ASSISTIVE TECHNOLOGIES AND TECHNIQUES FOR WEB

BASED EGOV IN DEVELOPING COUNTRIES

Heiko Hornung, M. Cec´ılia C. Baranauskas

Institute of Computing/NIED, State University of Campinas (UNICAMP), Cx. Postal 6176, Campinas, Brazil

Claudia A. Tambascia

CPqD – Telecommunication Research and Development Center, Brazil

Keywords:

Assistive Technologies, eGov, HCI.

Abstract:

Electronic government (eGov) is intended to serve the whole spectrum of the population. To be able to access

services and thus beneﬁt from eGov, many users require assistive technologies and techniques (ATT). This

demand is implied by auditory, visual or other impairments but also by low literacy skills. Thus, in the context

of this paper the term ATT is interpreted in a broader sense compared to the classic deﬁnitions of the term

“assistive technology” that relate to impairments or other special needs. This paper explores the range of ATT

on the background of the special conditions found in developing countries. We investigate which types of

users beneﬁt from ATT in which ways and discuss which categories of users have requirements that are not

yet covered by the current stage of development of ATT. The reasons for the remaining problems regarding

access may be due to factors related to the country context or due to current technological limitations. Some

lessons learned from our ﬁndings will be presented which indicate directions for further research.

1 INTRODUCTION

Electronic government(eGov) is intended to serve the

whole spectrum of the citizens in a society. Devel-

oping countries have characteristics that signiﬁcantly

differ from developed countries affecting directly the

design of eGov systems (Hornung and Baranauskas

2007). To illustrate our case we pick Brazil as a rep-

resentative of this category of nation. Knowing that

the situation in developing and emerging countries

throughout the world is not uniform we will point out

this fact whenever we believe that our considerations

are not universally valid. The context of this paper

is thence deﬁned by the design of web based eGov

services, the country context of Brazil and our HCI

perspective towards it. Guiding principles of our re-

search are those of Universal Design, i.e. instead of

focussing on solutions for people with speciﬁc im-

pairments we search for solutions that facilitate the

access for all, in the sense of the largest possible au-

dience.

A web based eGov service is accessible via the

Internet. Thus, it is typically accessed via a web

browser or other client-side interfaces (e.g. media

players with text and graphic rendering facilities or

any other custom-built application). Although web

based services can also be accessed via other clients

like mobile handheld devices or digital interactive

television, we limit our discussion in this paper to

computer based clients. The reasons for this de-

cision are a small (mobile handhelds) or virtually

non-existent (digital interactive television) market-

penetration on the one hand and too many open ques-

tions regarding universal access to these devices on

the other hand. Nevertheless, we believeour consider-

ations can also contribute to discussions of those two

classes of client devices.

We situate our discussion on web based eGov ap-

plications, i.e. services created by a governmental in-

stitution to be accessible via the Internet. eGov ser-

vices can be categorized in three different levels: a

ﬁrst-level service offers static information (e.g. texts

of laws, health information, etc.), a second-level ser-

vice enables the execution of an electronic process

(e.g. income tax declaration), whereas a third-level

service offers participation in democratic processes

(e.g. discussion forums, wiki-style creation of draft

laws). The main difference to non-eGov services is

248

Hornung H., Cecília C. Baranauskas M. and A. Tambascia C. (2008).

ASSISTIVE TECHNOLOGIES AND TECHNIQUES FOR WEB BASED EGOV IN DEVELOPING COUNTRIES.

In Proceedings of the Tenth International Conference on Enterprise Information Systems - HCI, pages 248-255

DOI: 10.5220/0001707102480255

 SciTePress

that the content of those services tends to be more

complex than that of other web applications such

as shopping sites or social networking services (by

which we do not mean that the design problem to cre-

ate easily accessible non-eGov sites is trivial).

The goal of this paper is to compile a list of tech-

nologies and techniques that can be employed for

building and using web based eGov services. Fur-

thermore we propose criteria to evaluate the items on

this list and give indications on technologies and tech-

niques best suited to our context.

The paper is organized as follows: Section 2 con-

textualizes the discussion on the challenges we face in

developing countries regarding interaction design for

eGov. Section 3 surveys recent literature on related

subjects. Section 4 presents our compilation of as-

sistive technologies and techniques focussing on so-

lutions that make sense in our country context. We

also present some lessons learned from the technolo-

gies and techniques analysis to deal with our major

challenge: illiteracy. Section 5 concludes.

2 COUNTRY CONTEXT AND

USERS – THE CHALLENGES

The Brazilian population is characterized by a vast di-

versity regarding different demographic dimensions,

disabilities, literacy and digital divide. The aim of

this section is not to draw a statistically comprehen-

sive picture of Brazil. Since the scope of our work is

the potential inclusion of all citizens, a small, but sig-

niﬁcant number in any given demographic dimension

is sufﬁcient for that dimension to be further consid-

ered. About 10% of the Brazilians have a visual im-

pairment, and about 5% have auditory or motor im-

pairments respectively (IBGE 2000).

As there are different methods for measuring lit-

eracy there are also different statistics regarding liter-

acy in Brazil (IPM 2005; UNDP 2007). In the con-

text of the latest Human Development Report of the

United Nations Development Programme, literacy is

deﬁned as the percentage of people of ages 15 and

older “who can, with understanding, both read and

write a short, simple statement related to their every-

day life.” (UNDP 2007, p. 368) and thus an adult

illiteracy rate of 11.4% is diagnosed for Brazil. Re-

gardless of the methods and deﬁnitions of literacy,

the proportionof illiterate and semi-literate Brazilians

cannot be ignored.

As to digital literacy, in a 2005 survey of the

IBGE, only 22.9% of the Brazilians over 10 years

stated that they had accessed the Internet during the

last three months. 50.0% of those have accessed the

Internet from their own homes. Considering those

who exclusively accessed the Internet from their own

homes, 38.5% had dial-up access only. The 77.1%

who didn’t access the Internet during the reference

period of three months, 37.2% didn’t have access to a

computer at all (IBGE 2005).

Considering that a great percentage of the Brazil-

ian population does not have access to computers and

the Internet from their homes, telecenters and internet

caf´es or other public spaces play an important role

in providing access to technology and will be consid-

ered accordingly in this paper. In Brazil, many tele-

center initiatives are government-driven, whereas in

other countries non-governmentorganizations play an

important role as well. In either case, long-term cost-

efﬁciency and sustainability are important require-

ments to consider.

We havenot yet deﬁned the term “digital literacy”,

neither do we intend to give a formal deﬁnition. In our

context, the minimum requirementsfor the interaction

with an eGov service are the proﬁciency in the use of

keyboard, mouse and other peripherals, as well as the

ability to use browser-like interfaces that contain text,

text input areas, multimedia areas (images, audio or

video), links, and buttons or other clickable active ar-

eas. Looking at the statistics we presented, we have

to presume that a signiﬁcant part of the Brazilian pop-

ulation does not show sufﬁcient competencies to in-

teract with web based eGov services without further

assistance, be it personal or via technology.

Given the context and user scenario described

above, we face some challenges that differ from the

canon of literature we encountered about assistive

technologies. Regarding the users, we are not in a po-

sition where we know their competencies and needs,

like for example a company that has to ﬁt out a num-

ber of workstations to suit the needs of its employees

with for example physical or other impairments. Nei-

ther is our situation that of an organization or insti-

tution that offers a facility where for example people

with visual impairments can access computer termi-

nals and count on the help of a trained attendant.

Following the principles of Universal Design, we

have to think about solutions that enable people with

all possible competencies to use eGov services. Re-

garding the country context, we can not simply create

incentives for people with special needs to buy their

own assistive technology that optimally suits them. In

the public or quasi-public areas of access (telecenters,

internet caf´es, etc.), we can not rely on trained per-

sonnel that always can assist users with special needs.

Since telecenters are only available in about 50% of

the Brazilian municipalities (IBGE 2006), we may of-

ten encounter places where private internet caf´es pro-

ASSISTIVE TECHNOLOGIES AND TECHNIQUES FOR WEB BASED EGOV IN DEVELOPING COUNTRIES

249

vide the only possibility to access computers and the

Internet. Considering the lack of public-private part-

nerships in this area, we can not expect that internet

caf´es provide all assistive technologies required by

potential users in their surroundings. Another chal-

lenge that we face are low digital literacy skills of po-

tential users.

3 RELATED WORK

This paper does not intend to give a comprehensive

overview of the ﬁeld, and some of the technologies

presented in this paper are already well known since

quite some time. However, we will present exemplary

solutions that address some special needs or compe-

tencies discussed in further sections and that represent

recent developments in the area. The solutions pre-

sented show some ideas regarding how to enable the

access to eGov services in general as well as how to

facilitate access to web based applications for users

with low literacy skills and users with auditory, cog-

nitive, motor or visual impairments.

Pilling and Boeltzig (2007) provide a starting

point by identifying the lack of assistive technolo-

gies as one barrier to the access to eGov services.

However, since they follow a strategically focused ap-

proach and base their investigations to initiatives in

the U.S. and U.K., their ﬁndings have only limited

applicability to our context.

Independently of different ways of measuring lit-

eracy, many people in developing countries have low

or no reading skills at all (UNDP 2007). Since eGov

services are supposed to reach and beneﬁt especially

people with low literacy skills, one essential challenge

is to provide access to these users. Medhi et al. (2007)

investigate different options to audio-visually repre-

sent healthcare-related concepts to people with low

or no literacy skills. Their main ﬁndings point out

that auditory information is very important for com-

prehension, but can confuse the subjects due to multi-

modal effects when used together with visual infor-

mation; richer information not always results in better

understanding. They also examine when to use static

images (i.e. photos and drawings) and when to use

videos or animations and conclude that it depends on

the content to be represented.

The averagereading level among the group of deaf

people is signiﬁcantly lower than that of the hearing.

Furthermore images and icons that are meaningful to

the hearing might not be so to the deaf and hard of

hearing. For the community of deaf people, the effort

in reading text in a spoken language is comparable to

the effort in reading a foreign language. Kennaway et

al. (2007) explore the possibilities of providing signed

content in web based applications. They identify ad-

vantages of signing avatars over videos and propose a

set of tools for the generation and delivery of signed

content via a browser plug-in. One of the challenges

lies in dynamically generated text with a previously

unknown structure; this problem has a similar com-

plexity to the automatic translation between spoken

languages and thus signing avatars yield comparable

results.

The literature review reveals that accessibility for

users with cognitive disabilities is a ﬁeld where much

work still has to be done. Based on the principles of

Universal Design, Sevilla et al. (2007) propose guide-

lines to redesign conventional web content in order to

make it cognitively accessible. A comparative study

shows that for example short-memory problems do

occur in the conventional version of a web page but

not in the cognitively accessible version of it.

Assistive technologies for users with motor im-

pairments are often hardware based solutions, e.g.

alternative input devices. However, there are also

software based solutions that improve the accessi-

bility of standard input devices like trackballs or

mice. These solutions are especially interesting in

our context since they are potentially easier to de-

ploy and maintain in a large scale basis. Wob-

brock and Gajos (2007) compare the target acquisi-

tion paradigms “area pointing” and “goal crossing”.

As opposed to area pointing, users do not click in an

area but pass over a target line. Although it seems

that users with motor impairments prefer goal cross-

ing and are able to achieve a better performance in

certain conditions, many open questions still remain,

e.g. how to design goal crossing interfaces or which

competencies of users with motor impairments are

best suited for this kind of interfaces.

Screen readers and refreshable Braille displays

represent the contents of web pages in a linear manner

to the blind user. Besides being time consuming, the

linearity also may complicate the comprehension of

the content, since images, tables and other structural

information are perceived differently than by a user

with no visual impairment. One approach to over-

come this limitation is the use of haptic devices that

provide tactile feedback. Kuber et al. (2007) present

a participatory approach to design feedback for web

based applications. Besides mostly used as a com-

plement for screen readers and other solutions, haptic

devices can possibly facilitate the access of users with

visual impairments that have no familiarity with the

use of screen readers. However, the work on haptic

devices has not yet reached a maturity that permits its

use in our context.

ICEIS 2008 - International Conference on Enterprise Information Systems

250

4 ASSISTIVE TECHNOLOGIES

AND TECHNIQUES

The term “assistive technology” is often deﬁned as

a set of technologies “[...] that increase, maintain,

or improve the functional capabilities of individuals

with disabilities

[...]” (U.S. Department of Health

and Human Services 2007). Although such technolo-

gies are not restricted to computer systems, but also

include devices found at the work places or homes of

people (e.g. phone foot switches or arm and elbow

supports), within the limits of this paper only com-

puter related artefacts are of interest to us. Due to the

context previously described, we will include people

without impairments but with low literacy skills into

our considerations regarding accessibility. Thus we

will take a broader view on the subject and use the

term “assistive technologies and techniques” (ATT)

to describe solutions that have the potential to enable

and facilitate the access and use of web based eGov

services for our target audience, i.e. citizens with all

possible special needs and competencies. This in-

cludes assistive technologies, but also methods, best

practices or other solutions like earcons (Brewster

1998), the use of multimedia content in web pages,

etc. Our work is therefore in-line with the shift from

research and development which considers assistive

technologies for people with disabilities to Universal

Design of solutions for the largest possible audience

(Law et al. 2007).

There exists a wide range of special needs implied

by conditions such as physical (motor, mobility), sen-

sory (auditory or visual) or cognitiveimpairments, de-

velopment disability or mental retardation. Table 1

displays a list of categories of assistive technologies

examined under the aspect of ﬁve signiﬁcant dimen-

sions:

• Beneﬁciary: denotes the category of user who

beneﬁts from the solution: users with auditory,

motor or visual impairments (columns A, M, and

V) or users with low literacy skills (column L ).

• Input/Output (I/O): denotes whether a technology

is used during data input (I) or output (O).

• Implementation in hardware or software (HW/

SW): software based solutions work without spe-

cial hardware (e.g. screen magniﬁers); hardware

based solutions generally work independently of

the application (e.g. braille embossers). The fact

that hardware based solutions generally require a

driver or other software to function is neglected

We intentionally ignore discussions about the appro-

priate use of terms like “disability” or “impairment” in this

paper.

here, however there are hardware based solutions

where the software part is signiﬁcant (e.g. bio-

metric devices) or where dedicated hardware so-

lutions exist besides the software-only solutions

(e.g. speech synthesizers). These cases are de-

noted by “HW/SW”.

• Maturity: in our context a technology is consid-

ered mature, if it functions with acceptable error

rates in our scenario of public access areas with

many different users that are potentially unskilled

in the use of the technology. The technology has

to function under a wide range of environmental

conditions since background noise, light, temper-

ature, etc. cannot be controlled. Our measure is a

qualitative one and can not be backed up by quan-

titative data. A “+” indicates a sufﬁcient, a “-” an

insufﬁcient maturity level. A “+/-” indicates that

the technology has an unacceptable maturity level

under certain circumstances (e.g. handwriting in

the case of optical character recognition and un-

availability or low quality of speech synthesizer

voices for some languages).

• Training: denotes whether an assistive technology

can be used without prior training. We do not

quantify the amount of training required. A “-”

means that no training is required, although the

user will improve her performance with increas-

ing familiarity. Technologies with a “+” require a

certain amount of prior training. Apart from that,

there also exist technologies that have to “learn

to interpret” the user input (e.g. voice recognition

software). A “(-)” means that the user does not

need any training but other knowledge to be able

to use the respective technology (e.g. users of re-

freshable Braille displays need to know Braille).

Disregarding entries that only indirectly beneﬁt

certain categories of user (entries marked as “(x)”),

a look at the table shows that there seems to be quite

some solutions for people with visual impairments,

not quite as many for people with auditory, cogni-

tive and motor impairments and very few solutions

for people with low literacy skills.

Regarding the deaf and the hard of hearing, the

solutions presented in the table either require liter-

acy skills (TTY/TDD conversion modems), are only

suited for very speciﬁc purposes (light signaler alerts)

or are not yet mature enough to be used in a large scale

off-laboratory scenario (gesture recognition, sign syn-

thesis).

Apart from technologies for data input (keyboard

ﬁlters and hardware based input devices), there seem

to be no technologies speciﬁcally developed for users

with cognitive impairments.

ASSISTIVE TECHNOLOGIES AND TECHNIQUES FOR WEB BASED EGOV IN DEVELOPING COUNTRIES

251

Table 1: Assistive technologies.

ASSISTIVE TECHNOLOGY

BENEFICIARY

I/0 HW/SW MATURITY TRAINING

A C L M V

Screen enlargers, screen magniﬁers x O SW + -

Braille embossers x O HW + (-)

Screen readers x O SW + +

Speech and voice recognition x x x x I SW - +

Text-to-speech (TTS) or speech synthe-

sizers

x x x O HW/SW +/- -

Refreshable Braille displays x O HW + (-)

Keyboard ﬁlters x (x) x x I SW + -

On-screen keyboards x I SW + -

Light signaler alerts x O HW + -

TTY/TDD conversion modems x x I/O HW + (-)

Alternative keyboards (x) x (x) x x I HW + (-)

Touch screens (x) (x) (x) (x) (x) I HW + -

Other alternative input devices (e.g. elec-

tronic pointing devices, wands, sticks)

x I HW + (-)

Peripherals (e.g. micro, web cam) (x) (x) (x) (x) x I/O HW + -

Scanners (x) (x) (x) (x) (x) I HW + -

Optical character recognition (OCR) (x) (x) (x) (x) (x) I SW +/- -

Biometric identiﬁcation devices (x) (x) (x) (x) (x) I HW/SW - -

Motion capture, gesture recognition x I HW/SW - +

Text-to-sign or sign synthesis x x O SW - (-)

Solutions for people with low literacy skills are

under-represented. This is not astonishing since many

assistive technologies emerged from the “computer at

work” context in developed countries that show high

literacy rates. Furthermore the challenge for users

with low literacy skills is not the sensory access of

computers or contents. Thus, solutions for this user

category will be mainly based on assistive techniques

described below.

Assistive technologies for people with motor im-

pairment are focused on hardware based solutions for

data input.

Although it seems that assistive technologies for

people with visual impairments are well represented

in the table, solutions for people with the most se-

vere impairments, i.e. blindness and very low vision,

require a certain amount of training (screen readers,

speech recognition) or further knowledge and skills

(Braille).

Another aspect that is not explicitly shown in Ta-

ble 1, but nevertheless important refers to cost. Re-

gardless whether public access points are run by gov-

ernment agencies or non-government organizations,

cost will play an important role for the deployment of

assistive technologies in developing countries. There

will be trade-offs between the academically desired

optimal solution (making available the technologies

that are best suited for the users that use a given point

of public access) and the practically feasible solution.

An example are alternative input devices for people

with motor impairments, some of which accommo-

date very special needs. Regarding the issue of cost,

solutions will be preferred that suit more than one spe-

cial need. Software based solutions are preferable to

hardware based ones: maintenance is cheaper and of-

ten free implementations already exist, although they

not always offer all functionality of commercial solu-

tions.

To summarize the analysis of Table 1, although

there exist many different categories of assistive tech-

nologies that attend many special needs, the number

of potential solutions diminishes when we consider

ICEIS 2008 - International Conference on Enterprise Information Systems

252

our speciﬁc context. Since according to our literature

research many authors have dedicated themselves to

assistive technologies for auditory, motor and visual

impairments and since solutions for cognitiveaccessi-

bility require a research proﬁle that differs from ours

we restrict our further considerations to people with

low literacy skills. Further comparative analysis of

technologies or concrete products are out of the scope

of this paper.

4.1 Lessons Learned for People with

Low Literacy Skills

This subsection discusses assistive techniques for

users with no or low literacy skills. In the context

of this paper, assistive techniques are methods for fa-

cilitating access to web based content. In contrast to

most assistive technologies, content becomes inher-

ently more accessible when applying assistive tech-

niques. Although limiting ourselves to the problem

of literacy, we borrow from the body of methods in-

tended to beneﬁt users with other special needs. On

the other hand, our considerations can possibly con-

tribute to the areas we borrow ideas from as well.

The challenge for people with low literacy skills is

twofold, since they are often novice computer users.

Since many usability guidelines deal with the perfor-

mance of novice users, we will exclusively focus on

the literacy aspect.

According to Table 1, the set of assistive technolo-

gies adequate for people with low literacy skills com-

prises speech and voice recognition systems, text-to-

speech synthesizers, keyboard ﬁlters, alternative key-

boards, touch screens, scanners and other peripherals,

optical character recognition software and biometric

identiﬁcation devices.

Voice recognition can be used for authentication

purposes, speech recognition for text entry or naviga-

tion. Although these solutions seem the most inter-

esting for this user-category, there are some indica-

tions that suggest they are not appropriate solutions

for our context. Apart from technical problems like

background noise or the recognition of natural, free-

style speech (Deng 2004), speech recognition systems

usually initially need to be trained and adjusted to the

individual user’s characteristics. Moreover, speech

recognition in public places collides with the user’s

privacy requirements.

Alternative keyboards with spatially clearly struc-

tured areas and differently colored keys (e.g. let-

ters, numbers and “function keys” like enter, space,

backspace, etc.) beneﬁt users with low literacy as well

as users with some cognitive or visual impairments

and are a feasible solution in our context of public

access points, since the coloring can be done even

for keyboards already purchased and in use. Touch

screens diminish the attention split between screen,

keyboard and mouse, and thus not only beneﬁt users

with cognitive impairments but also users with low

literacy skills.

Scanners and OCR software could be used to re-

place manual data entry that can be found on docu-

ments like utility or telephone bills. An aspect that

cannot be neglected, however, is that these documents

could contain other data that the user might not want

to divulge.

A similar argument applies to biometric identiﬁca-

tion devices. Although the identiﬁcation or authenti-

cation processes can be simpliﬁed, the registration of

biometric characteristics like ﬁnger prints or iris scans

in a government system can create other psychologi-

cal barriers of access.

Although we could identify assistive technologies

that bring beneﬁts to users with low literacy skills as a

side effect, much more potential lies in assistive tech-

niques discussed in the remainder of this section. We

identiﬁed the following categories of assistive tech-

niques:

• Accessibility, usability and other design guide-

lines and principles: to our knowledge there ex-

ist no guidelines, recommendations or principles

explicitly tailored to the requirements of people

with low literacy skills and published by consortia

like the W3C or other organizations. Apart from

sources directly related to the subject matter (e.g.

Huenerfauth (2002) or Medhi et al. (2007)), our

ﬁndings reveal that many guidelines, recommen-

dations or principles for users with other special

needs also bring beneﬁts to users with low liter-

acy skills. Although intended for the deaf and

hard of hearing, some of the ﬁndings of Fajardo

et al. 2007 are directly related to literacy and thus

applicable to our case. The simpliﬁcation of web

page structure and content that beneﬁts users with

cognitive impairments (Sevilla et al. 2007), also

beneﬁts users with low literacy skills, since a sim-

pliﬁcation will yield web pages with less textual

information and texts presented in a language and

grammar that is easier to understand.

• Standards or recommendations: generally, all

standards or recommendations that deal with al-

ternatives to text-based or visual interfaces, text

layout, and the augmentation of textual informa-

tion by multi-media content are relevant. This

list comprises but is not limited to the W3C

activities, recommendations or candidate recom-

mendations (W3C 2007) CSS (Cascading Style

Sheets, a mechanism for separating text content

ASSISTIVE TECHNOLOGIES AND TECHNIQUES FOR WEB BASED EGOV IN DEVELOPING COUNTRIES

253

from layout), the Multimodal Interaction Activ-

ity (provides the possibility to dynamically select

the most appropriate mode of interaction), SMIL

(Synchronized Multimedia, an XML-based lan-

guage for interactive multimedia presentations),

SVG (Scalable Vector Graphics, an XML-based

language for describing 2D graphics and graph-

ical applications), or the W3C Speech Interface

Framework (including markup speciﬁcations like

VoiceXML for telephone-based interaction with

web applications). Although having great poten-

tial, new challenges arise to make applications ac-

cessible that use these techniques (Gibson 2007).

• Individual solutions: The literature that elabo-

rates individual solutions ranges from case stud-

ies that try to identify techniques by analyzing de-

sign processes or results to proposals that study

novel approaches. Akan et al. (2006) develop an

electronic screening tool for rural primary care,

Plauch´e and Prabaker (2006) a telephone based

system for market and ambient information that

can be used uttering a set of pre-deﬁned command

words. A solution for novice computer users that

also facilitates access for users with low literacy

skills is presented by Chand and Dey (2006). Al-

though these three and other solutions for users

with low literacy skills are fundamentally differ-

ent from each other (in the three examples above

simple minimalist touch screen interface vs. com-

pletely auditory interface with speech recognition

vs. creation of printed macros that are activated by

a barcode scanner), they all follow similar princi-

ples: present a minimal and simpliﬁed interface,

minimize or eliminate the necessity of reading or

entering textual data, and using alternate media

(audio, images, video) either redundantly or ex-

clusively.

Our literature study showed that a body of best

practices seems to be evolving. Auditory feedback

is considered crucial. Apart from giving explicit lin-

guistic feedback, we could draw inspiration from so-

lutions that are intended for users with visual im-

pairments (e.g. Eiriksdottir et al. 2006), techniques

that use abstract non-speech sounds to facilitate menu

navigation or similar tasks (e.g. Brewster 1998; Dicke

et al. 2007).

The principle to create simple and minimalist in-

terfaces that avoid unnecessary complexity is also

found in the literature about Universal Design and

cognitive impairments.

Another best practice that seems to be consensus

is the use of visual information like drawings, photos,

animations or videos. Although certainly very impor-

tant, the use of related techniques has to be carefully

planned since there exist no universally valid rules re-

garding which kind of visual representation is opti-

mal. Furthermore cognitive effects have to be consid-

ered when using textual, auditory and visual represen-

tations simultaneously. Leahy et al. (2003) discovered

that depending on the task complexity and the redun-

dancy of auditory, textual and visual information, it is

sometimes better to use fewer types of different me-

dia.

5 CONCLUSIONS

This paper presented an overview of ATT in the con-

text of challenges we face in developing countries

when considering interaction design for eGov. Our

overview has shown that although there are a vari-

ety of assistive technologies, there remain many gaps,

especially considering the scenario of public access

points in developing countries. We have shown that

particularly for users with low literacy skills, assistive

techniques may offer many possible solutions to be

investigated. Although there is no clear body of rules

regarding the optimal employment of the techniques,

since many of them depend on various factors, they

encourage further research. Some of these solutions

are now being investigated in the context of interac-

tion design for an eGov project in Brazil.

ACKNOWLEDGEMENTS

This work was partially supported by FUNTTEL

– Fundo para o Desenvolvimento Tecnol´ogico das

Telecomunicac¸˜oes and CAPES.

REFERENCES

Akan, K. D., Farrell, S. P., Zerull, L. M., Mahone, I. H., and

Stephanie Guerlain, S. (2006). eScreening: Devel-

oping an Electronic Screening Tool for Rural Primary

Care. In Systems and Information Engineering Design

Symposium, 2006 IEEE, pages 212–215.

Brewster, S. A. (1998). Using nonspeech sounds to provide

navigation cues. ACM Trans. Comput.-Hum. Interact.,

5(3):224–259.

Chand, A. and Dey, A. K. (2006). Jadoo: a paper user in-

terface for users unfamiliar with computers. In CHI

’06: CHI ’06 extended abstracts on Human factors in

computing systems, pages 1625–1630, New York, NY,

USA. ACM Press.

Deng, L. and Huang, X. (2004). Challenges in adopting

speech recognition. Commun. ACM, 47(1):69–75.

ICEIS 2008 - International Conference on Enterprise Information Systems

254

Dicke, C., Sodnik, J., Billinghurst, M., and Tomazic, S.

(2007). Spatial auditory interfaces compared to vi-

sual interfaces for mobile use in a driving task. In

ICEIS 2007: proceedings of the Ninth International

Conference on Enterprise Information Systems, pages

282–285, Funchal, Madeira, Portugal.

Eiriksdottir, E., Nees, M. A., Lindsay, J., and Stanley, R. M.

(2006). User preferences for auditory device-driven

menu navigation. In Proceedings of the Human Fac-

tors and Ergonomics Society 50th Annual Meeting.

Fajardo, I., Ca˜nas, J. J., Salmeran, L., and Abascal, J.

(2006). Improving deaf users’ accessibility in hyper-

text information retrieval: are graphical interfaces use-

ful for them? Behaviour and Information Technology,

25(6):455–467.

Gibson, B. (2007). Enabling an Accessible Web 2.0.

In W4A ’07: Proceedings of the 2007 international

cross-disciplinary conference on Web accessibility

(W4A), pages 1–6, New York, NY, USA. ACM.

Hornung, H. and Baranauskas, M. C. C. (2007). Interac-

tion Design in eGov systems: challenges for a de-

veloping country. In XXXIV Semin´ario Integrado de

Software e Hardware, SEMISH/CSBC2007. Anais do

XXVII Congresso da SBC, pages 2217–2231, Rio de

Janeiro, Brazil.

Huenerfauth, M. P. (2002). Developing design recommen-

dations for computer interfaces accessible to illiterate

users. Master’s thesis, University College Dublin.

IBGE (2000). Instituto Brasileiro de Geograﬁa e Es-

tat´ıstica: Censo Demogr´aﬁco (in Portuguese). Online

available: http://www.ibge.gov.br/home/presidencia/

noticias/27062003censo.shtm. Last accessed:

12/12/2007.

IBGE (2005). Instituto Brasileiro de Geograﬁa e Estat´ıstica:

Acesso `a Internet e posse de telefone m´ovel celular

para uso pessoal (in Portuguese). Online available:

http://www.ibge.gov.br/home/estatistica/populacao/

acessoainternet/default.shtm. Last accessed:

12/12/2007.

IBGE (2006). Instituto Brasileiro de Geograﬁa e Es-

tat´ıstica: Perﬁl dos Munic´ıpios Brasileiros (in Por-

tuguese). Online available: http://www.ibge.gov.br/

home/estatistica/economia/perﬁlmunic/2006/default.s

htm. Last accessed: 12/12/2007.

IPM (2005). Instituto Paulo Montenegro: In-

dicador de Alfabetismo Funcional (in Por-

tuguese). Online available: http://www.ipm.org.br/

ipmb pagina.php?mpg=4.02.00.00.00&ver=por. Last

accessed: 12/12/2007.

Kennaway, J. R., Glauert, J. R. W., and Zwitserlood, I.

(2007). Providing signed content on the Internet by

synthesized animation. ACM Trans. Comput.-Hum.

Interact., 14(3):15.

Kuber, R., Yu, W., and McAllister, G. (2007). Towards

developing assistive haptic feedback for visually im-

paired internet users. In CHI ’07: Proceedings of

the SIGCHI conference on Human factors in comput-

ing systems, pages 1525–1534, New York, NY, USA.

ACM.

Law, C., Jacko, J., Peterson, B., and Tobias, J. (2005). Uni-

versal designs versus assistive technologies: research

agendas and practical applications. In Assets ’05: Pro-

ceedings of the 7th international ACM SIGACCESS

conference on Computers and accessibility, pages 2–

3, New York, NY, USA. ACM.

Leahy, W., Chandler, P., and Sweller, J. (2003). When Au-

ditory Presentations Should and Should not be a Com-

ponent of Multimedia Instruction. Applied Cognitive

Psychology, 17(4):401–418.

Medhi, I., Prasad, A., and Toyama, K. (2007). Opti-

mal audio-visual representations for illiterate users of

computers. In WWW ’07: Proceedings of the 16th

international conference on World Wide Web, pages

873–882, New York, NY, USA. ACM.

Petrie, H., Fisher, W., Weimann, K., and Weber, G. (2004).

Augmenting icons for deaf computer users. In CHI

’04: CHI ’04 extended abstracts on Human factors in

computing systems, pages 1131–1134, New York, NY,

USA. ACM.

Pilling, D. and Boeltzig, H. (2007). Moving toward e-

government: effective strategies for increasing access

and use of the internet among non-internet users in the

U.S. and U.K. In dg.o ’07: Proceedings of the 8th an-

nual international conference on Digital government

research, pages 35–46. Digital Government Research

Center.

Plauch´e, M. and Prabaker, M. (2006). Tamil market: a spo-

ken dialog system for rural India. In CHI ’06: CHI

’06 extended abstracts on Human factors in comput-

ing systems, pages 1619–1624, New York, NY, USA.

ACM Press.

Sevilla, J., Herrera, G., Mart´ınez, B., and Alcantud, F.

(2007). Web accessibility for individuals with cogni-

tive deﬁcits: A comparative study between an existing

commercial Web and its cognitively accessible equiv-

alent. ACM Trans. Comput.-Hum. Interact., 14(3):12.

Takagi, H., Asakawa, C., Fukuda, K., and Maeda, J. (2004).

Accessibility designer: visualizing usability for the

blind. SIGACCESS Access. Comput., (77-78):177–

184.

UNDP (2007). United Nations Development Programme:

Human Development Report 2007/2008. Online avail-

able: http://hdr.undp.org/en/reports/global/hdr2007-

2008/. Last accessed 12/12/2007.

U.S. Department of Health and Human Services (2007).

Research-based web design and usability guidelines –

Glossary. Online available: http://www.usability.gov/

guidelines/new/dictionary/index.html. Last accessed

12/12/2007.

W3C (2007). Homepage of the World Wide Web Consor-

tium (W3C). Online available: http://www.w3.org/.

Last accessed 12/12/2007.

Wobbrock, J. O. and Gajos, K. Z. (2007). A comparison

of area pointing and goal crossing for people with and

without motor impairments. In Assets ’07: Proceed-

ings of the 9th international ACM SIGACCESS con-

ference on Computers and accessibility, pages 3–10,

New York, NY, USA. ACM.

ASSISTIVE TECHNOLOGIES AND TECHNIQUES FOR WEB BASED EGOV IN DEVELOPING COUNTRIES

255