1
Weigh ed Shannon en opy as a quan i a i e measu e o açade
pa e n egula i y
Michał Malewczyk1*
1Facul y o A chi ec u e, Gdansk Uni e si y o Technology, Gdansk, Poland
*Facul y o A chi ec u e, ul. Gab iela Na u owicza 11/12, 80-233 Gdańsk, Poland.
Email: [email p o ec ed]; ORCID: 0000-0001-9585-1497
The a icle p esen s an o iginal measu e o acade pa e n egula i y, SVE
(Shannon-Va iance-En opy), based on weigh ed Shannon en opy, which aims o
quan i a i ely desc ibe he in e nal di e si y o a composi ion. The analysis
showed ha simply aking in o accoun he dis ance be ween elemen s does no
imp o e he accu acy o he egula i y desc ip ion. The a iance o elemen
p opo ions and angula ela ionships p o ed o be key ac o s, signi ican ly
inc easing he indica o 's consis ency wi h aes he ic pe cep ion. The bes esul s
we e ob ained o he composi e measu e SVEDAS, which combines all h ee
componen s. This measu e shows high consis ency wi h expe assessmen s and
e ec i eness in he algo i hmic gene a ion o acade pa e ns. The use o he
designa ed me ic as a design pa ame e enables quan i a i e con ol o aes he ics,
suppo ing sus ainable design in line wi h SDG 11.
Keywo ds: u ban planning; sma ci ies; u ban de elopmen ; ci y; da a-d i en
design
Decla a ions: no applicable.
Funding: no applicable.
Con lic s o in e es /Compe ing in e es s: he e a e no compe ing in e es s.
A ailabili y o da a and ma e ial: on demand.
Code a ailabili y: no applicable.
E hics app o al: no applicable.
Consen o pa icipa e: no applicable.
Consen o publica ion: sensi i e da a was no collec ed.
2
Weigh ed Shannon en opy as a quan i a i e measu e o açade
pa e n egula i y
In oduc ion
The aes he ics o u ban space, especially he composi ion o a chi ec u al
elemen s on building açades, play a key ole in he pe cep ion o he quali y o he
u ban en i onmen (Malewczyk e al. 2022a, 2024; Samala ičius 2021). The egula i y
and geome ic o de o açades, especially he a angemen o windows, a ec no only
he isual expe ience o esiden s and use s o public spaces, bu also he sense o o de ,
secu i y and iden i y o a place (Ezz e al. 2024; Wang & Munaka a, 2024). The pu sui
o aes he ically pleasing, o de ly and isually iendly u ban spaces is di ec ly in line
wi h Sus ainable De elopmen Goal 11 (SDG 11), which calls o he c ea ion o ci ies
and neighbo hoods ha a e sa e, inclusi e, sus ainable and o high quali y. Aes he ics
and a chi ec u al legibili y, including he geome ic egula i y o açades, suppo he
pe cei ed quali y o u ban li e, a sense o local iden i y and com o in he use o public
space (Webe e al. 2008).
Despi e g owing in e es in he p oblem o egula i y in he con ex o he aes he ics o
he buil en i onmen and i s ela ionship o he b oade psychophysical well-being o
he use s o ha en i onmen , he e a e s ill objec i e and unambiguous ma hema ical
ools ha would allow quan i a i e assessmen o he geome ic o de ing o açade
pa e n. Malewczyk's publica ion om 2025 esponded o his challenge by p oposing
he i s e sion o an index desc ibing he egula i y o pa e ns, consis ing o
ec angles (symbolizing windows o açade panels) in a wo-dimensional açade space.
The au ho 's index Hcen e dis co esponds o he en opy o he me ic dis ance be ween
he cen e s o adjacen composi ional elemen s. Despi e e y high e iciency in he
p edic ion o he deg ee o egula i y by Hcen e dis (co ela ion o abou 93%), in he
3
au ho 's opinion, he e emains some ield ha needs imp o emen .
The main objec i e o his pape is o de e mine a modi ied açade pa e n egula i y
measu e index (SVE - Shannon-Va iance-En opy), based on he o iginal H(cen e dis ). The
new app oach in ol es he in oduc ion o me ic en opy weigh ing, which should
inc ease he sensi i i y o he index o mo e sub le changes in composi ion egula i y.
An addi ional goal o he pape is o de elop a gene a i e algo i hm and e i y he
easibili y o designing a composi ion by op-down de e mina ion o i s egula i y as a
po en ial ool o suppo sus ainable a chi ec u al design in line wi h SDG 11 goals.
The a icle is based on wo esea ch hypo heses. Fi s , he weigh ed me ic en opy
index o he dis ance be ween he cen e s o adjacen composi ional elemen s o a
açade design desc ibes he egula i y o he design be e han pu e me ic en opy.
Second, he de e mined index o weigh ed me ic en opy can be e ec i ely used as an
objec i e unc ion in açade composi ion pa e n gene a ion algo i hms.
This pape a emp s o answe ou esea ch ques ions. Fi s , how o ma hema ically
de ine an op imal unc ion ha combines Shannon en opy wi h measu es o spa ial
a iance? Second, wha açade composi ion pa ame e s (sizes o composi ional
elemen s, spacing, loca ion) mos s ongly in luence he de eloped SVE index? Thi d,
how does he SVE index co ela e wi h he egula i y o açade pa e ns, and is his
co ela ion s onge han he o iginal Hcen e dis index? Fou h, is i possible o
algo i hmically gene a e açade pa e ns h ough a op-down de e mina ion o he SVE
pa ame e ?
This a icle builds on he ea lie s udy, howe e , based on an expanded esea ch sample
o 343 (o iginal 245) composi ions based on 49 esiden ial, mul i- amily buildings
nomina ed o he 2024 Eu opean Union P ize o Con empo a y A chi ec u e- Mies
4
Van de Rohe Awa d (EU Mies Awa d 2024) in he 'collec i e housing' ca ego y. A
new s a is ical analysis me hodology is also p oposed.
The es o he a icle p esen s an analysis o key concep s in he con ex o he cu en
s a e o knowledge. The au ho ocuses on measu es o egula i y in a chi ec u e,
en opy in e ms o in o ma ion heo y in e ms o spa ial analysis, and a e iew o ways
o algo i hmically gene a e composi ions.
Measu es o egula i y in a chi ec u e
In he analysis o a chi ec u al o ms, quan i a i e me hods o objec i ely assess hei
egula i y and o de ha e become inc easingly impo an . Resea ch on egula i y
coe icien s o açades ocuses, among o he hings, on he analysis o dis ances
be ween elemen s, hy hmici y and he occu ence o epe i i e geome ic pa e ns
(S iny & Mi chell 1978; Ilgaz e al. 2025; Med'dahi & Bousso a 2021).
Applica ions o ac al geome y o açade analysis a e also inc easingly epo ed in he
a chi ec u al li e a u e, indica ing a link be ween geome ic complexi y and he aes he ic
and en i onmen al pe cep ion o buildings (Bo ill 1996; Salinga os 1998). Measu es
such as ac al dimension o complexi y indices ha e been used o desc ibe he s uc u e
o his o ic u ban complexes and mode n açades, demons a ing hei po en ial as a ool
o assessing he quali y o space (Ali & Mus a a 2024; Lee 2014; Ka ona 2023).
Measu es o symme y and p opo ion, de i ed om classical heo ies o a chi ec u al
composi ion (Wi kowe 1949), a e also an impo an ca ego y o analy ical ools.
Symme y, bo h mi o and ansla ional, plays an impo an ole in he pe cep ion o
egula i y, as con i med by psychological esea ch and pe cep ual heo y (Jacobsen
2006; Lede e al. 2004). These wo ks indica e ha s uc u es wi h a high deg ee o
5
egula i y can be mo e easily ecognized and posi i ely e alua ed by space use s. A ich
li e a u e analysis o he issue o egula i y measu es in b oade con ex s was also made
in Malewczyk's 2025 a icle.
Shannon's en opy and i s applica ions
En opy is a concep ha was i s in oduced in he 19 h cen u y in
he modynamics o measu e he chao ic na u e o sys ems. In 1948, Claude Shannon
also began, h ough en opy (acco ding o Fo mula 1), o de e mine he a e age amoun
o in o ma ion in a message, a concep ha is pa icula ly ele an in he con ex o
in o ma ion heo y (Shannon 1948; Fou ie 2012).
𝐻=−$𝑝!×𝑙𝑜𝑔"𝑝!
#
!$%
Fo mula 1 En opy in in o ma ion heo y.
Since i s in oduc ion, Shannon's en opy has ound widesp ead use in a a ie y o ields
- including he desc ip ion o he design p ocess (K us 2013), medical diagnosis (Benish
2020), and esea ch on he ela ionship be ween di e si y and aes he ic p e e ences
(Redies e al. 2017; S anischewski e al. 2020; S amps 2004). The Shannon global
en opy coe icien is also widely used in buil en i onmen analysis (Güzelci e al.
2020, 2021; Güzelci & Alacam 2019; C omp on 2012; S amps 2012, 2014).
Based on me ic en opy, which is independen o he leng h o in o ma ion, he
Hcen e dis index was de eloped (Malewczyk, 2025). I is a measu e o he me ic en opy
o a sequence o med om he absolu e dis ances be ween he cen e s o composi ional
elemen s (e.g., windows, açade panels) on a building açade. A s udy o 245
composi ions based on 50 buildings nomina ed o he EU Mies Awa d 2024 in he
‘collec i e housing’ ca ego y ound a 93% co ela ion be ween he Hcen e dis index and
6
objec i e isual egula i y. Howe e , Malewczyk (2025) poin s ou he limi a ions o
he index, which does no conside he magni ude o dis u bance, only he p esence o
dis u bance, due o Shannon's en opy p ope y, which is sensi i e o a ia ion bu no o
he magni ude o change. Fo example, he s ings 1,3,7,3,5 and 2,35,192,79,192 ha e
he same en opy, despi e ma kedly di e en dispe sion. Mo eo e , his indica o does
no ake in o accoun he size o he composi ional elemen s.
This p oblem can be sol ed by weigh ed en opy, p oposed by Guiaşu (1971), which
conside s no only he p obabili y o e en s, bu also hei impo ance (e.g.,
signi icance). An example o i s applica ion is in in es men isk analysis (Naw ocki
and Ha ding, 1986). Weigh ed en opy is being de eloped bo h heo e ically (Ebanks
2010) and p ac ically, including in compu e science Kelbe e al. 2017) and u ban
scale buil en i onmen analysis (Boeing 2019; Huynh 2019; Wang 2018). Howe e , no
examples o i s applica ion in he analysis o a chi ec u al o ms using me ic en opy
ha e ye been ound.
𝐻=−$𝑤!×𝑝!×𝑙𝑜𝑔"𝑝!
#
!$%
Fo mula 2 En opy in in o ma ion heo y.
Algo i hmic açade gene a ion
O e he pas wo decades, algo i hmic açade gene a ion me hods ha e become
an impo an a ea o a chi ec u al esea ch and expe imen a ion, in eg a ing pa ame ic,
op imiza ion, en i onmen al and a i icial in elligence app oaches. O key impo ance is
pa ame ic design, which enables dynamic con ol o e he geome y, s uc u e and
pe o mance o a açade using ools such as G asshoppe , Ladybug o Oc opus (Cae ano
2020; Na ange el & S ou s 2016; Shen 2018; Ramadan 2024). This makes i possible
7
o simul aneously es and op imize o ms o dayligh ing, shading, na u al en ila ion,
and ene gy p oduc ion (Na ange el & S ou s 2016; Lahma e al. 2022; Chen & Tang
2024).
In he con ex o he goals o SDG 11 and he opic o his a icle, esea ch on he use o
algo i hmic me hods o gene a e açades wi h aes he ic quali y is pa icula ly ele an .
Rezakhani and Kim (2024) used gene ic algo i hms o op imize açades o aes he ics
and isibili y while main aining app op ia e sunligh pa ame e s. Alagöz and Jamal
(2024) used pa ame ic design s a egies in de eloping kine ic açades, s i ing o a
balance be ween unc ionali y and aes he ics. Simila issues a e also add essed by
Mahmoud and Elghazi (2016) and o he au ho s (Engin e al. 2023; Wang, Sun, Shao &
He 2024; Wang, Zhang, Zhang, Cui & He 2024; Cudzik 2019; Cudzik & A asoy 2023;
Cesbas e al. 2022).
Mo e ecen ly, he e ha e also been a emp s o apply a i icial in elligence o he design
p ocess, including h ough he use o he Mon e Ca lo T ee Sea ch (MCTS) algo i hm
in he ea ly s ages o design (Lin e al. 2025), as well as GAN algo i hms o gene a e
açades om images (Yu e al. 2020; Shan & Zhang 2022) and o he inpu s (Wan e al.
2023). Al hough hese algo i hms mainly ocus on op imizing selec ed aspec s, such as
sola iza ion, ma e ial consump ion o ene gy e iciency wi h aes he ics, none o he
publica ions ound a emp ed o quan i a i ely encode açade aes he ics.
Ma e ials and me hod
The p esen s udy is an expansion o Malewczyk's 2025 s udy. To inc ease he
sensi i i y o he s udy, he au ho pa ially uses, bu also pa ially expands he o iginal
esea ch sample. Modi ied s a is ical analysis p ocedu es a e also used o imp o e he
quali y o he esul s.
8
Ma e ials
The s udy is based on ma e ials, ha is, 245 composi ions o ec angles
symbolizing windows o açade panels, on which he o iginal Malewczyk (2025) s udy
was based. The composi ions we e based on 50 pho og aphs o mul i amily buildings
ha we e nomina ed in 2024 o he Eu opean Union P ize o Con empo a y
A chi ec u e- Mies an de Rohe Awa d (EU Mies Awa d 2024) in he 'collec i e
housing' ca ego y. In he o iginal s udy, pho og aphs we e downloaded om he
compe i ion o ganize 's websi e (h ps://miesa ch.com/a chi e). Basing he composi ion
on examples o a chi ec u e nomina ed o such a p es igious awa d was o ensu e he
high quali y o he esea ch ma e ial.
Me hods
Weigh ed en opy
The o iginal Hcen e dis index in Malewczyk's 2025 s udy was de e mined
based on me ic en opy, which was calcula ed in a simila way o in he wo ks o
S amps (2004, 2014), C opmp on (2012), Güzelci (2019, 2021) o S anischewski
(2020). Howe e , in o de o inc ease he sensi i i y o he de e mined index, he
heo e ical basis o Shannon's concep o weigh ed me ic en opy, which was i s
w i en abou by Guiaşu in 1971, is p oposed. O ele ance o his a icle, om he
pe spec i e o he p ac ical applica ion o weigh ed en opy, a e he wo ks o Boeing
(2019), Huynh (2019) and Wang and Zhao (2018). These wo ks, al hough ela ed o
u ban p oblems, make use o he au ho 's impo an idea o weigh ed en opy wi h
geome ic ela ions.
9
So wa e
The s udy mainly used Blende e sion 4.3.2 o macOS. This p og am was used
o c ea e and p ocess he s udy ma e ial and also o un sc ip s in Py hon e sion 3.12.
The sc ip s used in he s udy used lib a ies such as bpy, bmesh, andom, os, ma h, cs
and io.
C ea ing composi ions
In he o iginal s udy, 49 wo-dimensional composi ions we e c ea ed om
pho os o buildings nomina ed o he EU Mies Awa d 2024 in he 'collec i e housing'
ca ego y. Fo one building, a composi ion was no c ea ed. In a u he s ep, 4 a ian s
we e c ea ed based on each o he 49 o iginal composi ions, which clea ly di e ed om
each o he and om he o iginal composi ion in he deg ee o isual egula i y. This was
achie ed by using such p ocedu es as b eaking symme y, a ying he dis ance be ween
composi ional elemen s, a ying composi ional elemen s among hemsel es and he
like. In he end, 49 g oups o composi ions we e o med, each o which had 5
composi ions. A o al o 245 composi ions we e c ea ed.
The cu en s udy uses 245 composi ions om Malewczyk's 2025 s udy (sou ce iles
a ailable h ps://doi.o g/10.34808/bemn- d97) howe e , an addi ional 2 composi ions
we e c ea ed wi hin each o he 49 g oups. The pu pose o c ea ing he addi ional
composi ions was o en ich he s udy sample wi h specimens ha di e ed in isual
egula i y in a mo e sub le way, in o de o be able o e i y he sensi i i y o he new
weigh ed en opy-based indices. In he end, 343 composi ions (245 o iginal and 98 new)
we e collec ed o he s udy. The p ocess o c ea ing composi ions using a g oup o 12
composi ions as an example is isualized in Figu e 1.
16
Resul s
Analysis o i e di e en o mulas o he au ho 's Shannon-Va iance-En opy
(SVE) index showed signi ican di e ences in hei abili y o p edic he egula i y o
building açade pa e ns. The i e di e en weigh ing combina ions we e es ed agains
he o iginal Hcen e dis index (Malewczyk, 2025) in 49 composi ion g oups, each o which
con ained se en pa e ns anked acco ding o he REG scale ( om 1 o 7, wi h 1
indica ing he g ea es egula i y and 7 indica ing he g ea es i egula i y). Th ee
complemen a y measu es-Spea man's ank co ela ion coe icien , he pe cen age o
co ec ly o de ed pai s, and he sum o squa es o ank di e ences-we e used o assess
he pe o mance o he indica o s. These measu es we e chosen o accoun o he
o dinal na u e o ank o de egula i y (REG) and o a oid he limi a ions o linea
co ela ion assump ions (as in he case o he -Pea son co ela ion coe icien ).
Table 1 Summa y o esul s o s a is ical analyses.
Measu e
Spea man ρ
Co ec Pai s (%)
Sum o Squa ed Rank Di e ences
Hcen e dis
.929
88.7%
5.3
SVED
.871
90.0%
6.9
SVEA
.919
92.1%
4.8
SVES
.946
93.2%
3.3
SVEDS
.940
94.4%
3.1
SVEDA
.949
95.3%
2.8
SVEAS
.958
96.2%
2.5
SVEDAS
.964
96.8%
2.0
A compa ison o he pe o mance o he SVE coe icien a ian s agains he o iginal
Hcen e dis coe icien is shown in Table 1. The esul s show ha SVEDAS a iance
weigh ing o dis ance, p opo ion and angles) achie es he highes pe o mance in all
s a is ical indica o s wi h a Spea man co ela ion o .964, a co ec o de ing o 96.8%
and he lowes sum o squa es o ank di e ences (2.0).
17
Co ela ion o indi idual a iance componen s
The analysis e ealed di e en ial e ec s o indi idual a iance componen s on
he e ec i eness o he SVE index. SVE(D) (ρ=.871) weigh ed by dis ance a iance
(CVdis ance) shows he lowes e ec i eness, e en lowe han he baseline Hcen e dis index
(ρ=.929). An aspec ha was comple ely igno ed in he o iginal s udy by Malewczyk
(2025) is he a iabili y o he angles o med be ween he connec ions o he cen e s o
adjacen ec angles. The SVEA weigh ed CVangle ob ained a Spea man co ela ion
coe icien o ρ=.919 and is clea ly supe io o he baseline Hcen e dis in e ms o he
pe cen age o co ec ly o de ed pai s (92.1% s. 88.7%). Howe e , he mos signi ican
componen is he a ia ion in he p opo ion o ec angles (CVshape). SVES, weigh ed by
CVshape ob ained a Spea man co ela ion coe icien o ρ=.946.
Ve i ica ion o esea ch hypo heses
The s a is ical analyses pe o med on he da a collec ed in he conduc ed s udy
made i possible o e i y bo h esea ch hypo heses and a he same ime answe all ou
esea ch ques ions.
Rega ding hypo hesis one, which assumed ha weigh ed me ic en opy be e desc ibes
he egula i y o he ele a ion pa e n han unweigh ed en opy, i was con i med o
some o he a ian s o he SVE index. The combined SVEDAS index pe o med 3.5
poin s be e han he baseline Hcen e dis index and o de ed 8.1 % poin s mo e pai s
co ec ly. Thus, i was shown ha all aspec s- a iabili y o dis ances, p opo ions and
sizes o composi ional elemen s and angles in he connec ion g id-a e impo an in
egula i y analysis.
The high e iciency o he SVEDAS index (ρ=.964), he e y high pe cen age o
co ec ly o de ed pai s (96.8%) and i s simplici y indica e he g ea po en ial o his
18
measu e as an objec i e unc ion in gene a i e algo i hms. In he cou se o he esea ch,
such a gene a i e ool was de eloped, whose mechanism o ope a ion was based
di ec ly on he SVEDAS indica o . The ool was de eloped as an add-on o Blende and
made a ailable unde an open-sou ce license. The ool de ini ions o six composi ion
pa e ns c ea ed by Malewczyk, Ta aszkiewicz and Czyż (2022b) as a composi ion
ypology o açades o esiden ial, mul i- amily buildings. The ool allows you o selec
he speci ic ype o composi ion you wan o gene a e (acco ding o he pa e n
de ini ion), speci y he egula i y by p o iding expec ed nume ical alues o he
in e al o he SVEDAS index and o he pa ame e s, such as he numbe o windows in
ows and columns, window ypes, dis ances be ween window ows, heigh s o in e s o y
s ips and o he basic physical aspec s. Sc eensho s showing he use in e ace in
a ious con igu a ions a e shown in Figu e 3.
Fig. 3 Use in e ace o he de eloped acade composi ion gene a o .
19
Thus, a ool has been ob ained ha allows he c ea ion o an in ini e numbe o
ele a ion pa e ns wi h a gi en alue o egula i y, unde s ood p ecisely as he alue o
he SVE index, which con i ms i s use ulness no only as a diagnos ic measu e, bu also
as an objec i e unc ion in composi ion algo i hms.
Discussion
The analysis o SVE (Shannon Va iance-En opy) a ian s e eals he
complexi y o he pe cep ion o egula i y in acade pa e ns and he po en ial o
ad anced measu es in hei quan i a i e desc ip ion. The esul s con i m ha al hough
simple en opy (Hcen e dis ) is highly e ec i e, i s pe o mance can be signi ican ly
imp o ed by conside ing he in e nal a iabili y o he pa e n.
I has been shown ha dis ance a iance (CVdis ance) as a s andalone componen no only
ails o imp o e he accu acy o egula i y desc ip ion bu may e en wo sen i . This
sugges s ha in o ma ion abou he dis ances be ween elemen s (e.g., he cen e s o
g a i y o acade panels) does no cons i u e a su icien ly clea pe cep ual signal. This
may be since local symme y o hy hm o en occu despi e signi ican de ia ions om
equal dis ances.
In u n, he a iance o angles be ween lines connec ing he cen e s o composi ional
elemen s, al hough mode a ely e ec i e as a s andalone componen , has a s ong e ec
when combined wi h o he ea u es. This poin s o he impo ance o di ec ional
ela ionships in pa e n pe cep ion, especially in mo e complex and i egula
a angemen s. Angula di e ences can ac as ‘ isual dis u bances’ o hy hm and
con olling hem a ec s he pe cep ion o o de .
The s onges componen was ound o be he a iance in he p opo ions o elemen s
(CVshape). I s high e ec i eness, bo h on i s own and in combina ion, con i ms he
20
in ui i e unde s anding o egula i y as he epe i ion o o ms wi h simila dimensions,
which is consis en wi h p e ious quali a i e s udies on hy hm, p opo ion, and scale in
a chi ec u al composi ions.
The bes esul s we e ob ained o mul i-componen a ian s, especially SVEDAS, which
combines all h ee componen s: dis ance, angle, and p opo ion. Each o hem desc ibes
a di e en aspec o pe cep ion: spa ial a angemen , di ec ional ela ionships, and
simila i y o o ms, which ensu es high analy ical esolu ion and he abili y o cap u e
sub le di e ences in he o de o he composi ion.
The me hod o e alua ing he e ec i eness o he measu es is also wo h no ing. The
use o h ee indica o s (Spea man's ank co ela ion, pai o de accu acy, and sum o
ank di e ences) allowed o a mul i ace ed e alua ion ha be e e lec s he na u e o
o dinal da a and a oids o e in e p e a ion o linea co ela ion.
As a esul , SVEDAS can se e as an analy ical ool and indica o suppo ing he design
p ocess. I s high consis ency wi h expe assessmen s con i ms i s ope a ional po en ial-
bo h in he analysis o exis ing designs and in he au oma ic gene a ion o new a ian s,
which has been es ed du ing he esea ch. The algo i hm, which uses he SVEDAS alue
as a con ol pa ame e , con i med he unc ionali y o he o mula and i s use ulness in
design applica ions. Examples o composi ions gene a ed by he algo i hm a e shown in
Figu e 4.
The abili y o de ine egula i y h ough nume ical alues o o op imize he acade
design in ela ion o a minimum egula i y index (unde s ood as he SVEDAS alue)
opens a new chap e in a chi ec u al aes he ics, in line wi h he idea o da a-d i en
design. The key poin he e is ha SVEDAS does no desc ibe indi ec ac o s in luencing
aes he ics (such as sunligh ), bu aes he ics i sel . By aking his index in o accoun ,
21
designe s gain he abili y o objec i ely compa e aes he ic aspec s, which can con ibu e
o he c ea ion o a chi ec u e ha is mo e inclusi e, sus ainable, and suppo i e o SDG
11.
Conclusion
In conclusion, in he cou se o he conduc ed s udy based on he ma e ials and
me hod o he 2025 Malewczyk s udy, s a is ical analyses showed ha he SVEDAS
index , ha is, he me ic en opy o he leng hs o he connec ions be ween he cen e s
o he composi ional elemen s weigh ed by he indica o s o he a ia ion o hese
leng hs, he angles in he g id o he connec ions o he cen e s o he composi ional
elemen s and he a ia ion o he p opo ions o he composi ional elemen s, p o ed o
be ex emely e ec i e in p edic ing he isual egula i y o he composi ional designs.
A co ela ion o he au ho 's index wi h expe assessmen s o composi ional egula i y
o mo e han 96% was ob ained.
Fo nea ly 97% o he pai s co ec ly o de ed he egula i y o he composi ion. The
o iginal Hcen e dis index (Malewczyk 2025) admi edly also achie ed high esul s (93%
co ela ion and 89% co ec ly o de ed pai s), bu mos impo an ly, he new SVEDAS
index shows much g ea e sensi i i y o sub le di e ences in egula i y.
In he cou se o he wo k, an add-on o Blende was also de eloped o gene a e açade
composi ion pa e ns, based on he ypology o composi ion pa e ns by Male ich,
Ta aszkiewicz and Czyż (2022b). The add-on has he abili y o gene a e composi ions
wi h a p ede ined ange o SVEDAS ac o . Thus, he possibili y o using he de e mined
composi ional egula i y coe icien as an inpu unc ion in açade gene a i e algo i hms
was e i ied. Op imizing açade design o egula i y p omo es he goals o SDG 11
h ough a chi ec u al design. The op imiza ion o egula i y se es o imp o e
22
a chi ec u al aes he ics and u he - suppo s he c ea ion o sa e, inclusi e, sus ainable
and high-quali y ci ies and neighbo hoods.
Fig. 4 Th ee examples o acade composi ions wi h a compa able deg ee o egula i y,
gene a ed by he de eloped algo i hm..
23
Re e ences
Ali, Lana Abubak , and Fa is Ali Mus a a. 2024. F ac al dimensional analysis o
building açades: The case o o ice buildings in E bil Ci y. F ac al and
F ac ional 8(12): 746. h ps://doi.o g/10.3390/ ac al ac 8120746
Alagöz Konu , Me yem, and Ma wa Jamal. 2024. design app oaches o pa ame ic
kine ic wall sys ems: Inno a i e a chi ec u al solu ions. Konya Sana (Kona )
7: 26-43. h ps://doi.o g/10.51118/konsan.2024.41
Benish, William A. 2020. a e iew o he applica ion o in o ma ion heo y o clinical
diagnos ic es ing. En opy 22(1): 97. h ps://doi.o g/10.3390/e22010097
Besbas, Soumaya, F ancesco Noce a, Nou eddine Zemmou i, Mohamed Amine
Khad aoui, and Asma Besbas. 2022. Pa ame ic-Based Mul i-Objec i e
Op imiza ion Wo k low: Dayligh and Ene gy Pe o mance S udy o Hospi al
Building in Alge ia. Sus ainabili y 14(19): 12652.
h ps://doi.o g/10.3390/su141912652
Boeing, Geo . 2019. u ban spa ial o de : s ee ne wo k o ien a ion, con igu a ion, and
en opy. Applied Ne wo k Science 4: 67. h ps://doi.o g/10.1007/s41109-019-
0189-1
Bo ill, Ca l. 1996. ac al geome y in a chi ec u e and design. Bos on: Bi khäuse .
Cae ano, Inês, An ónio Lei ão, and Fe nando Bas os. 2020. om a chi ec u al
equi emen s o physical c ea ions: An algo i hmic-based app oach o açade
design. Jou nal o Façade Design and Enginee ing 8(2): 59-80.
h ps://doi.o g/10.7480/j de.2020.2.4543
Came on, Debo ah. 1985. eminism and linguis ic heo y. New Yo k: S . Ma in's P ess.
Came on, Debo ah. 1997. heo e ical deba es in eminis linguis ics: ques ions o sex
and gende . In Gende and discou se, ed. Ru h Wodak, 99-119. London: Sage
Publica ions.
Chen, Ping, and Hao Tang. 2024. A amewo k o adap i e açade op imiza ion design
based on building en elope pe o mance cha ac e is ics. Buildings 14(9): 2646.
h ps://doi.o g/10.3390/buildings14092646
C omp on, And ew. 2012. he en opy o LEGO®. En i onmen and Planning B:
Planning and Design 39 (1): 174-182.
24
Cudzik, Jan. 2019. changes aking place in he kine ic a chi ec u e o e he 20 h and
21s cen u ies. Space and Fo m 38: 21-32.
h ps://doi.o g/10.21005/pi .2019.38.b-02
Cudzik, Jan, and A i A asoy. 2023. he impac o con empo a y echnology on shell
s uc u es: ma e ial and ligh solu ions. Jou nal o he In e na ional Associa ion
o Shell and Spa ial S uc u es 61: 34-45.
h ps://doi.o g/10.20898/j.iass.2022.022
Ebanks, B uce. 2010. weigh ed en opies. Open Ma hema ics 8(3): 602-615.
h ps://doi.o g/10.2478/s11533-010-0026-3
Engin, Abdul Same , and Ahme Em e Dince . 2023. mo phological composi ion
assessmen o kine ic açade designs ia AHP and TOPSIS me hods based on
açade componen s and design inpu s. In e na ional Jou nal o A chi ec u al
Compu ing 22(3): 323-352. h ps://doi.o g/10.1177/14780771231209457
Ezz, Mohamed Salah, Essam Odah, Saleh Baha e ha, Abd Al-Kade A. Al Sayed, and
Doaa A. H. Salem. 2024. Analy ical examina ion o dynamic elemen s in
mode n a chi ec u al açades o ad anced s uc u al aes he ics. F on ie s in
Buil En i onmen 10. h ps://doi.o g/10.3389/ buil.2024.1302380
Fou ie, Ing id. 2012. en opy and in o ma ion heo y. 2nd ed. Online In o ma ion
Re iew 36(3): 481-482. h ps://doi.o g/10.1108/14684521211241477
Güzelci, O kan Zeynel and Sema Alacam. 2019. a S udy on Measu ing Complexi y in
Muqa nas Pa e ns. The Tu kish Online Jou nal o Design A and
Communica ion 9 (2), 191-201.
Güzelci, O kan Zeynel e al. 2020. Measu ing he En opy o Mass Housing P ojec s
Th ough Spa ial Rela ions. XXIII Gene a i e A Con e ence - GA2020, 229-
237.
Güzelci, O kan Zeynel, e al. 2021 Measu ing he En opy o Sinan's Muqa nas
Pa e ns. Nexus Ne wo k Jou nal 23: 85-105. h ps://doi.o g/10.1007/s00004-
020-00521-0
Guiaşu, Sil iu. 1971. weigh ed en opy. Repo s on Ma hema ical Physics 2(3): 165-
179. h ps://doi.o g/10.1016/0034-4877(71)90002-4
Huynh, Hoai Nguyen. 2019. spa ial poin pa e n and u ban mo phology: Pe spec i es
om en opy, complexi y, and ne wo ks. Physical Re iew E 100(2): 022320.
h ps://doi.o g/10.1103/PhysRe E.100.022320
25
Ilgaz, Eli Ece, Michael S. Bi e mann, and Dilek Yıldız Özkan. 2025. on he egula i y
o opening pa e ns in building açades. Nexus Ne wo k Jou nal 27(2): 373-389.
h ps://doi.o g/10.1007/s00004-025-00815-1
Jacobsen, Thomas. 2006. b idging he a s and sciences: A amewo k o he
psychology o aes he ics. Leona do 39(2): 155-162.
h ps://doi.o g/10.1162/leon.2006.39.2.155
Ka ona, Vilmos. 2023. The hidden dimension o açades: F ac al analysis e eals
composi ion ules in classical and Renaissance a chi ec u e. F ac al and
F ac ional 7(3): 257. h ps://doi.o g/10.3390/ ac al ac 7030257
Kelbe , Ma k, Izabella S uhl, and Iou i M. Suho . 2017 Weigh ed En opy: Basic
Inequali ies. Mode n S ochas ics: Theo y and Applica ions 4(3): 233-252.
h ps://doi.o g/10.15559/17-VMSTA85
K us, Paul. 2013. in o ma ion en opy in he design p ocess. In ICoRD'13, edi ed by
Ami Chak aba i and Rajee P akash. Lec u e No es in Mechanical
Enginee ing. Sp inge , India. h ps://doi.o g/10.1007/978-81-322-1050-4_8
Lahma , Imene, Alessand o Canna ale, F ancesco Ma ello a, and Nou eddine
Zemmou i. 2022. The impac o building o ien a ion and window- o-wall a io
on he pe o mance o elec och omic glazing in ho a id clima es: A pa ame ic
assessmen . Buildings 12(6): 724. h ps://doi.o g/10.3390/buildings12060724
Lede , Helmu , Benno Belke, And ies Oebe s , and Do o hee Augus in. 2004. a model
o aes he ic app ecia ion and aes he ic judgmen s. B i ish Jou nal o Psychology
95(4): 489-508. h ps://doi.o g/10.1348/0007126042369811
Lee, Minsuk Sang. 2014. applica ion o ac al geome y o a chi ec u al design.
A chi ec u al Ins i u e o Ko ea.
Lin, Vic o Yu-Chieh, and Shen-Guan Shih. 2025. AI-augmen ed pa ame ic açade
design: Explo ing MCTS o ea ly-s age decision-making. Nexus Ne wo k
Jou nal. h ps://doi.o g/10.1007/s00004-025-00822-2
Mahmoud, Ayman Hassaan Ahmed, and Yomna Elghazi. 2016. Pa ame ic-based
designs o kine ic açades o op imize dayligh pe o mance: Compa ing
o a ion and ansla ion kine ic mo ion o hexagonal açade pa e ns. Sola
Ene gy 126: 111-127. h ps://doi.o g/10.1016/j.solene .2015.12.039
Malewczyk, Michal. 2025. he ma hema ical egula i y coe icien o açade pa e ns.
Nexus Ne wo k Jou nal. h ps://doi.o g/10.1007/s00004-025-00827-x
