages with extra

information (we realized that it could be easily omit-

ted since it does not add real value to the page).

The post-test questionnaires pointed out that Visu-

alPedia is generally easy to use. Speciﬁcally, Figure

7 shows that the execution of the three tasks (login,

browsing, searching) did not pose any real problem to

the majority of them.

Preliminary Considerations. Running the tests with

this group of end-users has been particularly emo-

tional. Students really tried hard to do their best, also

those with difﬁculties in the use of computer and/or

assistive tools. We realized they really felt as part of

the project, trying to provide their help, each one ac-

cording to her/his capacities.

Moreover, these tests highlighted some problems

that were not detected during the heuristic evaluation.

As an example, in the second task – browsing to ﬁnd

some content – the interface displays a combo box to

ACCESSIBILITY AND USABILITY OF A COLLABORATIVE E-LEARNING APPLICATION

107

Figure 6: VisualPedia personalization.

select the educational object (see Figure 6). While the

combo box is generally easy to interact with, those

users which were not particularly expert in the use

of assistive tools had more difﬁculties in ﬁnding and

interacting with this element of the Web page.

As another example, consider the third task: ﬁnd-

ing content using the internal search engine. For each

object matching the query, VisualPedia returns three

different descriptions (if available) plus some addi-

tional details. This abundance of information has

been indicated as useless during the heuristic evalu-

ation and, even worst, resulted annoying for some of

the blind students. Jaws, in fact, started reading every

single word, causing confusion and disorientation.

6 FUTURE DEVELOPMENTS

We have reported the experiences gained in the eval-

uation of accessibility and usability of the VisualPe-

dia system, a collaborative wiki-based system for the

management of educational objects tailored also for

students with special needs.

The evaluation results highlight several problems

that need to be ﬁxed in the next version of VisualPe-

dia. Moreover, further improvements are required for

what concern the management of XML data and the

knowledge base of VisualPedia. We conclude the pa-

per with considerations on these aspects that open to

future research activities.

Usability. Innovative systems like VisualPedia rad-

ically change the way in which users participate to

the creation process of the knowledge at the base of a

system. (Bruns, 2008) shows how in the collaborative

content creation environments like WikiPedia what is

emerging is no longer just a new form of content pro-

duction, but a “new process for the continuous cre-

ation and extension of knowledge and art by collabo-

rative communities: produsage”. The usability evalu-

ations presented in this paper are devoted not only to

Figure 7: Post-test results.

ﬁnd simple usability bugs but to understand the use-

fulness and acceptance of an application like Visual-

Pedia even if, as discussed in Section 2, in some cases,

focusing on the usability can be harmful. We think

that the innovative ideas at the base of VisualPedia

should not be discouraged by some negative results of

the heuristic evaluation when they are balanced by the

genuine enthusiasm demonstrated during the test with

end-users. For this, starting from the current poor us-

ability, it is necessary to design new improvements,

following a star cycle of software development, able

to bring to the deﬁnition of an innovative e-learning

system for disseminating educational matters reﬂect-

ing the user’s expectations and needs.

XML Representation of Educational Objects.

The representation of educational objects as XML

documents has many relevant implications both for

what concerns accessibility of the content and effec-

tive organization and retrieval of information. In-

deed, one of the crucial aspects of Web-content ac-

cessibility is the organization of the data according to

HTML tags that lead to documents that are not acces-

sible. The possibility to organize information using

XML allows the separation of information representa-

tion from its logical organization. Therefore, speciﬁc

representations depending on the user proﬁles can be

tailored and thus overcoming the current limitations

(e.g. tables that are usually poorly accessible can be

structured in XML and speciﬁc layout based on div

tags can be obtained for visual impaired users through

XSL transformations). Moreover, XML based rep-

resentations of mathematical formulas (e.g. through

Math-ML) can be integrated in the educational object

representation. This can facilitate the access of math-

ematical formulas through screen readers and Braille

displays. Finally, in the spirit of semantic WikiPedia

(Kr

otzsch and Vrandecic, 2009), ontological organi-

zation of educational objects can be provided and ex-

ploited for searching facilities within the VisualPedia

repository.

CSEDU 2010 - 2nd International Conference on Computer Supported Education

108

VisualPedia Knowledge Base. As pointed out also

from the heuristic analysis, the number of educational

objects currently stored within VisualPedia is too low

(around 30 objects). Despite our great efforts in push-

ing the teachers to provide a certain number of educa-

tional objects, they end up with a really small number.

This is due, from the one hand, to the high working

load in the teaching activities and, to the other hand,

to a lack of motivation in participating to the project.

In order to face this problem we need to move

along different directions. First, we plan to consider

mash-up technologies for gathering educational ma-

terials from the Web (e.g. WikiPedia) and then tai-

loring the materials to the user with disabilities. In

this way, the original information can be left in the

original source and only a view of it can be stored

within VisualPedia with modiﬁcations necessary to

face the user difﬁculties. The second direction is to

directly involve the students in the production of ed-

ucational objects under the teacher supervision. By

integrating in VisualPedia tools like WikiDashboard

(http://wikidashboard.parc.com/), teachers can check

the students’ activities and identify those that effec-

tively worked on their homeworks. Last but not least,

we should work on teachers’ motivation, trying to un-

derstand how to involve them in the development of

new educational objects in the same spirit of authors

of user-generated content (Nov, 2007).

ACKNOWLEDGEMENTS

This work has been partially funded by the Minis-

tero Italiano della Pubblica Istruzione under Azione 6:

Progetto nuove tecnologie e disabilit

a. We would like

also to thank all the participants who took part to the

evaluation phase for their patience and enthusiasm.

REFERENCES

Boccacci, P., Ribaudo, M., and Mesiti, M. (2009). A collab-

orative environment for the design of accessible edu-

cational objects. IEEE/WIC/ACM International Con-

ference.

Bruns, A. (2008). Blogs, Wikipedia, Second Life, and Be-

yond: From Production to Produsage. Peter Lang

Publishing.

Canny, J. (1986). A computational approach to edge de-

tection. IEEE Trans. Pattern Analysis and Machine

Intelligence, 1(8).

Cathleen, W., John, R., Clayton, L., and Peter, P. (1994).

The cognitive walkthrough method: a practitioner’s

guide. In Usability inspection methods, pages 105–

140. John Wiley & Sons, Inc.

Dix, A., Finlay, J., Abowd, G., and Beale, R. (2003). Hu-

man Computer Interaction. Prentice Hall 3rd Edition.

Greenberg, S. and Buxton, W. (2008). Usability evaluation

considered harmful (some of the time). In CHI, pages

111–120.

Kelly, B., Sloan, D., Brown, S., Seale, J., Petrie, H., Lauke,

P., and Ball, S. (2007). Accessibility 2.0: people, poli-

cies and processes. In W4A ’07: Proc. of Int. Conf. on

Web Accessibility, pages 138–147. ACM.

otzsch, M. and Vrandecic, D. (2009). Semantic Wiki-

pedia. In Social Semantic Web, pages 393–421.

Springer.

Legge Stanca (2004). Disposizioni per favorire l’accesso

dei soggetti disabili agli strumenti informatici.

http://www.camera.it/parlam/leggi/04004l.htm,

retrieved at Jan. 2010.

Moss, T. (2008). 10 common errors when implement-

ing accessibility. http://www.webcredible.co.uk/user-

friendly-resources/ web-accessibility/errors.shtml, re-

trieved at Jan. 2010.

Nielsen, J. (1995). Usability inspection methods. In CHI

’95: Conference companion on Human Factors in

Computing Systems, pages 377–378. ACM.

Nov, O. (2007). What Motivates Wikipedians? Communi-

cations of the ACM, 50(1):60–64.

Petrie, H. and Kheir, O. (2007). The relationship between

accessibility and usability of websites. In CHI ’07:

Proc. of the SIGCHI Conf. on Human Factors in Com-

puting Systems, pages 397–406. ACM.

Polillo, R. (2004). Il check-up dei siti Web. Valutare la

qualit

a per migliorarla. Apogeo.

Sauvain, R. and Szilas, N. (2009). Edushare, a step be-

yond learning platforms. In EC-TEL, volume 5794 of

LNCS, pages 283–297.

Shneiderman, B. (2000). Universal usability. Communica-

tions of the ACM, 43(5):84–91.

Silva, P. A. and Dix, A. (2007). Usability - not as we know

it! In HCI 2007 - The 21st British HCI Group Annual

Conference.

Taras, C., Siemoneit, O., Weißer, N., Rotard, M., and Ertl,

T. (2008). Improving the accessibility of wikis. In

ICCHP ’08: Proc. of 11th Int. Conf. on Computers

Helping People with Special Needs, pages 430–437.

Springer-Verlag.

Thompson, A.-J. and Kemp, E. A. (2009). Web 2.0: extend-

ing the framework for heuristic evaluation. In CHINZ,

pages 29–36.

Veen, W. (2009). Homo Zappiens, learning in a digital age.

In 1st Int. Conf. on Computer Supported Education,

Lisbon, Portugal. Keynote speech.

Verpoorten, D., Glahn, C., Kravcik, M., Ternier, S., and

Specht, M. (2009). Personalisation of learning in vir-

tual learning environments. In EC-TEL, volume 5794

of LNCS, pages 52–66.

W3C (1999). Web content accessibility guidelines 1.0.

http://www.w3.org/TR/WCAG10/, retrieved at Jan.

2010.

ACCESSIBILITY AND USABILITY OF A COLLABORATIVE E-LEARNING APPLICATION

109