Structuring Guidelines for Web Application Designers

A Meta-model

Anh Do Tuan

, Isabelle Comyn-Wattiau

1,2

and Samira Si-saïd Cherfi

CEDRIC-CNAM, Paris, France

ESSEC Business School, Cergy-Pontoise, France

Keywords: Web Application Design, Guideline, Meta-model, Quality Characteristic, Knowledge Capitalization.

Abstract: Companies develop and maintain complex web sites. Literature provides them with many guidelines for these

tasks. However this knowledge is disseminated in many information sources and difficult to apply. This paper

is an attempt to address the following research question: How to structure the existing guidelines helping

website designers in order to facilitate their application? The contribution is twofold: i) we propose a meta-

model allowing a rich representation of these guidelines, ii) we feed this model with several hundreds of

guidelines thanks to a deep extraction and structuration. Future research will consist in enriching the UWE

(UML-based Web Engineering) method with this knowledge base leading to a quality based approach.

1 INTRODUCTION

Companies develop and maintain complex web sites

that allow them to communicate easily and

dynamically with their customers, suppliers, partners,

etc. In 2008, according to Krigsman, 24% web

projects fail to be delivered within budget and 5%

were unable to confirm the final cost of their web

development project. Moreover, 21% fail to meet

stakeholder requirements and nearly a third of web

based projects (31%) were not delivered within the

agreed timescales (Krigsman, 2008). More recently, a

research, conducted by McKinsey and the University

of Oxford on more than 5400 IT projects, concluded

that 45% of large projects are over budget, 7% are

over time and 56% delivered less value than predicted

(Bloch, 2013). The reasons vary: unclear objectives,

lack of business focus (missing focus), shifting

requirements, technical complexity (content issues),

unaligned team, lack of skills (skill issues), unrealistic

schedule, reactive planning (execution issues)

(Bloch, 2013), inconsistent stakeholder demands, and

insufficient time or budget (Krigsman, 2008).

Web sites and web applications are in fact

software applications. In this sense, the classical

application methodologies may be used manually or

with the help of computer aided software engineering

(CASE) tools. However, the very specific nature of

these applications led to the proposition of more

dedicated approaches. Indeed, during the two last

decades, research in Web Engineering brought a rich

contribution composed of methods and techniques to

support Web applications development. These

methods such as UWE, WebML, or others are

generally founded on a model-driven development

paradigm, and provide models and transformation

rules to handle several web applications' aspects such

as data, navigation, interaction, and presentation.

However and despite the research and the tooling

efforts, very few developers adopt these methods and

many continue to apply ad-hoc practices.

The main reason is that these approaches suffer

from lack of guidance. Even if web application

designers refer to these approaches, they do not have

sufficient knowledge and help in implementing them

efficiently. As a consequence, the resulting

applications are neither user-friendly nor easy to

maintain.

We argue that the current approaches are well

structured. However they need to be enriched with

guidelines helping designers in the numerous

decisions they have to make during the web

application development. Therefore, we have

collected the different sets of guidelines proposed in

the literature and organized them along different

dimensions. In particular, this structure allows us to

link the guidelines with the quality objectives

(maintainability, performance, functionality,

security, etc.) and with the relevant steps of the web

application design (content design, navigation design

and presentation design).

Tuan, A., Comyn-Wattiau, I. and Cherﬁ, S.

Structuring Guidelines for Web Application Designers - A Meta-model.

In Proceedings of the 18th International Conference on Enterprise Information Systems (ICEIS 2016) - Volume 1, pages 327-335

ISBN: 978-989-758-187-8

327

This article is organized as follows. Section 2

describes how we collected and selected the

guidelines, and a short experiment we conducted on

how methods and guidelines are followed in websites

construction. Based on the survey conclusions,

Section 3 motivates and describes our research

question. Section 4 describes the meta-model we

propose in order to represent the guidelines in a useful

way. Section 5 analyses the set of resulting

guidelines. Section 6 is dedicated to related works on

guidelines. Finally, the last section concludes and

sketches future research directions.

2 AN EXPERIMENT ON

METHODS AND GUIDELINES

USAGE

Before defining the research question, we performed

a quick inventory on how well web design best

practices and guidelines are followed by existing

websites. The objective was i) to analyze whether

existing practices and guides are used and ii) identify

how to facilitate their adoption and hence avoid ad

hoc approaches. Thus, we first collected 475

guidelines from several sources and confronted them

with three websites: the web site of our university

department (deptinfo.cnam.fr), the website of a

French newspaper (lemonde.fr) and a well-known e-

commerce web site (amazon.fr). We first describe

briefly the collected guidelines and then their

verification on the three websites.

2.1 Collecting the Guidelines

World Wide Web Consortium (W3C) is the main

international standards organization for the World

Wide Web. This consortium puts together around 400

organizations. They developed Web Content

Accessibility Guidelines (WCAG) with the goal of

proposing a single shared standard for web content

accessibility that meets the needs of individuals,

organizations, and governments (Web Accessibility

Initiative). Two versions of WCAG were published

until now. The first one was introduced in 1999. It

contains 14 large guidelines. Each main guideline is

composed of atomic guidelines addressing the same

topic. The second version was published in 2008. It

contains 12 guidelines organized into four categories,

targeting four desirable characteristics of websites:

perceivable, operable, understandable, and robust.

WCAG defines three levels of conformance,

respectively A, AA and AAA. Some of the related

guidelines could be automatically checked whereas

others require manual checking. Authors in (Trulock,

2008) conducted a case study on Irish websites

showing that web designers are aware of web

accessibility but they concentrate their efforts on

ensuring validation of automatically controlled

checkpoints and ignore those requiring additional

manual testing.

The guidelines of WCAG focus only on

accessibility. Thus, we collected other guidelines

which address all the characteristics of web site

quality. The literature contains guidelines for specific

web sites (for children for instance) as well as rules

available for all sites.

2.1.1 Identifying the Relevant Sources

For collecting guidelines from literature effectively,

we use some keywords when searching, such as

“website guideline”, “guideline for website”,

“guideline security web application” in title and

content of document, from main electronic libraries

and databases in computer science: IEEE Xplore,

Springer, ScienceDirect, ACM, and DBLP. As an

example, based on the keywords “web” and

“guideline”, we have 1273 results from IEEE, 273

results from ScienceDirect and 168 results from

DBLP. With Springer and ACM, we have much more

results in many domains, so we had to refine the

results and choose results with high relevance (as

computed by the search engines). Then we defined

inclusion criteria for selecting sources (primary

studies) and rejecting the other ones. The inclusion

criteria are presented in the table below (Table 1).

Table 1: Inclusion criteria.

Criterion Description

C1 The study focuses on guideline definition

for web sites

C2 The study mentions quality characteristics

of web sites

C3 The paper is recent, i.e. published since

2000

C4 The paper proposes original guidelines

(does not only mention guidelines from

other studies)

We found several guideline lists published since

2000. However, these documents are sparse and

address many domains. One objective is to gather

them, categorize, and model guidelines. Thus they

will be more usable for supporting web application

developers. Some guidelines are general and others

are dedicated to specific domains: education,

international, or for particular ages (children or

ICEIS 2016 - 18th International Conference on Enterprise Information Systems

328

seniors). As an illustration, the guidelines of

AgeLight Company (AgeLight LCC, 2001) are

divided in six categories: layout and style, color, text,

general usability testing, accessibility and disabilities,

user customization. Web sites for old people are the

research object of a number of studies (Xie, 2011)

(Sun, 2010). Meloncon et al., in contrast,

concentrated on guidelines for children (Meloncon et

al., 2010). Maguire focused on e-commerce

international sites (Maguire, 2011). Some papers

focused on the characteristics of quality directly, such

as (Chiuchi, 2011) which targeted portability and

efficiency. (Radosav, 2011) capitalizes on the 14

guidelines from WCAG, so we did not collect them.

Finally, we took into account fourteen sources. Their

analysis is described below.

2.1.2 Extracting the Appropriate Guidelines

Our systematic search followed by a scan of sources

allowed us to exhibit fourteen papers containing

relevant guidelines. The next step consisted in

studying all the guidelines and selecting the helpful

guidelines. In each source of guidelines, we found

some obsolete guidelines or some recommendations

which were out of our scope. For example, in (U.S.

Department of Health and Human Services, 2006),

guidelines in the last part (part 18), such as “Use an

iterative design approach” or “Solicit test

participants’ comments” were not selected, since they

are too general or dedicated to testing. So we

eliminated them from the list.

We found 14 sources with 475 guidelines (an

excerpt is listed at Annex 1). The number of

guidelines of each source is presented in Table 2. In

some cases, we split some guidelines, hence the

number of selected guidelines may be higher than the

number of guidelines proposed in the paper.

Some sources propose general guidelines. Others

are more specific. For example (Bargas-Avila, 2010)

concentrates on web forms or (Chiuchi, 2011) focuses

on portability and efficiency.

Some guidelines are too complex, so we had to

divide them into two parts or more. For example the

guideline for images in (Chiuchi, 2011) is separated

into two atomic guidelines: “The preferred use of

JPEG and GIF images” and “The resolution of image

should be set correctly inside the tags”.

2.2 Analyzing the Guidelines Usage

To analyze how well the guidelines are applied in

practice, we defined four levels namely: Yes, No,

Partial and NN. Yes means that the site satisfies

completely the guideline, No means that this site does

not satisfy it, Partial means that this site partially

meets the guideline and NN means that “We don’t

know”, since either the guideline cannot be applied to

the site or we don’t have enough information. The

result is synthesized at Figure 1. Each guideline

obtains the grade 1, 0.5, 0 point for Yes, Partial and

No respectively. After applying all guidelines to the

three websites, each guideline obtains a grade

between 3 and zero or is equal to NN. Thus, 206

guidelines are verified on the three selected sites

(totalizing 3 points). 33 guidelines reach 2.5, 46

guidelines obtain 2 points. 60 guidelines obtain

between 0.5 and 1.5. 47 guidelines obtain 0, meaning

that they are not respected on the three selected sites.

Table 2: Source, number and scope of guidelines.

Source Proposed

guidelines

Selected

guidelines

Scope

(AgeLight LCC, 2001) 53 35 General

(Bargas-Avila, 2010) 20 20 General but concentrating on web forms

(Xie, 2011) 7 10 Old people / medical information

(Chiuchi, 2011) 17 15 General / focusing portability and efficiency

(Carnegie Mellon University) 7 8 University

(U.S. Department of Health and

Human Services, 2006)

196 209 General

(Leuthold, 2008) 9 9 Blind people

(Lokman, 2009) 13 14 General

(Maguire, 2011) 20 8 International site

(Meloncon, 2010) 21 11 Children

(Microsoft Developer Network) 50 49 General

(Ministry of Community and

Social Services of Ontario, 2012)

11 11 General

(Ozok, 2004) 20 20 General

(Sun, 2010) 31 31 Old people

Structuring Guidelines for Web Application Designers - A Meta-model

329

But let us remind that 3 guidelines are dedicated to

international or children sites, and thus are not

required in the three tested web sites. Besides them,

there are 83 guidelines obtaining the NN value. For

guidelines which have NN value, many of them are

related to the security aspects. To check if they are

fulfilled, we require the admin authority, so we

cannot conclude about these guidelines.

Figure 1: The distribution of guideline grades.

As an illustration, the guideline G115

“considering both levels: ’high’ and ’low’ of cultural

context for satisfying both viewpoints” or G176

“Limit navigational topics” are not relevant for the

three web sites. Others may be irrelevant, such as

G217 “Inform users of long download times” or G247

“Limit homepage length” since we had high speed

connection for our tests.

Figure 2: Results for the three websites.

Figure 2 compares the scores obtained by the three

websites if we consider the rule: the more guidelines

the web site complies with, the better score it obtains.

deptinfo.cnam.fr obtains the score of 264.5 while

lemonde.fr obtains 287 points. Finally, amazon.fr is

the best one with score of 300.5. However,

deptinfo.cnam.fr has 124 NN guidelines, whereas

lemonde.fr has only 97 NN guidelines and amazon.fr

has 92 NN guidelines, so if we compare ratios,

deptinfo.cnam.fr achieves 75.3%, lemonde.fr 75.9%

and amazon.fr is the highest with 78.5%.

These figures show that either these guidelines are

not considered as references or these websites still

face quality problems. As an example, let us mention

G345 “Provide auto-tabbing functionality” for

increasing users’ convenience and G362 “Using

photographs of people” for increasing users’

reliability. The three websites are not aligned with

these two guidelines. That means that these

guidelines which were validated through complex

processes are not sufficiently known by web site

designers.

3 RESEARCH QUESTION

From the mid of 1990s, methods and approaches have

been created for helping developers to build web

applications more easily and constructively. The

Object‐Oriented Hypermedia Design Method

(OOHDM) was one of the first methods proposing a

rigorous process from requirements elicitation to

implementation including navigational and interface

design (Schwabe, 1995). The method relies on object-

oriented principle and proposes notation mainly

derived from UML. The transition from models to

specification is not supported and thus requires a

considerable effort.

The Web Modelling Language (WebML) is a

model driven web engineering method dedicated to

data-intensive web applications (Ceri, 2000).

WebML is one of the most used web engineering

methodologies. It is supported by a development

framework, Ratio5 (Acerbis, 2005) that is fully

integrated to the Eclipse framework. Several

extensions of the first version have been proposed

offering a rich modelling approach for developers.

However, the method relying very few on standards,

it led to a proliferation of proprietary notations

increasing the method complexity.

The UML‐based Web Engineering (UWE)

methodology (Hennicker, 2000) is a model‐driven

Web Engineering approach. It relies heavily on UML

and is extensively related to standards. The model

driven orientation allows generating platform specific

implementation through dedicated transformation

rules. Model driven approaches are based on four

levels of abstraction: the computer independent

model (CIM), the platform independent model (PIM),

the platform specific model (PSM), and the code.

Some methods address only the CIM level, other

methods focus on PIM level. In the same way, some

100

150

200

250

3 2,5 2 1,5 1 0,5 0 NN

Number of guidelines

Grades obtained by the guidelines

73,5

74,5

75,5

76,5

77,5

78,5

deptinfo.cnam.fr lemonde.fr amazon.fr

Ratio of conformance for the three websites

ICEIS 2016 - 18th International Conference on Enterprise Information Systems

330

methods deal with the transformation of CIM to PIM

(e.g. NDT, OOWS), others address the

transformation of PIM to PSM (e.g. WebML, UWE)

and others incorporate the transformation of PSM to

code (e.g. OOHDM, UWE) (Aragon, 2013). Even if

these methods offer a real support, they are still not

used by practitioners probably since they are complex

and they do not provide designers with sufficient

guidance.

We argue that most methods do not provide their

users with sufficient guidance in the design and

development process. Either in the same approaches

or in other sources, researchers propose many

guidelines in order to help designers and developers.

These guidelines may be very helpful to support

them.

Thus the research question we address in this

paper may be defined as follows: “How to structure

all the existing guidelines helping website designers

to understand and apply them?” To answer this

question the experiment presented in Section 2 helped

us to elicit the main characteristics of these

guidelines. We then defined a meta-model allowing

us to represent this knowledge. Finally we

categorized the selected guidelines based on our

meta-model. This categorization aims to facilitate

their reuse.

4 GUIDELINE CAPITALIZATION:

A MODEL-BASED APPROACH

In the literature, we find different ways to describe

guidelines: in (Chiuchi, 2011), they are represented

by three attributes: Category, Name and Content.

Meanwhile in (Ekberg, 2010) a guideline has three

parts: design/application solutions, objective and

description. We argue that this descriptive

information is not sufficient to facilitate the reuse of

guidelines by web application designers. In

particular, the latter must find easily the guidelines

using different criteria. For example, in case of

designing a web application for blind people: which

recommendations do they have to take into account?

If developers want mainly to facilitate the

maintainability of the web application: which

guidelines aim at this objective? Etc.

We first propose a model helping capitalizing and

structuring the guidelines. The meta-model is

depicted at Figure 3.

Following the general description of patterns for

decision processes (Harrison, 2007), we propose to

link each guideline with the following categories:

- The source where the guideline was found,

- The quality characteristics and sub-

characteristics that the guideline addresses,

Figure 3: The meta-model of guidelines.

Structuring Guidelines for Web Application Designers - A Meta-model

331

- The problem it aims to solve,

- The solution proposed,

- The particular domain concerned if any,

- The lifecycle aspect, meaning which web

application model (content model, navigation

model, presentation model) it deals with.

This structure will constitute a knowledge base for

automatic reuse through a web application design

tool. The meta-model is represented as a UML class

diagram at Figure 3. The related to relation between

guidelines allows us to represent potential links

between guidelines. Thus the attribute type of link

may take the values “in contradiction with”,

“specializes” or “similar to”.

Each guideline solves a problem; however several

guidelines may tackle the same problem. The solution

of the guideline describes the rules to be applied. As

explained above, in our process, we split some

guidelines such that each resulting guideline

recommends one and only one solution. The domain

may be general or it may be a specific one. The

quality characteristics (functional suitability,

performance/efficiency, compatibility, usability,

reliability, security, maintainability, portability) and

sub-characteristics refer to ISO25010 for software

quality. For space reasons we may not list all of them.

Some guidelines are common to several sources,

hence the multiplicity of the relation here is many-to-

many. Finally, the lifecycle aspect consists of three

elements: Content, Navigation, and Presentation.

In order to illustrate, let us describe the guideline

G37: “For body copy, the recommended faces for the

web, in order of preference, are Verdana, Arial and

Helvetica. The browser should use Verdana first; if it

is not available, use Arial and then Helvetica. If none

are available, use another Sans serif font”.

Number: #37

Content: For body copy, the recommended faces

for the web, in order of preference, are Verdana, Arial

and Helvetica. The browser should use Verdana first;

if it is not available, use Arial and then Helvetica. If

none are available, use another Sans serif font.

Problem: Choosing appropriate font for a website

Domain: web for university (even if it can also

apply to other types of site)

Lifecycle aspect: Presentation

Quality Sub-characteristics: User interface

aesthetics

Quality Characteristic: Usability

Solution: Choose Sans serif font, namely

Verdana, Arial and Helvetica.

Source: (Carnegie Mellon University)

5 GUIDELINES ANALYSIS

In this section, we provide the reader with an analysis

of the guidelines according to the different

dimensions of our meta-model. Let us remind that our

selection process led to the constitution of a set of 475

guidelines (the guidelines can be found at

http://deptinfo.cnam.fr/~wattiaui/Guidelines.html).

If we analyze them from the lifecycle dimension

(Content/ Navigation/ Presentation), we counted 203

guidelines for Presentation, 291 guidelines for

Content and only 40 guidelines for Navigation. Some

guidelines address more than one model. Hence the

total exceeds 475 (Figure 4).

The 475 guidelines were mapped with quality

sub-characteristics. Some guidelines are mapped with

several sub-characteristics. The characteristic

Usability, with sub-characteristics Operability and

User interface aesthetics is the most involved one. It

is easy to explain since many papers address interface

aspects (User interface aesthetics) and aim to build

easy-to-use interfaces (Operability).

Many guidelines are about font (G37, G42, G49,

G50, etc.) and color (G6, G8, G39, G41, G86, G185,

G186, etc.) of websites. White is the color which is

not recommended (G9, G39, G189, etc.).

We can detect some contradictory guidelines,

since some guidelines aim at different goals. In the

guidelines of a university (Carnegie Mellon

University) the documents should be opened in new

windows (G35), probably for legal responsibilities. It

is opposite to guideline G101 (AgeLight LCC, 2001)

which recommends not to open external links in new

windows, since it can cause user distracting.

Guideline G37 recommends using only Sans-serif

font, but meanwhile G85 accepts serif font in web site

for printing.

Some guidelines are dedicated to different types

of users, but finally they have same contents. As an

illustration, Sun et al. (Sun, 2010) focused on website

for old people; meanwhile Meloncon et al.

(Meloncon, 2010) concentrated on web applications

for children. Old people and children are two types of

users which have some specific characteristics in

comparison with others (e.g. not being able to

understand complex content).

The guideline about Security of web applications

in MSDN of Microsoft (Microsoft Developer

Network) contains about 50 sections. Many of them

address Integrity (prevent unauthorized access) (in 38

sections) and Confidentiality (data are accessible only

to those authorized) (in 18 sections). This is due to the

fact that Integrity and Confidentiality are important

for web applications which are designed for many

ICEIS 2016 - 18th International Conference on Enterprise Information Systems

332

Figure 4: Percentage of guidelines per lifecycle aspect.

kinds of users and also are the targets of attacks.

Among the eight quality characteristics,

Compatibility is not mentioned at all, since guidelines

focus on the site itself, and not on the relation of the

site with other sites or other applications (scope of

Compatibility).

6 RELATED WORKS

In this section, we synthesize the literature on

guidelines for web site design. We organized this

state of the art in two categories: first the approaches

which propose guidelines and, second, the

approaches which involve such guidelines.

6.1 Design for Guidelines

One of the most famous works delivering guidelines

for web site design is Web Accessibility Initiative

(WAI) of W3C (WAI, 2008). It is a collection of

standards, guidelines, and techniques for making

accessible products in four categories: websites,

authoring tools, browsers, and web applications. Each

category has a bunch of guidelines for constructing

web design and for improving accessibility. Other

sources of guidelines were listed in the paper and

enrich considerably the W3C recommendations.

Khlaisang is an example of research illustrating

how these guidelines may be either validated or

elicited (Khlaisang, 2015). The author developed user

interface guidelines and a prototype for evaluating

educational service websites. Based on source sites of

Thailand Cyber University Project (TCU), he studied

the use of sites, the website structure, the user

interface design and conducted usability tests of the

site. Resulting from these experiments, he presented

a model of suitable website for TCU service. Starting

from this website, model, he designed and developed

a prototype of site. The paper also mentions similar

approaches.

6.2 Design by Guidelines

Besides works creating guidelines, other works used

existing guidelines for proposing ways to improve

quality of websites.

Leuthold et al. (Leuthold et al., 2008) designed

enhanced text user interfaces for blind Internet users.

Starting from the guidelines of web content

accessibility guidelines (WCAG), they proposed

enhanced text user interface (ETI) helping blind users

in spending less time to complete tasks, making fewer

mistakes and expressing greater satisfaction when

surfing the website. This system contains nine

guidelines. For blind users, this system is more usable

than normal GUI.

Another work building on WCAG guidelines is

(Sloan, 2006). Using e-learning as an example, they

propose a framework that guides web authors and

policy makers in addressing accessibility at a higher

level, by defining the context in which a Web

resource will be used and considering how new

alternatives may be combined to enhance the

accessibility of the web site.

After a brief description of the 14 guidelines of

WCAG (version 1), Radosav et al. discussed the

choice of colours for adjusted web design (Radosav,

2011). They classified colour into several groups and

concluded that colours, which cannot be

differentiated by people with colour discrimination

disability, should not be placed next to each other.

For space reasons, we cannot provide a more

detailed literature review. As a conclusion, research

in this field is prolific and aims at i) proposing

guidelines for web site designers, ii) enriching

existing ones, iii) implementing guidelines into more

comprehensives approaches, iv) evaluating guidelines

through experiments. To the best of our knowledge, we

did not find any paper proposing a meta-model

allowing us to put together the different guidelines as a

first step for their reuse in an automatic way.

7 CONCLUSION AND FUTURE

RESEARCH

The companies grasp the importance of having usable

and efficient web applications. Thus, their

development and maintenance is of high importance.

The academic literature on the subject contains

hundreds of guidelines aiming at helping web site

291

203

Percentage of guidelines per lifecycle aspect

Content

Presentation

Navigation

Structuring Guidelines for Web Application Designers - A Meta-model

333

designers. The research question we addressed in this

paper may be expressed as follows: How to structure

the existing guidelines helping website designers in

order to facilitate their application? As a first

contribution, we defined a meta-model allowing us to

describe each guideline with six dimensions: the

problem it addresses, the solution it proposes, the

lifecycle aspect it deals with, the target quality

characteristics, the source it comes from, the potential

links (similarity, contradiction, specialization) with

other guidelines. Our search and selection process

allowed us to define 475 such guidelines and to feed

our meta-model with them. This required the

mapping of them with the relevant quality sub-

characteristics. As a first evaluation of these

guidelines, we checked whether they were compliant

with three very different web sites.

This research suffers from some limitations. Thus,

it is rather easy to check the contradiction between

guidelines attached to the same quality characteristics

and/or sub characteristics. However, contradictions

may also occur between guidelines associated with

different quality characteristics. Moreover, some

guidelines may become obsolete due to new technical

opportunities. It is not easy to ensure an easy update

of guidelines.

Future research will explore three directions: first

the definition of a grammar for expressing problem

and solution components of guidelines; second, the

implementation of these guidelines in a CASE tool

implementing UWE web application design method;

third, a validation of the approach through an

experiment with web site designers, in order to

evaluate how the guidelines help them when using the

CASE tool.

REFERENCES

Acerbis, R., Bongio, A., Brambilla, M., & Butti, S. (2007).

Webratio 5: An eclipse-based case tool for engineering

web applications. In Web Engineering (pp. 501-505).

Springer Berlin Heidelberg.

AgeLight LCC, 2001. Interface design guidelines for users

of all ages. Technical report, 2001.

Aragon, G., Escalona, M.J., Lang, M., Hilera J., 2013. An

Analysis of Model-Driven Web Engineering

Methodologies. International Journal of Innovative

Computing, Information and Control Volume 9,

Number 1, January 2013.

Bargas-Avila, J.A., Brenzikofer, O., Roth, S.P., Tuch, A.N.,

Orsini, S., Opwis, K., 2010. Simple but Crucial User

Interfaces in the World Wide Web: Introducing 20

Guidelines for Usable Web Form Design. In User

Interfaces, Rita Matrai (Ed.), InTech, 2010.

Bloch, M., Blumberg, S., Laartz, J., 2013. Delivering

large-scale IT projects on time, on budget, and on

value. McKinsey on Finance Number 45, Winter 2013,

pp 28 – 35.

Carnegie Mellon University. Web guidelines. http://www.

cmu.edu/marcom/brand-guidelines/print-web-products

/web/index.html.

Ceri, S., Fraternali, P., & Bongio, A. (2000). Web Modeling

Language (WebML): a modeling language for designing

Web sites. Computer Networks, 33(1), 137-157.

Chiuchi, C.A., de Souza, R.C.G., Santos, A.B., Valêncio,

C.R., 2011. Efficiency and portability: guidelines to

develop websites. Software Engineering and Knowledge

Engineering 2011: Miami Beach, USA, pp. 37 – 41.

Ekberg, J., Ericson, L., Timpka, T., Eriksson, H., Nordfeldt,

S., Hanberger, L., Ludvigsson, J., 2010. Web 2.0 Systems

Supporting Childhood Chronic Disease Management:

Design Guidelines Based on Information Behaviour and

Social Learning Theories. Journal of Medical Systems

April 2010, Volume 34, Issue 2, pp 107-117.

Harrison, N.B., Avgeriou, P., Zdun, U., 2007. Using

Patterns to Capture Architectural Decisions. IEEE

Software, Vol 24(4), 2007, pp. 38-45.

Hennicker, R., & Koch, N. (2000). A UML-based

methodology for hypermedia design. In

≪≫

UML

2000—The Unified Modeling Language (pp. 410-424).

Springer Berlin Heidelberg.

Khlaisang, J., 2015. Research-based Guidelines for

Evaluating Educational Service Website: Case Study of

Thailand Cyber University Project. Procedia - Social

and Behavioral Sciences, Volume 174, February 2015,

pp. 751-758.

Krigsman, M., 2008. Research: 25 percent of web projects

fail. http://www.zdnet.com/article/research-25-percent-

of-web-projects-fail/, May 2008.

Leuthold, S., Bargas-Avila, J.A., Opwis, K., 2008. Beyond

web content accessibility guidelines: Design of

enhanced text user interfaces for blind internet users.

International Journal of Human Computer Studies,

Volume 66, Number 4, April 2008, pp. 257-270.

Lokman, A.M., Noor, N.L., Nagamachi, M., 2009.

ExpertKanseiWeb: A Tool to Design Kansei Website.

Enterprise Information Systems, 11th International

Con-ference, ICEIS 2009, Milan, Italy, May 6-10,

2009: pp. 894-905.

Microsoft Developer Network. Chapter 4 - design guidelines

for secure web applications. https://msdn.microsoft.com/

en-us/library/ff648647.aspx .

Ministry of Community and Social Services of Ontario,

2012. Making your website more accessible. Queen's

Printer for Ontario, 2012. http://www.mcss.gov.on.ca/en/

mcss/publications/accessON/accessible

_websites/toc.aspx

Ozok, A., Salvendy, G., 2004. Twenty guidelines for the

design of Web-based interfaces with consistent

language. Computers in Human Behavior. Vol. 20(2),

149-161 (2004).

Maguire, M., 2011. Guidelines on Website Design and

Colour Selection for International Acceptance. HCI

International 2011, pp. 162–171, 2011.

Meloncon, L., Haynes, E., Varelmann, M., Groh, L., 2010.

ICEIS 2016 - 18th International Conference on Enterprise Information Systems

334

Building a Playground: General Guidelines for

Creating Educational Web Sites for Children.

Technical Communication, 57(4):398–415, 2010.

Radosav, D., Karuovic, D., Markoski, B., Ivankovic, Z.,

2011. Guidelines on accessible web portal design. 2011

IEEE 12th International Symposium on Computational

Intelligence and Informatics (CINTI), 21-22 Nov.

2011, pp. 297 – 302.

Sloan, D., Heath, A., Hamilton, F., Kelly, B., Petrie, H.,

Phipps, L., 2006. Contextual web accessibility,

maximizing the benefit of accessibility guidelines.

Proceeding W4A '06 Proceedings of the 2006

international cross-disciplinary workshop on Web

accessibility (W4A): Building the mobile web:

rediscovering accessibility? Pages 121-131.

Sun, Z., Zhao, Y., 2010. The preliminary construction of

accessibility design guidelines of learning website for

old people. International Workshop on Education

Technology and Computer Science, pages 612–615, 2010.

Schwabe, D., & Rossi, G. (1995). The object-oriented

hypermedia design model. Communications of the

ACM, 38(8), 45-46.

Trulock, V., & Hetherington, R. (2008). Assessing the

Progress of Implementing Web Accessibility-An Irish

Case Study. In ICEIS (5) (pp. 105-111).

U.S. Department of Health and Human Services, U.S.

General Services Administration, 2006. Research-

Based Web Design & Usability Guidelines.

Web Accessibility Initiative (WAI), 2008. Web Content

Accessibility Guidelines (WCAG) Overview. http://

www. w3.org/WAI/intro/wcag.

Xie, B., Watkins, I., Huang, M., 2011. Making web-based

multimedia health tutorials senior-friendly: design and

training guidelines. iConference 2011: Seattle,

Washington, USA, February 8-11, 2011, pp. 230-237.

APPENDIX

Annex 1: Excerpt from the set of guidelines.

Num

be r

Content S ub-characteri stics S ource

Related to

other

guidelines

Life cycle S pecific

Left just ified t ext, t ext line should

not be long

User interface aethestics

[8]

Similar to

G70

Present ation

Old people

Support users flexible op erations

(adjustable font size, color

conversion)

Accessibility

[8] Presentation

Old people

7 Ensure links change color after visit

Operability

User error protection

[8]

Similar to

G63

Present ation

Old people

20 Provide a site-map Operability

[8] Navigation

Old people

Search engine should have to check

and correct misspelled function

Operability

[8] Content

Old people

Consider p age download speed -

create 'small' p ages

Time behaviour

[8] Content

Old people

31 Do not require 'double clicks' Accessibility

[8]

Presentation Old people

All images should be JPGs, GIFs or

PNGs. JPGs are used for photos.

Grap hics should use GIF or PNG

formats

Functional

approp riateness

[15] Content

University

Images have a resolution of 72 dpi

and are in either RGB or indexed

color modes

Adap tability

[15] Content

University

Links should be relevant text. Do

not link words like "here" "this

page" etc.

Operability

[15] Content

University

Main background color should be

brown, not light blue User interface aethestics

[17] Presentation

Including hyp erlinks within longer

pages so viewers can “jump” with a

single click

Accessibility

[6] Navigation

Should not use exceptionally bright,

fluorescence or vibrant colors

User interface aethestics

[6] Presentation

Archive old art icles, while

maintaining the actual page URL

Functional

completeness

[6] Content

Trying to link to sites at the highest

possible level, in the case "p age not

found"

Fault tolerance

[6] Navigation

103

If using tables, provide an alternate

text-only version of page

Rep laceability

Fault tolerance

[6] Content

106

Site should have multi language

versions

Operability

[10] Content

International site

110

Getting the spelling right for the

correct market

Functional correctness

[10] Content

International site

117

Should build auto resp onse service

informing that t hey will receive a

full reply within 24 or 48 hours

Availabilit y

[10] Content

International site

191

Use concrete words, active verbs,

and concise sentence structure Learnability [9]

Similar to

G25

Content

Children sit e

246

Ensure the homep age looks like a

homep age

Approp riateness

recognizability

[13]

Content

Present ation

Structuring Guidelines for Web Application Designers - A Meta-model

335