![](bg6.png)
4 CONCLUSIONS AND FUTURE
WORK
On this paper we have presented AudioBrowser, a
talking browser that aims at enabling visually im-
paired users to efficiently navigate the web. Audio-
Browser takes advantage of the underlying structure
of the document in order to better present its content,
and is able to present different views of the same doc-
ument enabling the user to choose the view that better
fits its navigational needs at each moment. This fea-
ture is useful not only to visually impaired users, but
also for web designers since it allows them to have a
broader perspective of the construction of the page.
We have received reports of several web designers
that are using the AudioBrowser to this effect.
Features and issues related to talking browsers
were introduced. In this context some of the main
features of the AudioBrowser were described. Focus
was on features that provide increased accessibility.
From a usability point of view the main objection,
as mentioned previously, is the need to switch back
and forth from using the browser to a screen reader,
which is still needed for all the other tasks. This is one
usability problem that was not foreseen neither by the
developers nor the visually impaired people that par-
ticipated on the AudioBrowser project from the very
beginning.
A talking browser and a screen reader are incom-
patible because they both produce speech, and there-
fore they can not be working simultaneously.
However talking browsers have several advantages
over screen readers and should not be put aside. Talk-
ing browsers make it easy to manipulate the web page
and present different views, therefore minimizing the
scanning time for visually impaired users.
Can talking browsers and screen readers be com-
patible? A possible solution for this problem could be
to have the browser not producing any speech at all.
Instead the browser would produce text that would
then be read by the screen reader. It may seem ab-
surd to have a talking browser not talking at all, but
under this approach the browser is in fact "talking"
to the screen reader. In order for this solution to be
successful the process would have to be completely
transparent to the user. To the user the application
would be an enhanced web browser.
A totally different approach would be to have a web
service that provides a subset of the functionality of a
talking browser solution. This requires a proxy based
approach where a web address would be provided.
The page would be transcoded in several views in a
web server, and all those views would be supplied to
a regular browser, for instance using frames. The user
would be able to specify which views would be of in-
terest, and each view, plus the original page, would be
supplied in a separate frame properly identified.
Although this approach requires extra bandwidth
we believe that with the advent of broadband, and
considering that the different views have only text, i.e.
no images, the overhead would be bearable consider-
ing the extra functionality.
Combining a proxy based approach and a talking
browser can provide an enhanced web experience, as
the transcoded web pages may contain special tags
recognized by the talking browser.
ACKNOWLEDGEMENTS
The research reported in here was supported by
SNRIP (The Portuguese National Secretariat of Re-
habilitation and Integration for the Disabled) under
program CITE 2001, and also FCT (Portuguese Foun-
dation for Science and Technology) and POSI/2001
(Operational Program for the Information Society)
with funds partly awarded by FEDER.
REFERENCES
Fernandes, A. R., Martins, F. M., Paredes, H., and Pereira, J.
(2001). A different approach to real web accessibility.
In Stephanidis, C., editor, Universal Access in H.C.I.,
Proceedings of HCI International 2001, volume 3,
pages 723–727. Lawrence Erlbaum Associates.
Filepp, R., Challenger, J., and Rosu, D. (2002). Improving
the accessibility of aurally rendered html tables. In
Proceedings of the fifth international ACM conference
on Assistive technologies, pages 9–16. ACM Press.
IBM (2000). Web accessibility transcoding system.
http://www.trl.ibm.com/projects/acc_tech/attrans_e.htm.
Jackson-Sanborn, E., Odess-Harnish, K., and Warren,
N. (2002). Website accessibility: A study of ada
compliance. Technical Report Technical Reports
TR-2001-05, University of North Carolina
˝
UChapel
Hill, School of Information and Library Science,
http://ils.unc.edu/ils/research/reports/accessibility.pdf.
Rowan, M., Gregor, P., Sloan, D., and Booth, P. (2000).
Evaluating web resources for disability access. In Pro-
ceedings of ASSETS 2000, pages 13–15. ACM, ACM
Press.
Stevens, R. D. and Edwards, A. D. N. (1996). An approach
to the evaluation of assistive technology. In Proceed-
ings of ASSETS ’96, pages 64–71. ACM, ACM Press.
Sullivan, T. and Matson, R. (2000). Barriers to use: Usabil-
ity andcontent accessibility on the webŠs most popu-
lar sites. In Proceedings on the conference on univer-
sal usability, 2000, pages 139–144. ACM Press.
Zajicek, M., Venetsanopoulos, I., and Morrissey, W. (2000).
Web access for visually impaired people using active
accessibility. In Proc International Ergonomics Asso-
ciation 2000/HFES 2000.
ICEIS 2004 - HUMAN-COMPUTER INTERACTION
80