Exploring Twofold Adaptability to the User Interface

Different Users and Multiple Devices

Ig Ibert Bittencourt

1,2

, Maria Cecília Baranauskas

, Diego Dermeval

, Roberto Pereira

and Juliana Braga

Center of Excellence in Social Technologies, Computing Institute, Federal University of Alagoas, Maceió, Brazil

Computing Institute, University of Campinas, São Paulo, Campinas, Brazil

Federal University of Campina Grande, Paraíba, Campina Grande, Brazil

Federal University of ABC, São Paulo, Santo André, Brazil

Keywords: Adaptable User Interfaces, Inclusiveness, Diversity, Multiple Devices.

Abstract: The Web movement beyond desktop to different devices amplified the possibilities to ensure access to in-

formation for all. Nevertheless, the consideration of multiple devices and different user conditions for the

applications brings complexity to the design and development processes. On the one hand, some existing

tools take into account the design and development of user interfaces according to the target device, but do

not consider the diversity in terms of end user conditions (e.g., abilities, preferences, culture, limitations,

education). On the other hand, some systems provide the adjustment to users with special needs, but do not

adapt to multiple devices. In this paper, we present results of a systematic review on literature to build a

roadmap on the twofold adaptability of user interfaces: for different users and multiple devices. The results

point out a growing demand for solutions that consider adaptation to both: different users and devices;

moreover a distribution of the works per type of adaptation mechanisms found suggest aspects still to be

covered in further research in the field.

1 INTRODUCTION

The design and development of Web applications

portable to different devices have opened up a huge

set of possibilities for uncountable domains, such as

social systems, educational systems, literacy deve-

lopment, digital and social inclusion, and so on.

Indeed, several organizations around the world have

undertaken work programs to establish electronic

communication via Web technologies. As a result,

such technologies are improving the potential of

access to knowledge everywhere and at any time,

and are becoming a way for tackling the challenge of

providing a participative and universal access to

knowledge. The universal access has been consid-

ered one of the grand challenges of several Interna-

tional Communities around the world (Hoare and

Miller, 2004); (Medeiros, 2008). This challenge is

about the use of technologies to ensure the access to

knowledge in a participative and personalized way

for the citizen, taking into account people´s diversity

and, consequently, different users’ needs encom-

passing disability issues as well as social problems

(e.g., people living in underserved communities).

Although Web-based solutions represent a way

to reach a wide audience, for different reasons, such

technologies are not yet reaching everybody. On the

one hand, the available development tools take into

account the design of user interfaces according to

the device target, but do not consider the features of

the end users, especially those with special needs

(e.g., Haxe (http://haxe.org), Sencha

(http://sencha.com), Foundation Framework

(http://foundation.zurb.com). On the other hand,

some systems provide the adjustment to users with

special needs, but do not cover the different users’

needs and multiple devices. In addition, Web envi-

ronments are even more complex because they are

24/7 worldwide available environments, demand

robustness and are constantly changing require-

ments. As a result, there is a need for systems that

can be generalized for multiple devices and, at the

same time, specialized for users with different needs,

as presented in Figure 1. The main features to be

considered during the adaptation of an interface are:

i) device type which defines the features and re-

striction of each device, such as size, hardware, lan-

guage, and others; ii) adaptation type according to

Bittencourt I., Baranauskas M., Dermeval D., Pereira R. and Braga J..

Exploring Twofold Adaptability to the User Interface - Different Users and Multiple Devices.

DOI: 10.5220/0004418700380045

In Proceedings of the 15th International Conference on Enterprise Information Systems (ICEIS-2013), pages 38-45

ISBN: 978-989-8565-61-7

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

the hypermedia adaptive theory (Brusilovsky, 1996)

and software requirements needs; iii) adaptation

time, which represents the moment the adaptation

should occur; iv) adaptation techniques, which could

be represented as the computational techniques used

to adapt the interface or to design/create adaptable

interfaces; v) goals, which describe the domain goal;

and vi) users’ needs, which represent the particulari-

ties of each user intended to adapt the application.

Figure 1: Interface Adaptation mechanisms.

For this reason, several research and industrial

studies have been proposed to support the design of

adaptive systems that can change aspects of their

structure, functionalities, and/or interface in order to

accommodate the different needs of individuals or

groups of users and the changing needs of users over

time (Benyon and Murray, 1993).

As far as we know, no systematic review was

conducted to provide an overview on the design and

development of systems that adapt the user interface

based on the different users needs and for multiple

devices. The goal of this paper is to present results

of a systematic review on literature to build a

roadmap on the design of adaptable user interfaces

for different users and multiple devices.

2 THE METHOD

The research in this work was undertaken as a sys-

tematic literature review (SLR) to provide a repeata-

ble and formal process for documenting relevant

papers about the design of adaptable user interfaces

for different users and multiple devices.

According to (Kitchenham, 2004), a systematic

review is composed by three phases (planning, con-

ducting, and reporting) divided in several steps,

which are: 1) Planning the Review (Identification of

the need for a systematic review; Development of a

review protocol); 2) Conducting the Review (Identi-

fication of research; Selection of the studies; Study

Quality assessment; Data extraction and monitoring;

Data analysis; Data synthesis); 3) Reporting the

review (Report-writing).

2.1 Review Questions

Although many research and industrial studies have

been proposed to design interfaces for multiple de-

vices — such as (Calvary et al., 2003); (Gajos et al.,

2010); (Falb et al., 2009) — no systematic review

has been conducted to provide an overview on the

design and development of inclusive environments

which adapt the user interface based on the different

users’ need and for multiple devices.

As the goal of this systematic review was to

gather the knowledge about the design of inclusive

environments for different users and multiple devic-

es focusing on adaptation capabilities, the high-level

question of this study was:

How researchers are designing adaptable user interfaces

for different users and multiple devices?

This high-level question provides a starting point for

understanding how designers conduct the interface

adaptation taking into account the final user. Based

on this research question, two other more specific

questions were raised. The questions and their moti-

vations are described in Table 1.

Table 1: Research Questions and motivations.

Research Question Motivation

RQ1. Which Interface Design

approaches and support tools

are being used to adapt user

interfaces for multiple devices

and users with different

conditions?

The answer to this question is

important to understand how

people with different condi-

tions are being considered in

the interface design process

and the different solutions for

building applications adaptable

for all.

RQ2. Which mechanisms are

being used to adapt user

interfaces for different users

and multiple devices?

The answer to this question is

important to identify the dif-

ferent aspects for adapting an

interface to the user (as pre-

sented in Figure 1).

2.2 Sources and Search Selection

Criteria

The first step in performing the review was to define

the search selection criteria. Due to the fact that this

review has several sources to consider, two kinds of

search strategies were considered (i.e., the automatic

and the manual search).

The automatic search was done according to the

specification of the search terms (i.e., search string).

Although the automatic search covers a huge range

of relevant papers, it is also important to search on

specific and specialized sources to improve the co-

verage. For this reason, a manual search on some of

ExploringTwofoldAdaptabilitytotheUserInterface-DifferentUsersandMultipleDevices

the most important conferences and journals of HCI

area was conducted.

Hereafter, the search terms definition and the

digital libraries (DLs) selection regarding the auto-

matic search are explained. Based on the research

questions, a set of relevant terms was defined, such

as: cross-device, disabilities, underserved communi-

ties, interface design approaches, multiple devices,

different users, tools, and inclusive. After that, such

terms were categorized and their related terms were

identified. The terms were identified based on: i) the

expertise of the authors; ii) the analysis of terms

present in a HCI systematic review (Almeida and

Baranauskas, 2012); and iii) the TagCloud for HCI

presented in

(Buchdid and Baranauskas, 2012). By

contrast, the set of digital libraries was defined ac-

cording to the most popular and traditional DLs: ISI

Web of Science, Scopus, ACM Digital Library,

IEEE Xplore and ScienceDirect. The SpringerLink

digital library was excluded due to search re-

strictions. After the definition of the relevant terms

and DLs, the search string for automatic search on

the mentioned digital libraries was built as follows:

((tool OR environment OR framework OR au-

thoring OR architecture OR software OR ambi-

ent OR "reference model") AND

(inclusive OR inclusiveness OR inclusivity

OR "inclusive web" OR "inclusive social web"

OR accessibility OR disability OR disabili-

ties OR assistive OR underserved OR "margin-

alized communities" OR "design for all" OR

"universal access" OR "universal design" OR

"designing for diversity" OR "design for

diversity" OR "design diversity" OR diversi-

ty) AND

("Multiple Device" OR "Cross-device" OR

"Multimodal" OR migration OR "different de-

vices" OR "device-independent" OR "migratory

interfaces" OR "distributed interfaces" OR

"plastic user interfaces" OR "flexible user

interfaces" OR "flexible interfaces" OR

"distributed user interfaces" OR portability

OR "portable web applications" OR "portable

systems" OR "information interoperability"

OR "knowledge interoperability") AND

("Interaction design" OR "adaptable inter-

face" OR "adaptable user interfaces" OR "in-

teraction resources" OR "responsive web de-

sign" OR "universal design" OR "inclusive

design" OR "process model" OR "adaptable

model" OR "meta-design" OR "meta design" OR

metadesign OR "participative design"))

Moreover, in order to perform the manual search,

two relevant conferences and journals on Human-

Computer Interaction area were considered, as de-

picted in Table 2. Although this is only part of rele-

vant vehicles, this limitation is due to lack of access

to the library of some journals, such as the Interna-

tional Journal of Human-Computer Interaction. Re-

garding the conferences, some would desbalance the

study; for example, the ACM SIGCHI Conference

on Human Factor in Computing Systems would

represent half of the whole search space.

Table 2: Relevant sources on HCI considered in the

search.

Journals

1. International Journal of Human Computer Studies

2. Interacting with Computers

Conferences

1. IFIP INTERACT

2. Cross-Disciplinary Conference on Web Accessibility

2.3 Inclusion and Exclusion Criteria

The aim of defining a criterion is to identify the

primary studies that provide direct evidence about

the research questions and also to reduce the likeli-

hood of bias (Kitchenham, 2004). Regarding the

inclusion criteria, articles written in the last ten years

related to any of the research questions were consid-

ered. The exclusion criteria involve papers not relat-

ed to the research questions, papers that were not

written in English, short papers, duplicate studies

and papers before 2002. The summarized inclusion

and exclusion criteria are presented in Table 3.

Table 3: Inclusion and exclusion criteria.

Inclusion criteria

Peer-reviewed studies that answers to the research questions

Studies that focus on design approaches and tools to adapt

interfaces for multiple devices

Studies that focus on design approaches and tools to adapt

interfaces for different users’ needs

Studies published from 2002

Exclusion criteria

Short-papers

Non peer-reviewed studies

Studies that are not related to the research questions

Studies that do not consider user interface adaptation

Duplicated studies

Papers not written in English

Studies before 2002

2.4 Data Extraction

After the definition of the search and the selection

processes, a data extraction process was performed

by reading the abstract and full-text screening each

one of the selected papers. It is important to note that

this clustering is based on Figure 1. In order to guide

this data extraction, the data collection from

Biolchini et al., (2005) was adapted as follows:

ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems

─ Paper Information: Study Reference (ID);

Source; Year; Source Type (Journal or Confer-

ence); Affiliations; Authors list; Paper Title;

Google Scholar Citation;

─ Context (Industry and Academia);

─ Adaptation Mechanisms:

 Where – Device Types (Desktop, Web, Tablet,

TV, Mobile Phones, PDA, Tabletop, Braille

Notes);

 Why – Adaptation Goals: domain-dependent

or domain-independent;

 To Whom – Users’ Needs (Target Audience):

Blind/Visual Impairment; Deaf/Hearing Im-

pairment; Motor/ Mental; Underserved people;

Elder;

 What – Adaptation Type: Content presenta-

tion; Navigation support; Screen structure; Sys-

tem requirements;

 When – Adaptation Time: Design time; Use

Time;

 How (1) – Interface Design Approach: User-

Centered; Task-Centered; Participatory; Sce-

nario-Based; Ethnographic Methods; Design

per Target; Model-based; Automatically Gen-

erated; Multi-tier; Universal Design; User Sen-

sitive Inclusive Design;

 How (2) – Tool Technology: API; Design Pat-

tern; Framework; Platform; Software Product

Line; Authoring; MDA; Reference Model;

Middleware; Architecture;

 How (3) Adaptation Process – Manual; Hy-

brid (User and Software Adaptation).

─ Study Type (Controlled experiments; Quasi-

experiments; Case Study; Survey; Ethnography;

Action Research);

3 OVERVIEW OF THE

INCLUDED STUDIES

This section presents the included studies according

to the automatic and manual search (see Figure 2).

Firstly, the automatic search was conducted to each

digital library. Then, an iterative process was applied

to exclude the not relevant papers based on the ex-

clusion criteria. The exclusion criteria were applied

according to the analysis of the abstract, full-text

screening and, finally, the duplicate papers. A simi-

lar process was applied to the manual search. At the

end, only 2.19% of the selected papers were consi-

dered relevant to this systematic review. The list of

the 89 studies is available at www.nees.com.br/iceis.

Figure 2: Search process and selected studies.

The automatic and manual queries were conduct-

ed in the period between December 4

(2012) and

January 11

(2013). The manual search indicated the

W4A Conference (55%; 6 studies) as the largest

vehicle of relevant studies, and the IFIP Interact

(9%, 1 study) as the smallest vehicle of relevant

studies. However, other vehicles were identified as

relevant when the automatic search was applied,

such as the International Conference on Computers

Helping People With Special Needs (7.69%; 6 stud-

ies), Interacting with Computers (6.41%; 5 studies),

ACM SIGACCESS (6.41%; 5 studies), International

Conference on Universal Access in Human-

Computer Interaction (5%; 4 studies), and ACM

SIGCHI (3.85%; 3 studies).

From a temporal point of view, an increasing

number of publications in the context of this review

is noticed since 2005 (see Figure 3). It is important

to note also that 2012 is the year with more publica-

tions, which shows a demand for solutions to pro-

vide adaptable user interfaces for multiple devices

and users with different needs.

Figure 3: Distribution of publications per year.

As this review has started on December 2012, it

would be expected a decreasing in 2012 publications

because some papers might be under publication

processes. Nevertheless, we can see, in general, an

increase in the number of publications (based on the

linear progression). Despite this increase, only 2

studies have more than 60 citations (see Table 4);

they are depicted in Table 5.

The significant increase in publications reflects

the need for convergence of technologies and, at the

same time, the importance of deploying inclusive

solutions.

ExploringTwofoldAdaptabilitytotheUserInterface-DifferentUsersandMultipleDevices

Table 4: Publications per number (nº) of citations.

nº < 20 20 <= nº < 40 40 <= nº < 60 Nº >= 60

77 6 4 2

Table 5: List of studies with more than 60 citations.

Study (ID) Title Citations

SSD05

A Unifying Reference Frame-

work for multi-target user

interfaces

491

SACM64

Automatically Generating User

Interfaces Adapted to Users’

Motor And Vision Capabilities

The study holds contributions from 29 countries

located in all the continents. Although all the conti-

nents are represented by the included papers, there is

a concentration in the Europe (67.62%) and Ameri-

can Continent (20.00%). Table 6 presents the publi-

cations per country. Furthermore, according to the

distribution of the included papers, most of the relat-

ed studies were published in conferences (66.29%),

while (33.71%) are published in journals, consider-

ing the automatic and manual search.

4 RESULTS

As described in Section 2, three research questions

have driven this systematic review. Based on the

research questions, the string search was built and

the type of data extraction defined. Figure 4 presents

the type of empirical study of each included study.

More than 50% of the works report a case study as

empirical evaluation. Less than 20% of the works

report some kind of experiment. By contrast, survey

studies were 15%.

Table 6: Publications per country.

Country Works Total (%)

United Kingdom 13 12.38%

United States of America 12 11.43%

Spain 11 10.48%

Brazil 7 6.67%

Portugal 6 5.71%

France 5 4.76%

Germany 5 4.76%

Italy 4 3.81%

Greece 4 3.81%

Finland 4 3.81%

Sweden 4 3.81%

Others 30 28.58%

Figure 4: Distribution of empirical study type.

4.1 Adaptable User Interfaces

According to the data extraction, most users’ condi-

tions addressed were Blind or Visual Impairment

(26.45%; 41 studies), Motor (16.13%; 25 studies)

and Mental (15.48%; 24 studies). It is worth noting

that although blind and underserved are, respective-

ly, the more and the less discussed conditions, the

other disabilities are well balanced (see Figure 5).

Figure 5: Distribution of the studies per target audience.

Table 7: Distribution per device type.

Device type Works Total(%)

Mobile Phones 46 34.06%

Web 25 18.12%

Desktop 18 13.04%

PDA 18 13.04%

Tablet 13 9.42%

TV 7 5.07%

Others 10 7.25%

Table 7 presents the distribution of the papers

according to the device type. Most of the solutions

were proposed to mobile phones (34.06%; 46 stud-

ies) and Web (18.12%; 25 studies). It is worth noting

that the number of solutions for tablets is still low

(9.42%; 13 studies); it may happen especially be-

cause tablets have become popular more recently

than mobile phones. On the other hand, although the

number of studies for TV is not high (5.07%; 7 stud-

ies), one would expect its increasing over time due

to new platforms for SmartTV.

Table 8 presents the distribution of work per

support tool type. Frameworks, platforms, and refe-

ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems

rence models are in the top of the list, while some

approaches are not mentioned. In Table 9, it is pos-

sible to see that the “Automatically Generated” is

the most common ID Approach.

Table 8: Distribution per support tool type.

Tool type Works Total(%)

Framework 24 26.97%

Reference Model 15 16.85%

Platform 15 16.85%

Application 14 15.73%

Architecture 6 6.74%

Authoring 2 2.25%

API 2 2.25%

Design Pattern 1 1.12%

MDA 0 0.00%

Middleware 0 0.00%

Software Product Line 0 0.00%

Table 9: Distribution per Interface Design Approach.

ID Approach Works Total(%)

Automatically Generated 18 23.68%

User-Centred 15 19.74%

Participatory 13 17.11%

User-Sensitive Inclusive Design 8 10.53%

Task-Centered 5 6.58%

Design Per Target Device 5 6.58%

Model-Based 4 5.36%

Ethnographic 3 3.95%

Others 5 6.58%

Table 10: List of studies that propose support tools and/or

applications for users with different conditions and

multiple devices.

Study Paper Title Tool Type

SACM27

MyUI: Generating Accessible User Interfaces

from Multimodal Design Patterns

Framework

SACM64

Automatically Generating User Interfaces

Adapted to Users’ Motor And Vision Capabili-

ties

Platform

SACM68

Accessibility of Dynamic Adaptive Web TV

Applications

Framework

SIEEE11 A Framework for Designing Flexible Systems Framework

SIEEE30

i*Chameleon: A Unified Web Service Frame-

work for Integrating Multimodal Interaction

Devices

Framework

SSCOPUS38

A Novel Design Approach for: Multi-device

Adaptable User Interfaces: Concepts, Methods

and Examples

Framework

SSCOPUS80

Assistive smartphone for people with special

needs : The Personal Social Assistant

Application

SSCOPUS85

Attuning speech-enabled interfaces to user and

context for inclusive design: technology,

methodology and practice

Application

In contrast, only 18.18% (8 studies) among the

selected propose adaptable user interfaces tools for

multiple devices. Table 10 depicts these works by

presenting their study reference, paper title and the

tool type used.

4.2 Discussion

The twofold adaptability meant in this work refers to

the user interface; i.e. we were interested in the way

the adaptation occurs in the tools that proposed a

solution to different users and multiple devices. In

general, only a very limited number of papers (8

studies, as presented in Table 10) was identified

from the whole list of included studies (see Figure

2), representing 8.98%. It is worth noting that all

these 8 papers were published in the last 5 years; at

the same time 2012 has the highest number of stud-

ies (37.5%). This information demonstrates the cur-

rent (and growing) interest of the HCI community

with the development of solutions capable of adapt-

ing users’ interfaces for multiple devices and differ-

ent users conditions.

In order to answer the research questions raised

in this work, Figure 6 presents the distribution of

papers according to the following six dimensions

considering interface adaptation: (i) “where?”; (ii)

“why”; (iii) “what?”; (iv) “how?”; (v) “when?”; and

(vi) “to whom?”.

The “how?” dimension may be especially helpful

to answer the RQ1 (Which Interface Design ap-

proaches and Tools are being used to adapt user

interfaces for multiple devices and users with differ-

ent conditions?). The distribution on how the studies

provide interfaces adaptation is characterized by

interface design approaches (1), tool type (2) and

adaptation process (3). Hereafter, the results consid-

ering the interface design approaches and tool types

categories are discussed; the adaptation process – as

well the other dimensions – will be discussed under

the perspective of RQ2.

Thus, Figure 6 shows that automatic generated

approach represents only 1/3 of the total, indicating

that most of the studies presented manual adaptation

(in this review, 62.5%). Conversely, the participa-

tory and universal design approaches are representa-

tive, which could indicate the application of more

than one approach to enhance the adaptation.

Regarding the support tool type, more than a half

of the studies used framework as software design

approach. On the one hand, this is expected because

frameworks may support the extensibility of soft-

ware which can be used to adapt different configura-

tions of users’ interfaces. On the other hand, the use

ExploringTwofoldAdaptabilitytotheUserInterface-DifferentUsersandMultipleDevices

of frameworks shows the adaptation is done only by

designers/developers not the end users. It is im-

portant to note that none of the studies presented

software product lines, model-driven architecture or

middleware as support tool type. This absence may

suggest some kind of detachment between the HCI

researchers (related to adaptive user interfaces for

multi-users and multiple devices) and Software En-

gineering, since these approaches are some of the

best practices of software engineering research re-

garding software reuse, flexibility, and dynamic

evolution.

Aiming at answering the RQ2 (Which mecha-

nisms are being used to adapt user interfaces for

different users and multiple devices?), the adaptation

process – “How?”(3) – applied in the studies need to

be analysed. As Figure 6 shows, 62.50% of the stud-

ies present manual adaptation; whilst 37.50% pre-

sent hybrid (automatic and manual) solutions. The

interface adaptation technique is very close to the

adaptation time (“when?”), i.e., all studies that apply

manual adaptation also provide adaptation at design

time; in contrast, all hybrid adaptation studies pre-

sent adaptation both at design time and runtime.

Moreover, it is worth noting that the number of stud-

ies that uses frameworks is very similar to the num-

ber of studies providing adaptation at design time;

this information may suggest that the studies that

present framework also implement it at design time.

Alternatively, it can also be noted a similar distribu-

tion in the automatically generated interface design

approach studies. In fact, it is usual that studies that

support adaptation at runtime also provide automatic

generation of users’ interfaces.

Additionally, the outcomes show a regular distri-

bution in the number of studies regarding the target

audience (“to whom?”). The impairments disabilities

have similar percentages (with slightly higher to

blind/visual impairments). At the same time, the

elder audience is somewhat attended by the studies

(only 7.41%) and the underserved audience is not

addressed at all by the identified studies.

Considering the devices types (“where?”) ad-

dressed in the studies, the appearance of mobile

phones is highlighted. The desktop and PDA devices

are also representative in the studies. A smaller

amount of works focusing on TV or Web (only

11.76%) is observed, while none of the 8 studies

took into account the tablets.

The analysis of the studies found three different

adaptation types (“what?”) for users’ interfaces: i) in

the content presentation (e.g. font size and colour),

ii) in the screen structure (e.g. buttons position ac-

cording to screen density) and iii) in the system re-

quirements (e.g. check appointment). Figure 6 shows

that the number of studies found by each adaptation

type are similar, with a slightly decrease in screen

structure. Such distribution shows the importance

and relevance of the different kinds of adaptation.

Finally, it can be noted a concern (“why?”) of the

researchers to provide domain-independent tools

(75%) in designing users’ interfaces for multiple

devices and users.

4.3 Threats to Validity

This section discusses the threats to validity that

might have affected the results of this systematic

review. The review protocol was validated to ensure

that the research was as correct, complete and objec-

tive as possible. However, possible limitations in the

publication selection and in data extraction of the

process may have occurred.

The search for publications was performed in

two major steps: (i) automatic search and (ii) manual

search. In the step (i), there is a limitation because

the search string could not be used in SpringerLink

library, which possibly leads to a reduction in the

considered studies. In the step (ii), there is a limita-

tion concerning the papers included in the review.

The manual searching was only performed in a lim-

ited set of journals and conferences. Although it was

expected that relevant studies published in other

journals or conferences would be captured through

the automatic search realized in the previous step, it

cannot be guaranteed that all related papers pub-

lished are included in this systematic review.

5 CONCLUSIONS

The design and development of Web applications for

different devices have opened up a huge set of pos-

sibilities for uncountable domains, such as social

systems, educational systems, literacy development,

and so on. This paper presented a systematic review

to raise the adaptation mechanisms being considered

in the development of portable systems. Thus, the

study involved data extraction in order to answer the

two research questions. A systematic review proto-

col was defined and the search returned 4061 studies

undertaken between 2002 and 2012. After applying

the exclusion/inclusion criteria, it leads to the inclu-

sion of 89 studies in the review. After that, only 8

studies fulfilled the defined requirements, which

indicate a demand for more research results in the

context of the questions addressed in this paper.

ICEIS2013-15thInternationalConferenceonEnterpriseInformationSystems

Figure 6: Distribution of studies by interface adaptation mechanisms.

The results indicate that: i) there is a growth in

interest of the HCI community in the development

of solutions capable of adapting users’ interfaces for

multiple devices and diverse users; ii) the tools are

commonly directed to the technical developers; iii)

none of the studies presented a solution to under-

served audience; iv) there is a communication gap

between the Software Engineering and the HCI

communities; in the same way, software engineering

approaches need to consider inclusiveness as an

important non-functional requirement; v) there is a

trend in adapting interfaces to mobile devices; and

vi) the researchers are considering both adaptation in

the hypermedia and in the systems requirements.

As further work, we expected to: i) diminish the

threats to validity in this study; ii) extend the study

to consider more papers of relevant authors identi-

fied in this systematic review; and iii) analyse the

studies according to each type of user condition.

ACKNOWLEDGEMENTS

This work has been supported by the Brazilian Insti-

tutions: “Conselho Nacional de Desenvolvimento

Científico e Tecnológico (CNPq)”, processes:

150113/2013-7 and 560044/2010-0 (ecoWeb Pro-

ject).

REFERENCES

Almeida, L. D. A. & Baranauskas, M. C. C., 2012. Acces-

sibility in Rich Internet Applications: People and Re-

search. In IHC’12 Proceedings of the 11th Brazilian

Symposium on Human Factors in Computing Systems.

pp. 3–12.

Benyon, D. & Murray, D., 1993. Adaptive systems: from

intelligent tutoring to autonomous agents. Knowledge-

Based Systems, 6(4), pp.197–219.

Biolchini, J., Mian, P. G., Candida, A., Natali, C., 2005.

Systematic Review in Software Engineering. Springer

Berlin Heidelberg. V 679. N. May. pp, 165–176.

Buchdid, S. B.; Baranauskas, M. C. C, 2012. IHC Em

Contexto: O que as palavras revelam sobre ela. In

IHC'12, Brazilian Symposium on Human Factors in

Computing Systems pp. 199-208.

Brusilovsky, P., 1996. Methods and techniques of adaptive

hypermedia. User Modeling and User-Adapted Inter-

action, 6(2-3), pp.87–129.

Calvary, G., Coutaz, J., Thevenin, D., Limbourg, Q.,

Bouillon, L., Vanderdonckt, J. 2003. A Unifying Ref-

erence Framework for multi-target user interfaces. In-

teracting with Computers, 15(3), pp.289–308.

Falb, J., Popp, R., Rock, T., Jelinek, H., Arnautovic, E.,

Kaindl, H. 2009. Fully-automatic generation of user

interfaces for multiple devices from a high-level mod-

el based on communicative acts. International Journal

of Web Engineering and Technology. 5(2). pp 135-

161.

Gajos, K. Z., Weld, D. S. & Wobbrock, J. O., 2010. Au-

tomatically generating personalized user interfaces

with Supple. Artificial Intelligence, 174(12-13),

pp.910–950.

Hoare, T. & Miller, R., 2004 (Editors). Grand Challenges

in Computing Research Workshop. British Computer

Society.

Medeiros, C. B., 2008. Grand research challenges in com-

puter science in Brazil. Computer. 41(6) pp, 59-65.

Kitchenham, B., 2004. Procedures for Performing Sys-

tematic Reviews. Keele UK Keele University,

33(TR/SE-0401), pp.28.

ExploringTwofoldAdaptabilitytotheUserInterface-DifferentUsersandMultipleDevices