Tool Suppo o he Quali y Assessmen
o MDWE Me hodologies
F.J. Domı
´nguez-Mayo, M.J. Escalona, M. Mejı
´as, and J. Ponce
1 In oduc ion
An in e es ing conce n in de elopmen eams o Web applica ions is how o build
Web applica ions au oma ically, wi h he leas cos and he bes quali y. In his line,
he model-d i en enginee ing (MDE) app oach o so wa e de elopmen ocuses on
c ea ing models o abs ac ions close o some pa icula domain concep s a he
han compu ing concep s. I is mean o inc ease p oduc i i y by maximizing
compa ibili y among sys ems, simpli ying he design p ocess and p omo ing com-
munica ion be ween indi iduals and eams wo king on he sys em.
The ise o his pa adigm has al eady in luenced cu en Web de elopmen s. The
mos impo an esea ch ini ia i e in his a ea is he Model D i en A chi ec u e
(MDA) (OMG: MDA), which is being de eloped unde he umb ella o he Objec
Managemen G oup (OMG). Acco ding o his g oup, models may be independen
om he cha ac e is ics o any echnological pla o m. Besides, he li e cycle o a
so wa e sys em is comple ely co e ed, s a ing om equi emen s cap u e, passing
h ough code gene a ion up o he sys em main enance. In his con ex , Web
enginee ing is a speci ic domain whe e model-d i en so wa e de elopmen can be
success ully applied (Escalona and A ago
´n 2008). The use o MDE in Web
enginee ing is called model-d i en Web enginee ing (MDWE), and as i is no iced
in di e en pape s (ISO- In e na ional O ganiza ion o S anda diza ion, ISO/IEC
9126-1), in he las yea s, se e al esea ch g oups ha e p oposed di e en
me hodologies wi h p ocesses, models and echniques o build applica ions such as
UWE (UML-Based Web Enginee ing), WebML (The Web Modelling Lan-guage),
OOHDM (O ien ed Hype media Design Me hod), OOH4RIA (Melia
´ e al. 2010),
Rux-Me hod (P eciado e al. 2008)o NDT (Na iga ional De elop-men
Techniques). Some o hem in ol e mos o he le els o abs ac ion, and hey
e en ha e ools ha suppo he au oma ion o ans o ma ions in he de elopmen
p ocesses. A sui able design is needed as well as upda ed MDWE me hodologies
and e ec i e ools. To his end, ou wo k ocuses on e alua ing and compa ing
exis ing p oposals, bea ing in mind ha he amewo k could be ex ended o o he
a eas in he u u e. In o de o o e a sui able ool suppo o e alua ing he quali y
o app oaches, his chap e p oposes a ool suppo o QuEF (quali y e alua ion
amewo k) which is a amewo k o he quali y e alua ion o MDWE
me hodologies.
This chap e is o ganized in o he ollowing sec ions. Following his in oduc-
ion, Sec . 2o e s an o e iew o QuEF. Sec ion 3desc ibes he ool suppo
desc ip ion and basic p oposed in e aces oge he wi h he use s’ con ex . Finally,
Sec . 4s a es a se o conclusions and con ibu ions and p oposes possible u u e
wo k.
2 An O e iew o QuEF
QuEF (Domı
´nguez-Mayo e al. 2010a,b,c) is a amewo k o he assessmen o
MDWE me hodologies. This app oach is o ien ed owa ds he quali y e alua ion o
MDWE me hodologies in a speci ic en i onmen by means o objec i e measu es.
Gi en he high numbe o me hodologies a ailable and p oposed in ecen yea s, i
has become necessa y o de ine objec i e e alua ion ools which enable
o ganiza ions o imp o e hei me hodological en i onmen and help Web me hod-
ology designe s o c ea e new e ec i e and e icien ools, p ocesses and
echniques. Due o he p og essi e e olu ion o me hodologies, i is no only
necessa y o e alua e quali y bu also o ind ou how o imp o e and op imize
his p ocess in o de o educe cos s. In his wo k, an app oach o me hodology o
he de elopmen o Web applica ions is a model-d i en p oposal. I p o ides a se
o guidelines, echniques, p ocesses and/o ools o s uc u ing speci ica ions
which a e exp essed as models.
The e a e only conside ed Web modelling app oaches based on MDA in he
amewo k. In addi ion, a amewo k is a basic concep ual s uc u e composed o a
se o elemen s used o e alua e, in his case, MDWE me hodologies, al hough i
could be ex ended o o he a ea o domain. I consis s o he ollowing elemen s:
•Quali y Model. I includes he basis o quali y equi emen s speci ica ions wi h
he pu pose o e alua ing quali y. I speci ies each elemen and i s pu poses.
•Analysis P ocess. I includes he desc ip ion empla es o he e alua ion o
me hodologies and de ines he impo ance o each elemen . I depends on he
Quali y Model desc ip ion.
•E alua ion P ocess. I includes he de ini ion and speci ica ion o ca y ou he
quali y e alua ion p ocess.
•Mul i-C i e ia G oup Decision Making. I includes he necessa y elemen s o
be e he s anda diza ion o he access channel and communica ion among use s
o di e en MDWE me hodologies. I also eaches a consensus in he de ini ion,
he weigh alue impo ance o each de ini ion and weigh alue in luence
ela ionships among he di e en elemen s.
3 The P oposed Tool Suppo o QuEF
Figu e 22.1 shows he gene al sys em a chi ec u e o QuEF-TS (QuEF Tool Sup-
po ) o each componen . In his igu e, he Quali y Model componen is shown
oge he wi h each elemen and he ela ionship among hem. Acco ding o he
In o ma ion Needs desc ip ion, a Fea u e con ains a se o Sub-Fea u es ha ha e a
se o Me ics o measu e hem. Addi ionally, as Quali y Aspec s, Quali y Cha ac-
e is ic includes a se o Quali y Sub-Cha ac e is ics ha ha e o be associa ed o he
Sub-Fea u es.
The Quali y Model componen in QuEF-TS has a se o elemen s o be de ined.
Consequen ly, we ha e o de e mine a se o Fea u es, Sub-Fea u es, Me ics,
Quali y Cha ac e is ics, Quali y Sub-Cha ac e is ics and he ela ionships be ween
hese Sub-Fea u es and Quali y Sub-Cha ac e is ics. All hese elemen s and hei
ela ionships p o ide he basis o he quali y equi emen s speci ica ion and e alu-
a ion in a speci ic domain (in ou case, MDWE). Fea u es, Sub-Fea u es and
Me ics help o de ine he In o ma ion Needs. In o he wo ds, hey con o m o he
in o ma ion equi emen s o he e alua o , clien s and o he pe inen audiences o
Quali y Model Analysis P ocess
E alua ion P ocess
Mul i-C i e ia G oup Decision Making
Quali y Model::
Fea u e
Quali y Model::
Sub-Fea u e
Quali y Model::
Quali y
Cha ac e is ic
Quali y Model::
Quali y
Sub-Cha ac e is ic
Quali y Model::
Me ic
Analysis P ocess:
:Templa e
Analysis P ocess:
:Ma ix o
In luences
Analysis P ocess:
:MDWE
Me hodology
E alua ion
P ocess::Quali y
E alua ion
Mul i-C i e ia G oup
Decision Making::
Thesau us & Glossa y
Mul i-C i e ia G oup
Decision Making::
Consensus weigh
alues
Analysis P ocess:
:Weigh alues
1..*
1
1..* 1..* 1..* 1..*
1..*
1
1..*1..*
1..*
1
«use»
«use»
«use»
«use»
1
1..*
Fig. 22.1 The gene al sys em a chi ec u e o QuEF ool suppo
be me by means o he e alua ion. Quali y Cha ac e is ics and Quali y Sub-
Cha ac e is ics de ine he quali y aspec s o he equi ed model. Finally, hese
In o ma ion Needs ha e o be mapped wi h he quali y aspec s.
•Quali y Cha ac e is ic. This is a highe -le el quali y aspec . Fo ins ance,
Usabili y, Main ainabili y, Func ionali y, Reliabili y o Po abili y could be
conside ed Quali y Cha ac e is ics.
•Quali y Sub-Cha ac e is ic. In a hie a chy o quali y aspec s, his is a lowe -
le el quali y aspec . Fo ins ance, Usabili y is de ined h ough se e al Quali y
Sub-Cha ac e is ics, such as Lea nabili y, Unde s andabili y and Ope abili y.
Each cha ac e is ic in ISO/IEC 9126 (ISO- In e na ional O ganiza ion o
S anda diza ion, ISO/IEC 9126-1) and ISO/IEC 25000 (SQuaRE) is desc ibed in
ela ion o a so wa e p oduc . Howe e , in ou pa icula case, all Quali y
Cha ac e is ics and Quali y Sub-Cha ac e is ics a e desc ibed ega ding app oach
cha ac e is ics.
In MDWE, models a e p og essi ely e ined and ans o med in o new models o
codes. To his end, ools may also be use ul o es , e i y o alida e models.
Mo eo e , each me hodology may de ine i s de elopmen p ocess and/o
echniques. The quali y o me hodologies subsequen ly depends on he di e se
Fea u es, such as he MDE, he Web Modelling, he ma u i y o a me hodology
and he ool suppo applied o disco e aul s o weaknesses. The main idea is o
cha ac e ize he whole MDWE p ocess. The elemen s desc ibing he In o ma ion
Needs in he Quali y Model a e:
•Fea u e. This is a highe -le el desc ip ion concep o an app oach. I may be, o
example, he so wa e de elopmen p ocess, models, me amodels, languages,
ools o he ans o ma ions used.
•Sub-Fea u e. This is a lowe -le el desc ip ion concep o an app oach. Fo
example, he model-d i en enginee ing is a Fea u e wi h se e al Sub-Fea u es,
such as he Language De ini ion, T ans o ma ions and T ace.
•Me ic. In he Quali y Model, me ics should measu e he deg ee o which a Sub-
Fea u e eaches he equi ed model. To simpli y, a me ic is used o measu ing
Sub-Fea u es.
The Analysis P ocess componen would include he empla es which ha e o be
gene a ed in e ms o Quali y Model desc ip ion, speci ically he In o ma ion Needs
desc ip ion. The me hodology can be analysed wi h hese empla es conside ing he
Weigh alue de ini ion o Sub-Fea u es and Quali y Sub-Cha ac e is ics and he
Ma ix o In luences (MoI), which indica es he in luence in Sub-Fea u e and
Quali y Sub-Cha ac e is ics. Finally, he E alua ion P ocess componen pe o ms
he quali y e alua ion, and he Mul i-C i e ia G oup Decision Making is esponsible
o eaching a consensus on he e minology and weigh alues among use s.
Di e en basic in e aces a e p oposed and gi en in his chap e .
As a as use con ex is conce ned, his ool suppo would ha e h ee di e en
use con ex s o QuEF-TS. As i is shown in Fig. 22.2, he i s one would be he
De elope use , who canno modi y any hing. I would ha e a gene al iew o he
Quali y Model, he Analysis P ocess and he E alua ion P ocess. The second one
would be he Designe use , who would ha e mo e pe mission and could no only
iew bu also c ea e and edi he Analysis P ocess componen s such as he MoI, he
Weigh alues o Quali y Sub-Cha ac e is ics and Sub-Fea u es and Templa e
alues. This use could also log in he ool suppo . Thus, i would sa e di e en
MoIs, Weigh alues and Templa es and exchange hem wi h o he Designe s. I
would be use ul o e alua e me hodologies acco ding o o he designe s’ poin s o
iew. Finally, he hi d use would be he Adminis a o who would c ea e and edi
he Quali y Model and gene a e he empla es based on i in o de o be used in he
Analysis P ocess.
QuEF-TS can gene a e empla es and MoIs in e ms o he Quali y Model
speci ied. Thus, an impo an issue is o de ine he Quali y Model and i s
knowledge.
C ea e and Edi
Quali y Model
De elope
Designe
Adminis a o
Log in
Gene a e
Te mpla e s
View E alua ion
P ocess
C ea e and Edi
Analysis P ocess
View Analysis
P ocess View Quali y Model
«include»
«ex end»
«ex end»
Fig. 22.2 Use cases o he p oposed ool suppo
3.1 The Quali y Model
Figu e 22.3 ep esen s he basic in e ace o In o ma ion Needs. This in e ace
shows how he Fea u es, Sub-Fea u es and Me ics a e de ined. In o de o build
he di e en elemen s, a ee s uc u e is p oposed and used by he Adminis a o
wi h he aim o c ea ing and edi ing Fea u es, Sub-Fea u es and Me ics in e ms o
he In o ma ion Needs es ablished by he communi y o designe s. The de ini ion o
his hie a chy is cu en ly being de eloped, al hough we a e also wo king on Mul i-
C i e ia G oup Decision Making (MCGDM) (P eciado e al. 2008) o each a
consensual de ini ion o he hie a chy.
A se o Quali y Cha ac e is ics and Quali y Sub-Cha ac e is ics o each Quali y
Cha ac e is ic based on cu en li e a u e, such as ISO/IEC 9126 o o he s anda ds
which a e adap ed o MDWE me hodologies, a e iden i ied, classi ied and placed in
a hie a chy. Likewise, he Quali y Cha ac e is ics and Quali y Sub-Cha ac e is ics
a e de ined in he Quali y Model, as shown in Fig. 22.4. On i s le side, his igu e
shows a ee s uc u e whe e he use would c ea e and edi Quali y Cha ac e is ics
and Quali y Sub-Cha ac e is ics in e ms o he Quali y Aspec s conside ed. Fu -
he mo e, in o de o de ine a Quali y Model, i mus con ain associa ion links
be ween he Sub-Fea u es and he Quali y Sub-Cha ac e is ics. These links ep e-
sen he dependencies be ween Sub-Fea u es and Quali y Sub-Cha ac e is ics and
show how each Sub-Fea u e a ec s e e y Quali y Sub-Cha ac e is ic. Associa ion
links may be based on p o en eal-wo ld expe ience o de e mined by eal case
s udy applica ions in a numbe o eal p ojec s. We a e cu en ly wo king on a
MCDGM me hod which could pe mi eaching a consensus alue acco ding o
di e en expe s’ iews.
Fig. 22.3 Basic in e ace o in o ma ion needs desc ip ion
3.2 The Analysis P ocess
In his componen , he associa ion links be ween Sub-Fea u es and Quali y Sub-
Cha ac e is ics a e de ined in he MoI, which indica es, by means o a igu e o
alue, how e e y Sub-Fea u e a ec s Quali y Sub-Cha ac e is ic.
Fo ins ance, Usabili y is desc ibed as a se o Quali y Sub-Cha ac e is ics. As
shown is Fig. 22.5, hese Quali y Sub-Cha ac e is ics could be a ec ed by one o
he Sub-Fea u es. This igu e also ep esen s he basic in e ace o he MoI. Each
alue indica es he in luences be ween Sub-Fea u es and Quali y Sub-
Cha ac e is ics appea ing in his igu e o e e y Sub-Fea u es o MDE, such as
Le els o Abs ac ion, S anda d De ini ion, Model-Based Tes ing, T ans o ma ions
and T aces, as well as o Quali y Sub-Cha ac e is ics o Usabili y such as
Lea nabili y, Unde s andabili y, Simplici y and In e p e abili y.
The empla es in he Analysis P ocess a e based on he Quali y Model de ini ion,
and hey a e used o desc ibe an inpu me hodology. Figu e 22.6 ep esen s he
sugges ed basic in e ace. These empla es would be used as inpu in he en i on-
men , analysed in he e alua ion p ocess and compa ed wi h he equi ed quali y
model o he Quali y Model de ini ion. Templa es o MDE, Web Modelling, Tool
Suppo and Ma u i y ha e al eady been de eloped. Finally, he E alua ion P ocess
would con as he in o ma ion om each inpu app oach empla e wi h he in o -
ma ion om he Quali y Model de ini ion. The main e alua ion pu pose is o
iden i y ade-o s and sensi i i y poin s o he me hodology unde s udy. The
aim is o de e mine which aspec needs o be imp o ed on a MDWE me hodology.
Each Sub-Fea u e has o de ine a se o quan i a i e o quali a i e me ics used o
measu e each Sub-Fea u e. In hese lines, i migh be in e es ing o es ablish
s anda d me ics on MDWE o cen alize hem. In he li e a u e, nume ous
Fig. 22.4 Basic in e ace o quali y aspec s desc ip ion
e e ences o me ics can be ound, bu s anda diza ion has no been ca ied ou ye .
The ypes o me ic de ined a e ei he quali a i e o quan i a i e. Quali a i e me ic
ange indica es i he me ic is Suppo ed, Pa ly Suppo ed o No Suppo ed, as
appea ed in Fig. 22.6.
On he con a y, Fig. 22.7 o e s he Weigh alues se ing o he In o ma ion
Needs and Quali y Aspec s. Weigh alue ep esen s he impo ance o each Sub-
Fea u e in he se o Sub-Fea u es o a Fea u e. Simila ly, a weigh alue is also
used o de ine he impo ance o each Quali y Sub-Cha ac e is ic in he se o
Quali y Sub-Cha ac e is ics o a Quali y Cha ac e is ic. MGDCM me hods a e
being cu en ly de eloped o de ine weigh alues o Sub-Fea u es and Quali y
Sub-Cha ac e is ics.
3.3 The E alua ion P ocess
The E alua ion P ocess is he componen which use s can e alua e whe he hey
ha e eached he equi ed model o quali y o no . The p oposed basic in e ace, as
shown in his example o Fig. 22.8, conside s he Sub-Fea u es o Web Modelling.
Fig. 22.5 Basic in e ace o he ma ix o in luences (MoI)
We can see he Sub-Fea u e alues o he Web Concep ual Le els, In e aces,
De elopmen P ocess, Con en Modelling, P esen a ion Modelling, Na iga ion
Modelling and Business Modelling. All o hem a e Web Modelling Sub-Fea u es.
In o he wo ds, he iews o all Sub-Fea u e alues o a speci ic Fea u e o a iew o
all Fea u e alues in gene al can be shown wi h his in e ace.
Quali y Cha ac e is ics and Quali y Sub-Cha ac e is ics a e shown in he same
way ha Fea u es and Sub-Fea u es appea ed, al hough in his in e ace, he use
has o indica e which Fea u es has o be conside ed in o de o calcula e he Quali y
Sub-Cha ac e is ics and Quali y Cha ac e is ics.
The eason is ha hese alues depend on he in luences ha Sub-Fea u es ha e
on Quali y Sub-Cha ac e is ics. In o he wo ds, he alue de ini ion o Quali y
Cha ac e is ics and Quali y Sub-Cha ac e is ics is de e mined by he ela ionship
(s) be ween he In o ma ion Needs and he Quali y Aspec s. Figu e 22.9 ep esen s
he basic in e ace o Quali y Aspec s and p o ides a gene al iew o Quali y
Cha ac e is ics. We can also ob ain all Quali y Sub-Cha ac e is ic alues o a
Quali y Cha ac e is ic by selec ing he app op ia e Da a alue op ion. In his
example, we can obse e all Quali y Cha ac e is ic alues in e ms o Web
Modelling, MDE and Ma u i y, which a e he Fea u es ha ha e been selec ed in
“In e ms o ” op ion.
Fig. 22.6 Basic in e ace o he empla es
