Treść książki
Przejdź do opcji czytnikaPrzejdź do nawigacjiPrzejdź do informacjiPrzejdź do stopki
14
EdytaAbramek,TomaszWachowicz
thetruebehaviorandthesequenceoffunctionalitiestheusersfocusonbyana-
lyzingthepaththeireyes“draw”inthewebsitelayout.Theresultsoftheeye-
-trackinganalysisallowedtheNielsenNormanGrouptoelaboratethebestprac-
ticesinbuildingandusingthewebsites[Kaus09,p.92-93].
2.2.MCDMmethodologiesforbuildingthewebsitesranking
Nomatterwhichoftheorganizingtechniquesdescribedaboveis
selected,finallytheevaluationrequiresthemulti-attributecomparisonofthe
websitesthemselves.Thelistofsupportivetoolsforsuchanevaluationislong
andconsistsofthemethodslikeclassicMAUTmodels[KeRa76],ELECTRE
[Roy96],PROMETHEE[BVMa86,p.228-238],AHP[Saat80],TOPSIS
[HwYo81]orhybridinteractivemethods[Nowa06,p.1413-1430].Most
ofthesetoolsaredescribedindetailinthesummarizingworkbyFigueira
[Figu04].Inthisworkwepointtoonlyfourselectedanalyticmethodsthatbase
ondifferentassumptionsandapplydifferentalgorithms.Itwillallowusto
analyzetheimpactofthesealgorithmsonthefinalrankingofalternatives.
Additivescoringmodel(ASM)
AMS[KeRa76]allowstoscorethealternativesthataredescribedin
adiscretewayintermsofthepredefinedcriteria.ItisthemostpopularMCDM
method,claimedtobeeasytouse.Howeveritisbasedonmanytiresomescores
assignmentsthatrequirebasicmathematicalskillsandanelementaryknowl-
edgeonthetheoryofdecisionmaking.Itsdisadvantageisthatthescoresare
simplyassignedinsteadofbeingdetermined[FoSe01].
ThemainideaofAMSistoscoreeachoptionaccordingtodecision
maker’s(DM)subjectivepreferencesbyusinganartificialcriterionlikeutility.
Theprocessofscoringoffersrequires:
1.Assigningtheweights
wforeachcriterion
i
i
=
1
,...,
I
,suchas:
∑
i
w
i
=
P
,
(1)
wherePisthetotalpoolofscoringpoints(usuallyapoolof100pointsis
used).
2.Assigningscorestoeachoptionwithineachcriterion.DMassignsthescor-
ingpointstoeachoption(xjk)describingtheperformanceofalternativejfor
thecriterionkuptothelimitdefinedbytheweightofthiscriterion:
u
(
x
jk
)
∈
[
0
,
w
k
]
.
(2)
