Web information system quality

According to the descriptions given above, we can say that the foundation for both length and complexity measures is “solution-orientated,” whereas the foundation for a functionality siz

Web Information Systems Quality

Coral Calero Muñoz

University of Castilla-La Mancha, Spain

Ma Ángeles Moraga

University of Castilla-La Mancha, Spain

Mario Piattini

University of Castilla-La Mancha, Spain

Hershey • New York

InformatIon scIence reference

Typesetter: Michael Brehm

Published in the United States of America by

Information Science Reference (an imprint of IGI Global)

701 E Chocolate Avenue, Suite 200

Hershey PA 17033

Tel: 717-533-8845

Fax: 717-533-8661

E-mail: cust@igi-global.com

Web site: http://www.igi-global.com

and in the United Kingdom by

Information Science Reference (an imprint of IGI Global)

Web site: http://www.eurospanonline.com

Copyright © 2008 by IGI Global All rights reserved No part of this publication may be reproduced, stored or distributed in any form or by any means, electronic or mechanical, including photocopying, without written permission from the publisher.

Product or company names used in this set are for identification purposes only Inclusion of the names of the products or companies does not indicate a claim of ownership by IGI Global of the trademark or registered trademark.

Library of Congress Cataloging-in-Publication Data

p cm.

Summary: "This book integrates invaluable research on the models, measures, and methodologies of Web information systems, software quality, and Web engineering into one practical guide to Web information systems quality, making this handbook of research an essential addition to all library collections" Provided by publisher.

Includes bibliographical references and index.

ISBN 978-1-59904-847-5 (hardcover) ISBN 978-1-59904-848-2 (ebook)

1 World Wide Web Handbooks, manuals, etc 2 Information technology Handbooks, manuals, etc 3 Computer software Quality control Handbooks, manuals, etc 4 Application software Development Handbooks, manuals, etc 5 Web services Handbooks,

TK5105.888.H362 2008

004.67'8 dc22

2007032032

British Cataloguing in Publication Data

A Cataloguing in Publication record for this book is available from the British Library.

All work contributed to this book set is original material The views expressed in this book are those of the authors, but not necessarily of the publisher.

If a library purchased a print copy of this publication, please go to http://www.igi-global.com/reference/assets/IGR-eAccess-agreement.pdf for information on activating the library's complimentary electronic access to this publication.

Silvia Abrahão

Valencia University of Technology, Spain

Manuel Ángel Serrano

University of Castilla-La Mancha, Spain

University of New South Wales, Australia

Eduardo Fernández Medina

University of Castilla-La Mancha, Spain

Carlos García

Indra Software Factory, Spain

Ignacio García-Rodríguez de Guzmán

University of Castilla-La Mancha, Spain

Preface xv Acknowledgment xix

Section I Effort and Quality Assessment Chapter I

Sizing Web Applications for Web Effort Estimation 1 Emilia Mendes, The University of Auckland, New Zealand

Patterns for Improving the Pragmatic Quality of Web Information Systems 57

Pankaj Kamthan, Concordia University, Canada

Chapter IV

Evaluation of the Effectiveness of Small and Medium Sized Businesses Web Sites

in a Business to Business Context 71

Rosemary Stockdale, Massey University, New Zealand

Chapter V

Anomaly Detection and Quality Evaluation of Web Applications 86

May Haydar, Université de Montréal, Canada

Chapter VII

A General View of Quality Models for Web Portals and a Particularization to

E-Banking Domain 113

Mª Ángeles Moraga, University of Castilla—La Mancha, Spain

Chapter VIII

A Data Quality Model for Web Portals 130

Angélica Caro, University of Bio Bio, Chile

Section II Accessibility and Usability

Chapter IX

Specification of the Context of Use for the Development of Web-Based Applications 146

Marta Fernández De Arriba, University of Oviedo, Spain

Chapter X

Web Accessibility 163

Carlos García Moreno, Indra, Spain

Chapter XI

Comparing Approaches to Web Accessibility Assessment 181

Adriana Martín, Universidad Nacional del Comahue, Argentina

Chapter XII

Maximizing Web Accessibility Through User-Centered Interface Design 206

Soonhwa Seok, The University of Kansas, USA

Pascual González, University of Castilla-La Mancha, Spain

Chapter XIV

The Usability Dimension in the Development of Web Applications 234

Maristella Matera, Politecnico di Milano, Italy

Chapter XV

Handling Usability Aspects for the Construction of Business Process Driven

Web Applications 250

Victoria Torres, Technical University of Valencia, Spain

Joan Fons, Technical University of Valencia, Spain

Section III Metadata, MDE, Metamodels, and Ontologies

Chapter XVI

New Approaches to Portletization of Web Applications 270 Fernando Bellas, University of A Coruña, Galicia

Chapter XVII

Towards the Adaptive Web Using Metadata Evolution 286 Nicolas Guelfi, University of Luxembourg, Luxembourg

Chapter XVIII

Looking for Information in Fuzzy Relational Databases Accessible via Web 301 Carmen Martínez-Cruz, University of Jaén, Spain

Geraldo Xexéo, COPPE—Federal University of Rio de Janeiro, Brazil

Chapter XX

Towards Quality Web Information Systems Through Precise Model-Driven Development 344

Fernando Molina, University of Murcia, Spain

Chapter XXI

The Use of Metamodels in Web Requirements to Assure the Consistence 363

M J Escalona, University of Seville, Spain

G Aragón, Everis, Spain

Chapter XXII

A Quality-Aware Engineering Process for Web Applications 378

Cristina Cachero Castro, Universidad de Alicante, Spain

Chapter XXIII

Restrictive Methods and Meta Methods for Thematically Focused Web Exploration 405

Sergej Sizov, University of Koblenz-Landau, Germany

Chapter XXIV

WSRP-O: An Ontology to Model WSRP Compliant Portlets 424

Mª Ángeles Moraga, University of Castilla-La Mancha, Spain

Chapter XXV

Philosophy of Architecture Design in Web Information Systems 443

Tony C Shan, Bank of America, USA

Improving the Quality of Web Search 463

Mohamed Salah Hamdi, University of Qatar, Qatar

Chapter XXVII

The Perspectives of Improving Web Search Engine Quality 481 Jengchung V Chen, National Cheng Kung University, Taiwan

Chapter XXVIII

Web Search Engine Architectures and their Performance Analysis 491

Xiannong Meng, Bucknell University, USA

Chapter XXIX

Towards a Model for Evaluating Web Retrieval Systems in Non-English Queries 510

Fotis Lazarinis, University of Sunderland, UK

Chapter XXX

Web Information Resources Vis-à-Vis Traditional Information Services 528 John D’Ambra, The University of New South Wales, Australia

About the Contributors 541 Index 553

Preface xv Acknowledgment xix

Section I Effort and Quality Assessment

This section is related to effort and quality assessment and is composed of eight chapters The first two chapters deal with Web development effort estimation The other six are related to several aspects of Web quality such as context of use, pragmatic quality, effectiveness of small and medium size business Web sites, anomaly detection, and quality evaluation and assessment Also, two chapters are included where quality models for Web portals and data portal quality are presented.

Chapter II

Web Development Effort Estimation: An Empirical Analysis 26 Emilia Mendes, University of Auckland, New Zealand

The objective of this chapter is to introduce the concepts related to Web effort estimation and effort estimation techniques It also details and compares, by means of a case study, three effort estimation techniques, chosen for this chapter because they have been to date the ones mostly used for Web effort estimation: multivariate regression, case-based reasoning, and classification and regression trees

The chapter emphasizes the significance of approaching Web information systems (WIS) from an gineering viewpoint A methodology for deploying patterns as means for improving the quality of WIS

en-as perceived by their stakeholders is presented

Chapter IV

Evaluation of the Effectiveness of Small and Medium Sized Businesses Web Sites

in a Business to Business Context 71

Rosemary Stockdale, Massey University, New Zealand

This chapter discusses and analyses the effectiveness of SME business to business Web sites from a user perspective, under the premise that an effective method of evaluating a Web site can contribute to the development of more quality Web sites and greater realization of benefits

Chapter V

Anomaly Detection and Quality Evaluation of Web Applications 86

May Haydar, Université de Montréal, Canada

This chapter addresses the problem of Web application quality assessment from two perspectives First,

it shows the use of model checking of properties formulated in LTL to detect anomalies in Web tions Second, the chapter explains how probabilistic models (Bayesian Networks) can be built and used

applica-to evaluate quality characteristics The two proposed approaches are evaluated and a discussion on how they complement each other is presented

Mª Ángeles Moraga, University of Castilla—La Mancha, Spain

In this chapter, several portal quality models are presented and compared Authors have adapted one of the best portal quality models proposed in the literature to the e-banking context In addition, the new e-banking portal quality model has been compared with the original portal quality model, as well as with the main portal quality characteristics

Chapter VIII

A Data Quality Model for Web Portals 130

Angélica Caro, University of Bio Bio, Chile

The chapter proposes a model for data quality in Web portals (PDQM) built on the foundation of three key aspects: (1) a set of Web data quality attributes identified in the literature in this area, (2) data qual-ity expectations of data consumers on the Internet, and (3) the functionalities that a Web portal may offer its users

Section II Accessibility and Usability

This section is divided into two main topics The first chapter works on both topics The next three ters deal with accessibility, one of them from a general point of view, another one comparing approaches

chap-to Web accessibility assessment, and the last one about maximizing Web accessibility The other three chapters are about usability from the point of view of ergonomic criteria as part of the development

of Web applications or as an important aspect for the construction of business process driven Web plications.

ap-Chapter IX

Specification of the Context of Use for the Development of Web-Based Applications 146

Marta Fernández De Arriba, University of Oviedo, Spain

This chapter is presented in the structure of an index, which serves as support that allows the development team to create the specification of the context of use document for the development of Web applications, bearing in mind characteristics of usability and accessibility

This chapter studies the Web accessibility issue from the perspective of Web information systems quality

In addition, the closed relation between accessibility and standard Web technologies is explained

Chapter XI

Comparing Approaches to Web Accessibility Assessment 181

Adriana Martín, Universidad Nacional del Comahue, Argentina

In this chapter the importance of Web accessibility assessment is discussed and 15 different approaches found in literature are compared

Chapter XII

Maximizing Web Accessibility Through User-Centered Interface Design 206

Soonhwa Seok, The University of Kansas, USA

The user interface is the place where users can interact with the information by using their minds ers with special needs can acquire information by using a human centered user interface This chapter highlights the need to investigate the relationship between cognition and user interface

Us-Chapter XIII

Usability-Oriented Quality Model Based on Ergonomic Criteria 220

Francisco Montero, University of Castilla-La Mancha, Spain

This chapter proposes a quality model that focuses on quality in use or usability for the product characterization of the World Wide Web

Chapter XIV

The Usability Dimension in the Development of Web Applications 234

Maristella Matera, Politecnico di Milano, Italy

This chapter surveys the most emergent usability evaluation models to be adopted during the whole lifecycle of Web information systems for promoting usability For each evaluation method, the main features, as well as the emerging advantages and drawbacks, are illustrated

Victoria Torres, Technical University of Valencia, Spain

Joan Fons, Technical University of Valencia, Spain

In this chapter, the authors gather a set of guidelines provided by experts in Web usability and present the solution designed in a particular Web engineering method that follows a model driven development approach

Section III Metadata, MDE, Metamodels, and Ontologies

Section III is related to Metadata, MDE, metamodels, and ontologies The first four chapters are focused

on Metadata issues In concrete, the former presents a proposal annotated-based to portletizing existing Web application, the second one uses metada evolution for adaptive Web, and the last two are related

to information The next chapter presents a proposal for developing quality Web information systems through precise model driven development In the next three chapters, different metamodels oriented to Web requirements, development of Web applications, and Web exploration are shown Finally, an ontology for WSRP standard is presented and a philosophy of architecture design in Web Information Systems.

Chapter XVI

New Approaches to Portletization of Web Applications 266 Fernando Bellas, University of A Coruña, Galicia

This chapter focuses on “portletizing” existing Web applications, that is, wrapping them as portlets, without requiring any modification After providing some background on portlet technology, they discuss two kinds of approaches to portletization: automatic and annotation-based

Chapter XVII

Towards the Adaptive Web Using Metadata Evolution 286 Nicolas Guelfi, University of Luxembourg, Luxembourg

The authors survey techniques for ontology evolution The authors detail the various existing languages and techniques devoted to Web data evolution, with particular attention to Semantic Web concepts, and

Looking for Information in Fuzzy Relational Databases Accessible via Web 301 Carmen Martínez-Cruz, University of Jaén, Spain

In this chapter, the importance of using ontologies to represent database schemas is highlighted The representation of the fuzzy data in fuzzy databases management systems (FDBMS) has certain special requirements, and these characteristics must be explicitly defined to enable this kind of information to

be accessed

Chapter XIX

A Web Metadata Based-Model for Information Quality Prediction 324 Ricardo Barros, COPPE—Federal University of Rio de Janeiro, Brazil

This chapter addresses those issues by proposing a Web metadata-based model to evaluate and mend Web pages based on their information quality, as predicted by their metadata

recom-Chapter XX

Towards Quality Web Information Systems Through Precise Model-Driven Development 344

Fernando Molina, University of Murcia, Spain

This chapter presents one WIS development methodology (MIDAS) that has been completed with the definition of a strategy for the formal specification of its models with V&V objectives

Chapter XXI

The Use of Metamodels in Web Requirements to Assure the Consistence 363

M J Escalona, University of Seville, Spain

G Aragón, Everis, Spain

This chapter presents NDT (navigational development techniques), a Web methodological approach to deal with requirements, based on model-driven engineering The proposal is composed of a set of pro-cedures, techniques, and models to assure the quality of results in the Web requirements treatment

Coral Calero, University of Castilla-La Mancha, Spain

This chapter introduces the necessity to consider quality management activities as part of the Web engineering (WE) process to improve the final quality of Web applications with respect to creative practices

Chapter XXIII

Restrictive Methods and Meta Methods for Thematically Focused Web Exploration 405

Sergej Sizov, University of Koblenz-Landau, Germany

This chapter addresses the problem of automatically organizing heterogeneous collections of Web ments for generation of thematically focused expert search engines and portals As a possible application scenario for the presented techniques, the authors show a Web crawler that aims to populate topics of interest by automatically categorizing newly fetched documents

docu-Chapter XXIV

WSRP-O: An Ontology to Model WSRP Compliant Portlets 424

Mª Ángeles Moraga, University of Castilla-La Mancha, Spain

This chapter presents an ontology for WSRP standard The aim of this standard is to provide a common interface in order to allow the communication between portal and portlets Bearing this in mind, in this work the authors propose an ontology for the WSRP standard The ontology offers an understandable summary of the standard

Chapter XXV

Philosophy of Architecture Design in Web Information Systems 443

Tony C Shan, Bank of America, USA

This chapter defines a comprehensive set of guiding principles—philosophy of architecture design (PAD)—as a means of coping with the architecture design complexity and managing the architectural assets of Web information systems in a service-oriented paradigm

in this section, four of them are related to the first topic, whereas the last two are related to the second one The chapters focused on search engine deal with the following issues: improve the quality of Web search, enhance Web search engine performance, and Web search engine architectures Regarding the chapters which are focused on information, we can highlight that one of them proposes a model for evaluating Web retrieval systems in non English queries, and the other one presents Web information resources vis-à-vis traditional information services.

Chapter XXVI

Improving the Quality of Web Search 463

Mohamed Salah Hamdi, University of Qatar, Qatar

In this chapter, the authors propose the improvement of the quality of Web search by combining search and self-organizing maps This can help users both in locating interesting documents more easily and in getting an overview of the retrieved document set

meta-Chapter XXVII

The Perspectives of Improving Web Search Engine Quality 481 Jengchung V Chen, National Cheng Kung University, Taiwan

In this chapter, some past research in Web search and current trends focusing on how to improve the search quality in different perspectives of “what,” “how,” “where,” “when,” and “why” are discussed

Chapter XXVIII

Web Search Engine Architectures and their Performance Analysis 491

Xiannong Meng, Bucknell University, USA

The objectives of the chapter are to review the theories and technologies pertaining to Web search, helping in the understanding of how Web search engines work, and how to use the search engines more effectively and efficiently

Chapter XXIX

Towards a Model for Evaluating Web Retrieval Systems in Non-English Queries 510

Fotis Lazarinis, University of Sunderland, UK

The purpose of this chapter is to describe the methods and the criteria used for evaluating search gines, and proposes a model for evaluating the searching effectiveness of Web retrieval systems in non English queries

Nina Mistilis, The University of New South Wales, Australia

The chapter proposes that visitor information centres are analogous to a general information system and that centre user experience can partially be explained by their perception of the information resource quality

About the Contributors 541 Index 553

of Web applications in order to give a good service and assure the loyalty of users In the second case, the quality is important in order that users can achieve their objectives in a proper way

Therefore, it is essential not only to develop new Web information systems, but also to take into count their quality With regard to this, new methodologies and models are being developed to improve and assess the quality of Web information systems In such an ever-evolving environment, Web engineers, software quality managers, software engineers, software architects, MSc Students, and university-level professors of the discipline need access to the most current information about the models, measures, and methodologies in this emerging field

ac-The Web age has modified our society, and new business models have appeared, while others have been modified In addition, the relationships between the different actors have changed

It may be worth emphasizing that Web Technologies have become very important in information systems Web Technologies are essential for organizations Currently, it is indispensable that the devel-oped Web products, such as Web pages, Web applications, Web portals, and so forth, achieve a minimum level of quality

According to Offutt 20021, several factors inherent to Web development affect to the quality:

1 Developers build Web-based software systems by integrating numerous diverse components from disparate sources, including custom built special-purpose applications, customized off-the-shelf software components, and third-party products

2 Much of the new complexity found with Web-based applications also results from how the ent software components are integrated Not only is the source unavailable for most of the com-ponents, the executables might be hosted on computers at remote, even competing organizations

differ-To ensure high quality for Web systems composed of very loosely coupled components, we need novel techniques to achieve and evaluate these components’ connections

3 Finally, Web-based software offers the significant advantage of allowing data to be transferred among completely different types of software components that reside and execute on different computers However, using multiple programming languages and building complex business ap-plications complicates the flow of data through the various Web software pieces

The Handbook of Research on Web information systems Quality provides comprehensive coverage

of the most important issues, such as: effort and quality assessment, accessibility, usability, metadata, MDE, metamodels, ontologies search engine, and information All of them focus on Web information systems

The handbook is divided into four sections that cover the main tendencies on the Web information systems research and gives a clear vision of the main actual lines of work and also the topics where more effort is being developed

The first section is on quality assessment, where different approaches, whose central point is quality, are presented The second is on accessibility and usability, perhaps two of the most important factors related to Web information systems and where more research and development efforts are deployed from the beginning of this discipline The third section approaches the technological point of view with chapters about metadata, MDE, metamodels, and ontologies Finally, the last section works on Web engines and information on the Web

As we have already mentioned, the first section is related to effort and quality assessment and is composed of eight chapters The first chapter presents a survey literature of size measures (attributes) that have been proposed for Web effort estimation These measures are classified according to a pro-posed taxonomy In addition, the authors discuss ways in which Web companies can devise their own size measures The objective of the second chapter is to introduce the concepts related to Web effort estimation and effort estimation techniques It also details and compares, by means of a case study, three effort estimation techniques

Chapter III emphasizes the significance of approaching Web information systems (WIS) from an engineering viewpoint A methodology for deploying patterns as means for improving the quality of WIS as perceived by their stakeholders is presented The fourth chapter discusses and analyses the ef-fectiveness of SME business to business Web sites from a user perspective under the premise that an effective method of evaluating a Web site can contribute to the development of more quality Web sites and greater realization of benefits In Chapter V, the problem of Web application quality assessment is assessed from two perspectives

Chapter VI presents the most prominent systems and prototypes implemented for the automatic ity assessment for Internet pages and analyzes the knowledge sources exploited for these approaches

qual-In Chapter VII several portal quality models are presented and compared Authors have adapted one of the best portal quality model proposed in the literature to the e-banking context

Finally, the section ends with a chapter that proposes a model for data quality in Web portals (PDQM) built upon the foundation of three key aspects: (1) a set of Web data quality attributes identified in the literature in this area, (2) data quality expectations of data consumers on the Internet, and (3) the func-tionalities that a Web portal may offer its users

The second section of the book is divided into two main topics and is composed of seven chapters The first works on both topics The three next chapters deal with accessibility, one of them from a gen-eral point of view, another one comparing approaches to Web accessibility assessment, and the last one about maximizing Web accessibility The other three chapters are about usability from the point of view

of ergonomic criteria as part of the development of Web applications or as an important aspect for the construction of business process driven Web applications

Concretely, Chapter IX is presented in the structure of an index, which allows the development team

to create the specification of the context of use document for the development of Web applications, ing in mind characteristics of usability and accessibility

bear-Chapter X studies the Web accessibility issue from the perspective of Web information systems Quality In addition, the closed relationship between accessibility and standard Web technologies is

explained In the eleventh chapter, the importance of Web accessibility assessment is discussed and 15 different approaches found in literature are compared

The user interface is the place where users can interact with the information by using their minds Users with special needs can acquire information by using a human centred user interface Chapter XII highlights the need to investigate the relationship between cognition and user interface

Chapter XIII proposes a quality model that focuses on quality in use or usability for the product acterization of the World Wide Web, and Chapter XIV surveys the most emergent usability evaluation models to be adopted during the whole lifecycle of Web information systems, for promoting usability

char-In the last chapter of this section, the authors gather a set of guidelines provided by experts in Web usability and present the solution designed in a particular Web engineering method that follows a model driven development approach

The third section is related to metadata, MDE, metamodels, and ontologies The first four chapters focus on metadata issues In the next three chapters, different metamodels oriented to Web requirements, development of Web applications, and Web exploration are shown Finally, an ontology for WSRP stan-dard is presented and a philosophy of architecture design in Web information systems

The section starts with Chapter XVI, focused on “portletizing” existing Web applications, that is, wrapping them as portlets, without requiring any modification After providing some background on portlet technology, they discuss two kinds of approaches to portletization: automatic and annotation-based In Chapter XVII, the authors survey techniques for ontology evolution The authors detail the various existing languages and techniques devoted to Web data evolution, with particular attention to Semantic Web concepts, and how these languages and techniques can be adapted to evolving data in order to improve the quality of Web information systems applications

In Chapter XVIII, the importance of using ontologies to represent database schemas is highlighted The representation of the fuzzy data in fuzzy databases management systems (FDBMS) has certain special requirements, and these characteristics must be explicitly defined to enable this kind of information to

be accessed Chapter XIX addresses those issues by proposing a Web metadata-based model to evaluate and recommend Web pages based on their information quality, as predicted by their metadata

Chapter XX presents one WIS development methodology (MIDAS) that has been completed with the definition of a strategy for the formal specification of its models with V&V objectives, and chapter XXI presents NDT (navigational development techniques), a Web methodological approach to deal with requirements, based on model-driven engineering The proposal is composed of a set of procedures, techniques and models to assure the quality of results in the Web requirements treatment

Chapter XXII introduces the necessity to consider quality management activities as part of the Web engineering (WE) process to improve the final quality of Web applications with respect to creative practices

The problem of automatically organizing heterogeneous collections of Web documents for generation

of thematically focused expert search engines and portals is the focus of Chapter XXIII As a possible application scenario for the presented techniques, the authors show a Web crawler that aims to populate topics of interest by automatically categorizing newly fetched documents

An ontology for WSRP standard is presented in Chapter XXIV The aim of this standard is to provide

a common interface in order to allow the communication between portal and portlets Bearing this in mind, in this work the authors propose an ontology for the WSRP standard that offers an understandable summary of the standard

Closing this section, Chapter XXV provides a comprehensive set of guiding principles—philosophy

of architecture design (PAD)—as a means of coping with the architecture design complexity and ing the architectural assets of Web information systems in a service-oriented paradigm

manag-The last section of the book focuses on two main topics: search engine and information Among the chapters classified in this section, four of them are related to the first topic, whereas the last two are related to the second one

The section starts with Chapter XXVI, where the authors propose the improvement of the quality of Web search by combining meta-search and self-organizing maps This can help users both in locating interesting documents more easily and in getting an overview of the retrieved document set

In Chapter XXVII, some past research in Web search and current trends focusing on how to improve the search quality in different perspectives of “what,” “how,” “where,” “when,” and “why” are discussed The objectives of Chapter XXVIII are to review the theories and technologies pertaining to Web search, helping in the understanding of how Web search engines work, and how to use the search engines more effectively and efficiently

The purpose of the Chapter XXIX is to describe methods and criteria used for evaluating search engines The chapter also proposes a model for evaluating the searching effectiveness of Web retrieval systems in non English queries

Finally, the last chapter of the book proposes that visitor information centres are analogous to a eral information system and that centre user experience can partially be explained by their perception

gen-of the information resource quality

Due to the variety of topics and the different aspects related to the research on quality for Web formation systems, this handbook can be used by software engineering researchers and practitioners (professors, PhD, and postgraduate students, industrial R&D departments, etc.) for helping in the un-derstanding of the topic, knowing about the main current tendencies of research and the future lines of research on Web information systems quality

in-EndnotE

1 Offutt, A J (2002) Quality attributes of Web software applications IEEE Software, 19(2),

25-32

Special thanks also go to the publishing team at IGI Global, whose contributions throughout the whole process from inception of the initial idea to final publication have been invaluable Thanks in particular

to Kristin Roth and Jessica Thompson, who continuously prodded via e-mail for keeping the project on schedule, and to Michelle Potter, whose enthusiasm motivated me to initially accept her invitation for taking on this project

Special thanks go to Spanish Ministry of Education that financed this work, under the research projects ESFINGE (TIN2006-15175-C05-05) and CALIPSO (TIN2005-24055-E)

In closing, we wish to thank all of the authors for their insights and excellent contributions to this handbook

Dra Coral Calero

Dra Mª Ángeles Moraga

Dr Mario Piattini

Ciudad Real, Spain

June 2007

Effort and Quality Assessment

Chapter I

Sizing Web Applications for

Web Effort Estimation

Emilia Mendes

The University of Auckland, New Zealand

IntroductIon

The purpose of estimating effort is to predict the

necessary amount of labour units to accomplish

a given task, based on knowledge of previous

similar projects and other project characteristics

that are believed to be related to effort Project

characteristics are the input, and effort is the

output we wish to predict

A task to be estimated can be as simple as

developing a single function (e.g., creating a Web

form with 10 fields) or as complex as developing

a large application, and the one input always

found to have the strongest influence on effort

is size Thus, using an adequate size measure is fundamental to building adequate and accurate effort estimation models

One of the main challenges in Web effort estimation is to determine what is/are the best measure(s) to be used to size an application There are no standards and throughout industry and academia different size measures are used Since 1998, numerous size measures have been proposed for Web effort estimation and it is important that such body of knowledge be struc-tured and made available such that practitioners may look at existing measures and assess whether

or not they are applicable to their own



ment; in addition, researchers may use this body

of knowledge as a starting point to understand

trends in sizing Web applications

The literature to date has published three

sur-veys on Web measures (Calero, Ruiz, & Piattini,

2004; Dhyani, Ng, & Bhowmick, 2002; Mendes,

Counsell, & Mosley, 2005) Of these, only Mendes

et al (2005) has included measures that are

ap-plicable for Web effort estimation

Each survey is briefly described below:

• Dhyani et al (2002) concentrates on

mea-sures that belong to one of the following six

categories:

◦ Web graph properties: measures

that quantify structural properties

of the Web on both macroscopic and

microscopic scales

◦ Web page significance: measures used

to assess candidate pages in response

to a search query and have a bearing

on the quality of search and retrieval

on the Web

◦ Usage characterization: measures

that quantify user behavior aiming at

improving the content, organization,

and presentation of Web sites

◦ Web page similarity: measures that

quantify the extent of association

be-tween Web pages

◦ Web page search and retrieval:

mea-sures for evaluating and comparing

the performance of Web search and

retrieval services

◦ Information theoretic: measures that

capture properties related to

informa-tion needs, producinforma-tion, and

consump-tion

• Calero et al (2004) provides a survey where

Web measures are classified into three

di-mensions, all related to Web quality:

◦ Web features dimension: incorporates

content, navigation, and presentation

measures

◦ Quality characteristics dimension:

incorporates functionality, reliability, efficiency, portability, and maintain-ability measures

◦ Life cycle processes dimension:

Process measures related to a Web development life cycle

In addition to the above classification, Calero

et al (2004) also assess the surveyed sures according to an additional criteria:

mea-◦ Granularity level: whether the

meas-ure’s scope is a “Web page” or “Web site.”

◦ Theoretical validation: whether or not

a measure has been validated cally

theoreti-◦ Empirical validation: whether or

not a measure has been empirically validated

◦ Automated support: whether or not

there is a support tool that facilitates the calculation of the measure

• Mendes et al (2005) provided a survey and taxonomy of hypermedia and Web size measures based on literature published since 1992 The criteria they used to clas-sify measures will be detailed in the next section because this is the same criteria we use in this chapter A taxonomy represents a model that is used to classify and understand

a body of knowledge

This chapter’s objectives are twofold: first,

to complement Mendes et al.’s work by focusing further on size measures for Web effort estima-tion We employ the same taxonomy proposed in Mendes et al (2005) to classify the existing body

of knowledge; second, to make recommendations

to Web companies on how to define their own size measures, whenever that seems applicable.The remainder of this chapter is organised

as follows: First, it introduces the taxonomy we employ, explaining terms and definitions that are part of this classification Second, it presents our

literature review, which was based on 10 papers

Note that we only included in our literature review

papers that proposed a new set of measures, that

is, if two or more papers used the same set of size

measures we included only the first one published

Third, it applies the taxonomy to classify each of

the papers from our literature review Fourth, it

discusses the change in trends that have occurred

in the area of Web sizing Fifth, it details the

necessary steps to be used by a Web company to

derive its own size measures Finally, it presents

its conclusions

sIzE MEAsurEs tAxonoMy

The taxonomy that was proposed by Mendes et

al (2005) uses as its basis software measurement

concepts 0 and literature in software size

mea-sures and measurement (Briand & Wieczorek,

2002) It originally comprises nine categories,

which are applied to each size measure identified

in the literature We will only use eight of these

categories, which are as follows:

The harvesting time category describes when, in

a project’s development life cycle, the measure

should be obtained (measured) Whenever a

measure has to be obtained early on in a project’s

development life cycle it is very likely that it will

need to be estimated Otherwise, it may be directly

measured The main motivation for including this category to classify the size measures in this chapter is that measures for effort estimation should ideally all be gathered early in a project’s development life cycle, and thus we want to also assess to what extent what has been proposed in the literature complies with this premise This category can have simple values such as “early size measure” or “late size measure;” however,

a longer description can also be given whenever necessary (e.g., “late size measure to be measured after the implementation is finished”)

Measure Foundation

The measure foundation category describes whether the size measure is a problem-orientated measure or a solution-orientated measure (Briand

& Wieczorek, 2002):

• Problem-orientated measure: a

problem-orientated measure assumes that an tion’s size corresponds directly to the size of the problem to be solved in order to deliver

applica-a corresponding applica-applicapplica-ation Therefore, the greater the problem, the greater the size

In this context, the problem to be solved is denoted by the functionality of the applica-tion to be developed Problem-orientated size measures generally take the form of surrogate measures of functionality These measures can be extracted from the speci-fication or design documents (e.g., use case diagrams 0, data flow diagrams (DeMarco, 1982), or entity-relationship models (Mendes

et al., 2002)) An example of a common problem-oriented metric is function points, which aims to measure the size of an applica-tion in terms of the amount of functionality within the application, as described by its proposed specification 01997)

• Solution-orientated measure: in contrast,

a solution-orientated measure assumes that

an application’s size corresponds to the



actual delivered size of an application A

frequently used size measure is lines of code

(LOC), which measures the size of a given

software implementation This measure has

been frequently criticised for its difficulty in

being measured consistently (Jones, 1998)

and for being a difficult measure to estimate

early in the development life cycle Finally,

another source of criticism is that LOC is

a measure that is highly dependent on the

programming paradigm, language and style

employed (Briand & Wieczorek, 2002)

class

The class category allows for the classification of

size measures into either of three possible classes:

length, complexity, and functionality (Fenton &

Pfleeger, 1997):

• Length: measures the physical size of an

application;

• Functionality: measures the functions and

features supplied by the application to the

user; and

• Complexity: measures the structural

complexity of an application, where the

application’s structure is represented by the

way in which nodes (e.g., Web pages) are

interconnected via links The assumption

behind complexity size measures is that by

analysing the application’s structure, the

application’s development (authoring) can

be improved to create more comprehensible

structures These therefore improve the

application’s usability, as they enable users

to better traverse (navigate) the application

More comprehensible structures also reduce

the disorientation caused by traversing a

complex structure

According to the descriptions given above, we

can say that the foundation for both length and

complexity measures is “solution-orientated,”

whereas the foundation for a functionality size measure is “problem-orientated.”

• Web hypermedia application

(Christo-doulou, Zafiris, & Papatheodorou, 2000): a nonconventional application characterised

by the authoring of information using nodes (chunks of information), links (relations between nodes), anchors, access structures (for navigation), and delivery over the Web Technologies commonly used for develop-ing such applications are HTML, XML, JavaScript, and multimedia In addition, typical developers are writers, artists, and organisations who wish to publish informa-tion on the Web or CD-ROM without the need to know programming languages such

as Java These applications have unlimited potential in areas such as software engineer-ing, literature, education, and training

• Web software application (Christodoulou

et al., 2000): a conventional software tion that relies on the Web or uses the Web’s infrastructure for execution Typical appli-cations include legacy information systems such as databases, booking systems, knowl-edge bases, and so forth Many e-commerce applications fall into this category Typically, they employ development technologies (e.g., DCOM, ActiveX, etc.), database systems, and development solutions (e.g., J2EE) De-velopers are in general young programmers fresh from a Computer Science or Software Engineering degree course, managed by a few more senior staff

applica-• Web application: an application delivered

over the Web that combines characteristics

of both Web hypermedia and Web software

applications

• Media: a multimedia component, for

ex-ample, graphic, audio, video, animation,

and photograph

• Program/Script: code employed to add

functionality to an application (e.g., Perl

scripts, javascript)

• Web application design model: a

concep-tual representation of a Web application

Such representations are characterised by

models, for example, navigation model

and presentation model These models are

abstractions that later are translated into an

implementation of the Web application

Measurement scale type

To understand the measurement scale type

cat-egory it is first important to understand what

measurement means Measurement represents a

process by which numbers or symbols are assigned

to attributes (measures) of entities in the real world

such that these entities can be described according

to clearly defined rules For example, in relation

to an entity “Person,” the attributes (measures)

height, weight, and gender are used as

character-istics of “Person.” Each attribute (measure) can be

measured using one of five different measurement

scale types Each scale type represents a set of

characteristics associated with a measure that help

interpret this measure and also determine what

sort of manipulations can be applied to it The

five scale types are Nominal, Ordinal, Interval,

Ratio, and Absolute (Fenton & Pfleeger, 1997),

and their descriptions are as follows:

• Nominal: Defines classes or categories,

and places entities in a particular class or

category, based on the value of the attribute

Let’s suppose we wish to measure the

at-tribute application type for the entity “Web

application,” and that the types of application considered were:

attribute application type Note that there is

no notion of ordering between the classes This means that, even if we had used instead

of symbols, numbers from 1 to 15, they would not represent any notion of ranking between numbers The same also applies to symbols, that is, an application of type “News and Information” is not more or less important than an application of type “Personal.”

• Ordinal: Augments the nominal scale with

information about an ordering of classes or categories This means that entities belong

to classes that are ordered with respect

to the attribute Let’s suppose we wish to

measure the attribute application structural

complexity for the entity “Web application,”

and that structural complexity is measured using the following classes:



attribute application structural complexity

Here, there is a notion of ordering (ranking)

between the classes This means that

applica-tions that belong to class “Very high” have

greater structural complexity than those that

belong to class “High,” and so forth Classes

can be represented by numbers or symbols;

however, it is important to note that even if

we had used number (e.g., 1 to 5) to represent

classes, these numbers would only represent

ranking, so addition, subtraction, and other

arithmetic operations have no meaning

• Interval: Augments the ordinal scale with

information about the size of the intervals

that separate the classes Thus, the ranking

between classes is preserved, however, now

the interval between two classes is constant

For example, the difference between 20°C

- 10°C is the same as that for 30°C - 20°C

However, it does not make sense to say

that 20°C is twice as hot as 10°C Another

example of an interval scale is relative time,

for example, the number of calendar days

since the start of a given Web project The

difference between two consecutive

calen-dar days is always the same Note that this

measurement scale type does not have a

natural zero representing the complete

ab-sence of a class Addition and subtraction are

acceptable operations between two classes,

but not multiplication and division.

• Ratio: Preserves ordering, the size of

in-tervals between classes, and ratios between

classes People’s heights and weights are

typical ratio-scale measures It is meaningful

to say that someone who has a height of 180

cm is twice as tall as someone who has a

height of 90 cm, and this holds true

regard-less of whether height is being measured in

centimetres, meters, or yards All

arithme-tic operations are acceptable between two

classes

• Absolute: The measure always takes the

form “number of occurrences of x in the

entity E.” For example, to measure the size

of a “Web application” using as measure the

number of new Web pages uses an absolute

scale because there is only one choice here which is to count the number of new Web pages This scale type is very often used

to measure software and Web application attributes

computation

The computation category describes whether a size

measure can be measured directly or indirectly (Fenton & Pfleeger, 1997) Indirect measurement

means that the measure is computed based on other

measures Conversely, direct measurement means

that the size measure does not rely on other sures in order to be measured For example, assume the three size measures presented below:

mea-• Page count: Number of HTML or SHTML

files

• Connectivity: Number of internal links, not

including dynamically generated links

• Connectivity density: Computed as

Con-nectivity divided by page count.

Page count and connectivity are both direct

measures because they can be measured without

using other measures However, connectivity

density is an indirect measure because to be

computed it uses other two measures: connectivity and page count

Validation

The validation category describes whether a size measure has been validated To be validated means that evidence has been gathered regard-ing the measure’s usefulness to measure what it purports to measure Validations can be carried out empirically, where generally data is used to provide evidence of a measure’s usefulness; or theoretically, where the measurement principles associated with a proposed measure are checked

to make sure that they are in line with the

mea-surement theory that supports the definition of

that measure

Possible values for the validation category are

“validated empirically,” “validated theoretically,”

“both,” and “none.” This is similar to one of the

criterion suggested by Calero et al (2004)

Model dependency

This represents whether a size measure requires

the use of a specific Web methodology or model

in order to be measured For example, as will be

discussed later, Mangia and Paiano (2003)

pro-posed size measures to estimate effort to develop

Web applications that have been modeled using

the W2000 methodology This means that unless

Web companies use the W2000 methodology to

design and model their Web applications they are

unlikely to find the size measures proposed by

Mangia and Paiano (2003) useful for their own

context On the other hand, Mendes, Mosley, and

Counsell (2003) has proposed size measures that

are applicable to measure the size of any Web

application, where applications can be designed

using the W2000 methodology or not The two

possible values that this category takes are

“Spe-cific” and “Nonspecific.”

LItErAturE rEVIEw oF wEb

sIzE MEAsurEs

This section presents a literature review of Web

size measures proposed since 1998, described in

chronological order

1998, 2000: size Measures by

cowderoy

Cowderoy, Donaldson, and Jenkins (1998) and

Cowderoy (2000) organised their proposed

mea-sures into four distinct categories: Web

applica-tion, Web page, media, and program The sets

of measures within each category are presented

below:

Web Application

• Web pages: Measures the number of Web

pages in a Web application

• Home pages: Measures the number of major

entry points to the Web application

• Leaf nodes: Measures the number of Web

pages, in a Web application, that have has

no siblings

• Hidden nodes: Measures the number of Web

pages excluded from the main navigation buttons

• Depth: Measures the number of Web pages

on the second level that have siblings

• Application Paragraph count: Measures

the number of Page paragraph count

(de-scribed later) for all Web pages in a Web application

• Delivered images: Measures the number of

unique images used by a Web application

• Audio files: Measures the number of unique

audio files used in a Web application

• Application movies: Measures the number

of Page movies (described later) for all the

Web pages in an application

• 3d objects: Measures the number of files (including 3D objects) used in a Web ap-plication

• Virtual worlds: measures the number of

files (including virtual worlds) used in a Web application

• External hyperlinks: Measures the number

of unique URLs in a Web application

Web Page

• Actions: Measures the number of

indepen-dent actions by use of Javascript, Active X, and so forth

• Page paragraph count: Measures the

number of paragraphs in a Web page

• Word count: measures the number of words

in a Web page



• Navigational structures: Measures the

number of different structures in a Web

page

• Page movies: Measures the number of movie

files used in a Web page

• Interconnectivity: Measures the number of

URLs that link to other pages in the same

application

Media

• Image size (IS): Measures the size of an

image, computed as width * height

• Image composites: Measures the number

of layers from which the final image was

created

• Language versions: Measures the number

of image versions that must be produced to

accommodate different languages or

differ-ent cultural priorities

• Duration: Measures the summed duration

of all sequences within an audio file

• Audio sequences: Measures the number of

sequences within the audio file

• Imported images: Measures the number

of graphics images imported into an audio

file

Program

• Lines of source code: Measures the number

of lines of code in a program/script

• McCabe ciclomatic complexity: Measures

the structural complexity of a program/

script

2000, 2001: size Measures by

Mendes et al

Mendes, Counsell, and Mosely (2000) and

Mendes, Mosely, and Counsell (2001) organised

their proposed size measures into five distinct

categories: hypermedia application, Web

ap-plication, Web page, media, and program The

sets of measures proposed for each category are presented below:

Web Application

• Page count: Measures the total number of

HTML or SHTML files that a Web tion has

applica-• Media count: Measures the total number

of unique media files that a Web application has

• Program count: Measures the total number

of CGI scripts, JavaScript files, and Java applets that a Web application has

• Total page allocation: Measures the total

amount of space (Mbytes) allocated for all the HTML or SHTML pages of a Web ap-plication

• Total media allocation: Measures the total

amount of space (Mbytes) allocated for all media files used by a given Web applica-tion

• Total code length: Measures the total

num-ber of lines of code for all the programs in

a given Web application

• Reused media count: Measures the total

number of reused or modified media files that a given Web application has

• Reused program count: Measures the total

number of reused or modified programs that

a given Web application has

• Total reused media allocation: Measures

the total amount of space (Mbytes) allocated for all reused media files that belong to a given Web application

• Total reused code length: Measures the

total number of lines of code for all reused programs that belong to a given Web ap-plication

• Code comment length: Measures the total

number of comment lines in all the programs that belong to a given Web application

• Reused code length: Measures the total

number of reused lines of code in all the

programs that belong to a given Web

ap-plication

• Reused comment length: Measures the

total number of reused comment lines in

all the programs that belong to a given Web

application

• Total page complexity: Measures the

av-erage number of different types of media

used, excluding text, for a given Web

ap-plication

• Connectivity: Measures the total number

of internal links, not including dynamically

generated links, for a given Web

applica-tion

• Connectivity density: Computed as

Con-nectivity divided by page count Measures

the average number of links per Web page

for a given Web application

• Cyclomatic complexity: Computed as

Connectivity - page count) + 2 Measures

the cyclomatic complexity for a given Web

application

Web Page

• Page allocation: Measures the total

allo-cated space (Kbytes) of a HTML or SHTML

file

• Page complexity: Measures the total number

of different types of media used on a Web

page, not including text

• Graphic complexity: Measures the total

number of graphics media on a Web page

• Audio complexity: Measures the total

number of audio media on a Web page

• Video complexity: Measures the total

number of video media on a Web page

• Animation complexity: Measures the total

number of animations on a Web page

• Scanned image complexity: Measures the

total number of scanned images on a Web

page

• Page linking complexity: Measures the

total number of links on a Web page

Media

• Media duration: Measures the total

dura-tion (minutes) of audio, video, and tion

anima-• Media allocation: Measures the total size

(Kbytes) of a media file

Program

• Program Code length: Measures the total

number of lines of code in a program

2000: size Measures by rollo

Although Rollo (2000) did not suggest any new size measures we have included his work here because he was the first, as far as we know, to investigate the issues of measuring functional-ity of Web applications specifically aiming at cost estimation, using numerous function point analysis methods:

• Functional size: Measures the total number

of function points associated with a Web application Function points were measured using COSMIC-FFP, Mark II, and Albrecht

(Rollo, 2000)

Later, other studies have also employed the COSMIC full function points method to size Web applications (Mendes et al., 2002; Umbers,

& Miles, 2004) These studies are not described here as the size measure employed is the same one used by Rollo

2000: size Measures by cleary

Cleary (2000) proposed size measures organised into three categories: Web hypermedia applica-tion, Web software application, and Web page Each measure is detailed below:

0

Web Hypermedia Application

• Nontextual elements: Measures the total

number of unique nontextual elements

within a Web hypermedia application

• Externally sourced elements: Measures the

total number of externally sourced elements

Being externally sourced means that such

elements were not developed by the

devel-opment team responsible for developing the

given Web hypermedia application They

can be developed within the same company

by a different group of developers, or even

developed by third party

• Customised infrastructure components:

Measures the total number of customised

infrastructure components Such

compo-nents would not have been developed from

scratch for the given Web hypermedia

ap-plication, but rather, reused from elsewhere

and adapted to the given application

• Total Web points: Measures the total size

of a Web hypermedia application in Web

points The Web points measure computes

size by taking into account the complexity

of the Web pages contained within an

ap-plication Complexity of a page is a function

of the number of words this page contains,

number of existing links, and number of

nontextual elements Once the complexity

of a page is measured, it leads to a number

of Web points for that page (Abrahao, Poels,

& Pastor, 2004)

Web Software Application

• Function points: measures the

function-ality of a Web software application using

any existing function points measures (e.g.,

IFPUG, Mark II, COSMIC)

Web Page

• Nontextual elements page: Measures the

total number of nontextual elements in a Web page

• Words Page: Measures the total number of

words in a Web page

• Web points: Measures the total length of

a Web page This measure uses an ordinal scale with scale points “low,” “medium,” and

“high.” Each point is attributed a number

of Web points, previously calibrated to a specific dataset of Web projects data

• Number of links into a Web page: Measures

the total number of incoming links (internal

or external links) Incoming links are links that point to a given Web page

• Number of links out of a Web page:

Mea-sures the total number of outgoing links (internal or external links) Outgoing links are links that have their origin at the given Web page and destination elsewhere

• Web page complexity: Measures the

com-plexity of a Web page based upon its number

of words, and combined number of ing and outgoing links, plus the number of nontextual elements

incom-2000: size Measures by reifer

Reifer (2000) proposed a single size measure to

be used to estimate effort to develop Web plications:

ap-• Web objects: Measures the total number

of Web Objects in a Web application using Halstead’s equation for volume, tuned for Web applications The equation is as fol-lows:

N = number of total occurrences of operands and

operators

n = number of distinct operands and operators

N1 = total occurrences of operand estimator

N2 = total occurrences of operator estimators

n1 = number of unique operands estimator

n2 = number of unique operators estimators

V = volume of work involved represented as Web

objects

Operands are comprised of the following

measures:

• Number of building blocks: Measures the

total number of components in a Web

ap-plication, for example, Active X, DCOM,

and OLE

• Number of COTS: Measures the total

number of COTS components (including

any wrapper code) in a Web application

• Number of multimedia files: Measures

the total number of multimedia files, except

graphics files, in a Web application

• Number of object or application points

(Cowderoy et al., 1998; Cowderoy, 2000):

Measures the total number of

object/ap-plication points, and so forth, in a Web

application

• Number of Lines: Measures the total

num-ber of xml, sgml, html, and query language

lines in a Web application

• Number of Web components: Measures

the total number of applets, agents, and so

forth, in a Web application

• Number of graphics files: Measures the

total number of templates, images, pictures,

and so forth, in a Web application

• Number of scripts: Measures the total

num-ber of scripts for visual language, audio,

mo-tion, and so forth, in a Web application

2003: size Measures by Mendes et al

Mendes et al (2003) proposed size measures, organised as follows:

Web Application

• Web pages: Measures the total number of

Web pages in a Web application

• New Web pages: Measures the total number

of Web pages created from scratch in a Web application

• Customer Web pages: Measures the total

number of Web pages, provided by the customer, in a Web application

• Outsourced Web pages: Measures the total

number of outsourced Web pages in a Web application

• Text pages: Measures the total number of

text pages (A4 size), part of a Web tion, which had to be typed

applica-• Electronic text pages: Measures the total

number of reused text pages, part of a Web application, which are in electronic for-mat

• Scanned text pages: Measures the total

number of reused text pages, part of a Web application, which had to be scanned with OCR

• New images: Measures the total number of

new images/photos/icons/buttons created from scratch for a given Web application

• Electronic images: Measures the total

number of reused images/photos, contained

in a given Web application, which are in electronic format

• Scanned images: Measures the total

num-ber of reused images/photos, contained in

a given Web application, which need to be scanned

• External images: Measures the total

num-ber of images, contained in a given Web application, which were obtained from an image/photo library or outsourced



• New animations: Measures the total

num-ber of new animations (Flash/gif/3D, etc.),

contained in a given Web application, which

were created from scratch

• External animations: Measures the total

number of reused animations (Flash/gif/3D

etc.) contained in a given Web application

• New audio: measures the total number of

new audio/video clips created from scratch

for a given Web application

• External audio: Measures the total number

of reused audio/video clips contained in a

given Web application

• High Fots: Measures the total number of

High-effort features off-the-shelf (FOTS)

contained within a given Web application

Features off-the-shelf are features that have

been reused as they are, without any

adapta-tion High effort represents the minimum

number of hours to develop a single

func-tion/feature by one experienced developer

that is considered high (above average)

This number is currently set to 15 hours

based on collected data from industrial Web

projects

• High FotsA: Measures the total number of

High-effort FOTS contained in a Web

ap-plication, which were reused and adapted

to local circumstances High effort here

represents the minimum number of hours

to adapt a single function/feature by one

experienced developer that is considered

high (above average) This number is

cur-rently set to 4 hours based on collected data

from industrial Web projects

• High new: Measures the total number of new

High-effort Feature/Functionality contained

in a Web application, which was developed

from scratch

• Fots: Measures the total number of

Low-ef-fort FOTS contained in a Web application

• FotsA: Measures the total number of

Low-effort FOTS contained in a Web application,

which were adapted to local circumstances

• New: Measures the total number of

Low-effort Feature/Functionality contained in

a Web application, which were developed from scratch

Examples of feature/functionality are as follows:

Mangia and Paiano (2003) proposed size measures

to be used to estimate the necessary effort to velop Web applications that have been modeled according to the W2000 methodology

de-Web Application

• Macro: Measures the total number of

mac-rofunctions in a Web application, which are required by the user

• DEI: Measures the total number of input

data for each operation

• DEO: Measures the total number of output

data for each operation

• Entities: Measures the total number of

in-formation entities that model the database conceptually

• AppLimit: Measures the total application

limit of each operation

• LInteraction: Measures the total level of

interaction that various users of the

applica-tion have in each operaapplica-tion

• Compatibility: Measures the total

com-patibility between each operation and an

application’s delivery devices

• TypeNodes: Measures the total number of

types of nodes that constitute the

naviga-tional structure

• Acessibility: Measures the total number

of accessibility associations and pattern of

navigation between node types

• NavCluster: Measures the total number of

navigation clusters

• ClassVisibility: Measures the total

vis-ibility that classes of users have of a Web

application’s navigational structure

• DeviceVisibility: Measures the total

vis-ibility that delivery devices have of a Web

application’s navigational structure

2003: size Measures by baresi et al

Baresi, Morasca, and Paolini (2003) proposed

size measures to estimate the effort required to

design Web applications that have been designed

according to the W2000 methodology Their size

measures were organised according to the three

different types of design models that result from

using W2000: information model, navigation

model, and presentation model These measures

are detailed below:

Information Model

enti-ties in the model

components in the model

in the model

number of slots per semantic association center

number of slots per collection center in the model

number of components per entity

number of slots per component

semantic associations in the model

seman-tic association centers in the model

of nodes per cluster

slots per node

links in the model

clusters in the model

Presentation Model

in the model

pub-lishing units in the model

in the model

sec-tions in the model



2006: size Measures by costagliola

et al (2006)

Costagiola, Di Martino, Ferruci, Gravino,

Tor-toro, and Vitiello (2006) organised their size

measures into two separate categories: length

• CSAPP: Measures the number of client side

scripts and applications

• SSApp: Measures the number of server side

scripts and applications

• IL: Measures the number of internal links

• EL: Measures the number of external

• Scr: Measures the number of scripts

• Lin: Measures the number of links

APPLIcAtIon oF tAxonoMy to surVEyEd sIzE MEAsurEs

This section discusses the literature review sented in the previous section in light of the tax-onomy presented previously In order to provide a more effective discussion, we present the detailed findings in Table 1, followed by a summary of the main findings from the literature review in Table

pre-2 The literature review was based on 10, where

144 measures were proposed in total

Out of the 144 measures proposed for effort

estimation, 47 measures (33%) are early measures;

of these, 10 measures (21%) can only be obtained after a Web application has been designed These results therefore show that, of the 144 measures proposed for Web effort estimation only 37 (26%) can be gathered very early on in the development life cycle, even before a detailed requirements stage These 47 measures were proposed by only three studies (Costagliola et al., 2006; Mangia & Paiano, 2003; Mendes et al., 2003)

Most of the proposed measures are orientated (71%) and length (69%) measures Twenty (77%) measures, out of a total of 26 func-tionality measures, measure functionality using some of the function points analysis methods, and the remaining six base their measurement

solution-on a list of features/functisolution-ons to be provided to customers at the start of the development (Mendes

et al., 2003)

A large number of the proposed size measures (74.4%) relate to the entities Web application or Web application design model, which suggests they can be used for static as well as dynamic Web applications

Only 31 size measures (22%) are bottom-up measures, allowing for the measurement of “parts”

of an application (e.g., Web page, media) The maining size measures (78%) target at the whole application, where application can be represented

re-as Web hypermedia (3%), Web software (0.6%), Web (52.2%), or also represented as a conceptual abstraction using a Web design model (22.2%)

Table 1 Taxonomy applied to all 144 measures

continued on following page