Visualiza ion o On ology E olu ion Using
On oDi G aph∗
And é La a1, Ped o Rangel Hen iques2, and Alda Lopes Gança ski3
1 Cen o ALGORITMI, Uni e sidade do Minho, B aga, Po ugal
[email p o ec ed]
2 Cen o ALGORITMI, Uni e sidade do Minho, B aga, Po ugal
[email p o ec ed]
3 Ins i u Telecom, Telecom SudPa is, CNRS Samo a , E y, F ance
[email p o ec ed]
Abs ac
On ologies e ol e wi h he passing o ime due o imp o emen s, co ec ions o changes in e-
qui emen s ha need o be made. In his pape we desc ibe a hesis wo k aiming a he c ea ion
o a isualiza ion echnique wi h he objec i e o allowing he iewe o easily iden i y changes
made in an on ology. Wi h he use o a speci ica ion based on he al eady exis ing Visual No a-
ion o OWL On ologies (VOWL) i is possible o display he di e ences ha exis be ween wo
e sions o an on ology. The p oposed app oach will be implemen ed in an applica ion, ha is
also discussed in he pape .
1998 ACM Subjec Classi ica ion H.3.1 Con en Analysis and Indexing
Keywo ds and ph ases On ology E olu ion, On ology Visualiza ion
Digi al Objec Iden i ie 10.4230/OASIcs.SLATE.2017.14
1 In oduc ion
Th ough he use o on ologies i is possible o s o e en i ies and hei ela ions wi h each
o he . Howe e wi h he inc ease in complexi y o an on ology, he isk o he use becoming
unable o keep up wi h he changes ha a e made migh make i necessa y o change he
me hod used o isualize he on ology. One o he mos in ui i e o ms o displaying an
on ology is h ough he use o a g aph, howe e he e a e a ious di e en o ms o display
he same in o ma ion in a g aph. G aphs can be p esen ed h ough he use o o ce-di ec ed,
o hogonal, adial, ees, and many o he ypes o layou s as i can be seen in [
3
]. In o de o
map on ology elemen s o g aphical en i ies he e a e al eady exis ing no a ions, one example
o his is he Visual No a ion o OWL On ologies (VOWL) no a ion [9].
The F iend o a F iend (FOAF) On ology
1
, is an on ology ha con ains in o ma ion
abou people and he connec ions hey ha e be ween each o he . Since he yea his on ology
was c ea ed (2000) un il he elease o he mos ecen e sion (2014), his on ology wen
h ough se e al di e en e sions. Tools o analyze his e olu ion in he on ology al eady
exis [
7
,
10
,
5
] howe e he isualiza ion aspec o hese ools can s ill be g ea ly imp o ed.
The exis ence o hese ools con i ms ha he e is a p oblem ha needs o be sol ed. The
c ea ion o a isualiza ion echnique ha allows use s o easily iden i y changes would help
∗
This wo k has been suppo ed by COMPETE: POCI-01-0145-FEDER-007043 and FCT – Fundação
pa a a Ciência e Tecnologia wi hin he P ojec Scope: UID/CEC/ 00319/2013.
1A ailable om h p://xmlns.com/ oa /spec/.
©And é La a, Ped o Rangel Hen iques, and Alda Lopes Gança ski;
licensed unde C ea i e Commons License CC-BY
6 h Symposium on Languages, Applica ions and Technologies (SLATE 2017).
Edi o s: R. Quei ós, M. Pin o, A. Simões, J. P. Leal, and M. J. Va anda; A icle No. 14; pp. 14:1–14:8
Open Access Se ies in In o ma ics
Schloss Dags uhl – Leibniz-Zen um ü In o ma ik, Dags uhl Publishing, Ge many
14:2 Visualiza ion o On ology E olu ion Using On oDi G aph
use s analyze how on ologies a e e ol ing and quickly de e mine wha has been changed
be ween di e en e sions o an on ology.
Wi h he use o a g aph o display an on ology i is possible o display he esul s o a
s uc u al di inside he g aph. I his is success ully execu ed i is possible o c ea e a way o
easily display he e olu ion o an on ology. The exis ence o an applica ion ha implemen s
hese ea u es will allow use s o analyse he s uc u al changes made in di e en e sions o
an on ology and isualize he impac ha hese changes made in he g aph displaying all
concep s and ela ionships o he on ology.
The p ojec he e discussed has he ollowing objec i es:
P oposal o a isualiza ion echnique ha allows he use s o easily iden i y changes in
an on ology. Ou con ibu ion ocuses on he possibili y o he use s o isualize he
changes made on an on ology wi hou ha ing o see he en i e on ology, which is di icul
o manage. They can easily selec he desi ed changes o isualize hem di ec ly in he
on ology display window and a ex ual desc ip ion.
Implemen a ion o his isualiza ion echnique in an applica ion using Ja a.
This hesis esea ch hypo hesis is ha h ough he use o he p oposed echnique, a isual
di using a g aph o display he on ology and he changes made o i , i will be easie and
as e o iden i y and loca e he di e ences in e sions o an on ology and help he use
comp ehend hei meaning and impac .
This pape is di ided in he ollowing sec ions:
In oduc ion: In his sec ion we in oduce he eade o he con ex o his pape and
wha i aims o achie e.
Rela ed Wo k: This sec ion iden i ies and discusses wo k ela ed o he p oposal he e
epo ed. A b oad and deep esea ch was made on he opics conce ned wi h his p ojec :
isualiza ion o on ology e olu ion, and change de ec ion. Howe e o he sake o space
i is no possible o include he e all he ma e ial disco e ed; o ead all he in o ma ion
collec ed and o ganized, he eade should see [8].
On oDi G aph: The P oposal: In his sec ion we discuss he p oposed echnique and he
a chi ec u e o he applica ion whe e i is going o be implemen ed, On oDi G aph.
On oDi G aph: De elopmen : This sec ion con ains in o ma ion ela ed o ea u es and
decisions ha we e aken du ing he de elopmen o he applica ion.
Conclusion: In his sec ion, we analyse wha has been desc ibed in his pape and he
u u e wo k ha s ill needs o be done.
2 Rela ed Wo k
2.1 On ology Visualiza ion
The e a e languages ha can be used o se ialize he con en and s uc u e o an on ology.
One o hese languages is he Web On ology Language (OWL [
1
]), buil on he al eady exis ing
Resou ce Desc ip ion F amewo k (RDF [2]) and RDF Schema (RDFS [4]) speci ica ions.
Visual No a ion o OWL On ologies (VOWL) [
9
] is a isual language o on ologies
wi h he aim o help use s unde s and he s uc u e o an on ology in ui i ely. The VOWL
no a ion p o ides a g aphical e sion o he a ious exis ing on ology elemen s and has been
imp o ed since i s ini ial elease. In Figu e 1 i is possible o obse e he ep esen a ion o
he a ious elemen s o an on ology. This no a ion can be ead in mo e de ail in [9].
A. La a, P. R. Hen iques, and A. L. Gança ski 14:3
Figu e 1 Visualiza ion o an on ology wi h he VOWL no a ion.
2.2 Change Visualiza ion
PROMPT-Viz [
11
] is a P o égé plugin ha ex ends he PROMPTDi plugin so ha he
di e ences be ween on ologies can be displayed in ui i ely. This plugin uses a zoomable
eemap layou o op imize he u iliza ion o he sc een, acili a ing he isualiza ion o la ge
on ologies.
A cs a e used o display changes in he loca ion o classes in he eemap and i s
connec ions o he o he nodes. The selec ion o a node e eals i s a cs o o he nodes,
hese a cs a e colou ed depending on how he des ina ion class has been changed. This
ea u e clu e ed he isualiza ion i i was ac i e o all he nodes all he ime, he e o e his
in o ma ion is only displayed on a selec ion e en . Howe e he use s ha e alua ed his
plugin s ill had di icul y unde s anding he in o ma ion he a cs we e ying o con ey.
Abe OWL [
12
] is an on ology eposi o y and amewo k o on ology-based da a access
h ough he use o a web in e ace. In Abe OWL i is possible o isualize on ologies as
di ec ed g aphs whe e nodes ep esen classes and edges ep esen axioms. I is also possible
o isualize he di e ence be ween se e al e sions o an on ology, o achie e his Abe OWL
uses a di e en colou o each e sion o he on ology o di e en ia e hem. Howe e he
isualiza ion o he e olu ion o an on ology is no he main ea u e o his sys em. I is no
possible o sea ch o he changes be ween wo di e en e sions, he use mus expand he
ee manually o isualize he en i e on ology and be able o de ec he e olu ion, his is
no a p ac ical solu ion because i he use needs o isualize a la ge on ology wi h a la ge
numbe o nodes i will ake a e y long ime o ge he desi ed esul s.
3 On oDi G aph: The P oposal
We p opose a sys em, called On oDi G aph, explo ing a new isualiza ion echnique o easily
iden i y he di e ences be ween wo e sions o an on ology. This echnique uses a isual
no a ion ha is based on he al eady exis ing VOWL speci ica ion [
9
], used o map on ology
elemen s o isual g aph elemen s in o de o easily display an on ology, wi h he objec i e
o displaying he di e ences be ween wo e sions o an on ology.
To make he use able o easily iden i y wha has been changed and he ype o change
made o an elemen we decided o add bo de s o he isual ep esen a ion o he a ious
on ology elemen s. The ypes o changes ha will be iden i ied a e c ea ions, modi ica ions
and dele ions. Each one o hese ypes will ha e a colou associa ed wi h hem: c ea ion will
SLATE 2017
14:4 Visualiza ion o On ology E olu ion Using On oDi G aph
Figu e 2 Class C ea ion, Modi ica ion and Dele ion.
Figu e 3 Rela ion C ea ion, Modi ica ion and Dele ion.
ha e a g een bo de , modi ica ion will ha e a blue bo de and dele ion will ha e a ed bo de .
Figu e 2 shows wha a owl:Class node would look like when displayed wi h his no a ion.
Figu e 3 shows wha he edges would look like when he no a ion is applied on a
owl:Objec P ope y.
Changes o he domain o ange o he objec p ope y will make his elemen ha e a blue
bo de , because hese changes a e conside ed a modi ica ion o he p ope y. The change in
domain and ange will also be seen in colou s o he edges connec ing he p ope y o i s
a ge s. Figu e 4 displays a change in he objec p ope y domain and ange, he domain
and he ange used o be Resou ce bu in he new e sion i is Thing.
In Figu e 5 i is possible o obse e wha he edges would look like when he no a ion is
applied on a owl:Da a ypeP ope y.
Howe e , due o he la ge size o some on ologies, i can s ill be di icul o loca e he
changes exis ing om one e sion o he o he one, e en wi h he added isual in o ma ion.
To sol e his, he use will also ha e access o he ex ual e sion o he changes made in he
on ology. When he use selec s one o he changes in he ex ual e sion he iewable a ea
in he applica ion will eposi ion i sel so ha he change in cen e ed on ha a ea.
On oDi G aph: A chi ec u e
The p e iously s a ed ea u es a e being implemen ed in a Ja a p og am, using Ja aFX o
d aw he g aph and OWL API [6] o ex ac in o ma ion om he on ology.
Figu e 6 con ains he a chi ec u e o he p oposed sys em. This sys em will be able o
load on ologies and display he di e ences be ween hem h ough he use o a g aph.
To load on ologies, we use o he OWL API [
6
], a Ja a API used o c ea e, pa se,
manipula e and se ialize OWL on ologies. This p ocess can be obse ed in Figu e 6 p ocess
named Load On ologies. A e he loading ask he sys em compa es he on ologies ha we e
loaded and iden i ies he di e ences ha exis be ween hem, in he Calcula e Di e ences
p ocess. Wi h his in o ma ion, he sys em will be able o gene a e a g aph in he Gene a e
G aph p ocess. This g aph con ains he on ology in o ma ion and he di e ences be ween
he on ologies. A e c ea ing he g aph s uc u e, he only s ep le o do is d awing he
A. La a, P. R. Hen iques, and A. L. Gança ski 14:5
Figu e 4 Example o a change in domain and ange.
Figu e 5 Da a ype C ea ion, Modi ica ion and Dele ion.
g aph so ha he esul s can be displayed o he use . This will be done wi h he use o a
Ja aFX Can as in he D aw G aph p ocess.
4 On oDi G aph: De elopmen
This sec ion con ains de ailed in o ma ion abou he ea u es and decisions made in he
de elopmen o he applica ion. I does no only con ain he implemen a ion o he echnique
bu all he necessa y ea u es o a comple e on ology isualiza ion applica ion, such as
manipula ion o g aph elemen s, layou cus omiza ion o he isualiza ion o he on ology.
4.1 Basic ea u es
Some impo an ea u es need o be added o imp o e he usabili y o he applica ion. The
ollowing ea u es we e conside ed mus -ha e and ha e been implemen ed in he applica ion:
Zooming, allows he use s o zoom in and ou so ha he g aph can be seen om di e en
dis ances.
Panning, allows he use s o mo e he isible g aph a ea o he applica ion.
D agging, allows he use s o d ag nodes o he g aph o hei desi ed posi ion.
I will also be possible o easily ind he loca ion o an elemen in he g aph by he use o
he menu con aining he lis o classes, objec p ope ies and da a p ope ies o he on ology.
By selec ing an elemen o he lis , he node in he isualiza ion will be cen e ed in he
iewable a ea. This menu is shown in he ollowing igu e:
In his menu i is also possible o see he elemen s o he on ology ha we e added,
modi ied and emo ed.
The backg ound colou o each cell shows wha happened o ha elemen be ween he
di e en e sions, he possible colou s a e:
Whi e, he elemen did no change om one e sion o he o he .
G een, he elemen was added o he on ology.
Blue, he elemen was modi ied.
Red, he elemen was emo ed om he on ology.
SLATE 2017

14:6 Visualiza ion o On ology E olu ion Using On oDi G aph
Figu e 6 A chi ec u e o he On oDi G aph sys em.
Figu e 7 Menu wi h all on ology elemen s.
4.2 Layou Cus omiza ion
In he applica ion i is possible o change gene al and layou speci ic alues du ing un ime.
This makes he use able o easily lea n he impac o each pa ame e in he g aph isualiza ion
and is a be e op ion han eloading he en i e g aph when changes a e made o hese.
The cu en ly a ailable a iables ha can be changed in he o ce-di ec ed layou a e he
ollowing:
Edge spacing, con ols he dis ance be ween edges wi h he same node ends.
Repulsion o ce, con ols he epulsion o ce o he nodes o he g aph.
A ac ion o ce, con ols he a ac ion o ce o he edges o he g aph.
Sp ing leng h, con ols he leng h o he edges.
Damping, con ols he damping o he o ces o he g aph.
Times ep, con ols he seconds ha pass in e e y i e a ion o he algo i hm.
4.3 Change de ec ion
In o de o ind all he di e ences be ween wo e sions o an on ology i is necessa y o
compa e i s con en s. All on ology elemen s con ain an unique IRI ha iden i ies hem,
his will help iden i y elemen s ha we e added o emo ed. The ollowing lis con ains he
change made in an on ology and how i is possible o de e mine ha change:
Elemen was added, an elemen is classi ied as added o he on ology i i s IRI does no
exis in he ini ial e sion bu exis s in he inal e sion.
Elemen was emo ed, an elemen is classi ied as emo ed om he on ology i i s IRI
exis s in he ini ial e sion bu does no exis in he inal e sion.
Elemen was modi ied, an elemen is classi ied as modi ied i in o ma ion con ained in
his elemen is modi ied. An example o his is a change in he elemen domain o ange
as seen in Figu e 4.
Howe e his me hod has some weaknesses, one o hose is ha i does no classi y a
change in an elemen name as a modi ica ion. The enaming o an elemen will change i s
IRI, he e o e his change will be displayed as he emo al o he elemen wi h he old name
and he addi ion o a new elemen wi h he new name.
A. La a, P. R. Hen iques, and A. L. Gança ski 14:7
Figu e 8 Visualiza ion o classes and objec p ope ies in he applica ion.
Figu e 9 Visualiza ion o classes and da a p ope ies in he applica ion.
4.4 Visualiza ion
In o de o c ea e he isualiza ion i is necessa y o ob ain all he elemen s o he on ology and
ans o m hese in o nodes and edges wi h hei espec i e isual ep esen a ion. The ollowing
igu es displays classes, objec and da a p ope ies as seen in he de eloped applica ion.
In he g aph i is also possible o obse e he union node ha is c ea ed when he p ope y
domain o ange con ains mo e han one elemen .
5 Conclusion
In his pape he con ex , mo i a ion and objec i es o he unde lining wo k we e p esen ed
and discussed. The ou comes conce ning he esea ch done on he opics in ol ed in he s a e
o he a o ou wo king a ea we e epo ed. Conce ning he opic o on ology isualiza ion
we jus i ied he adop ion o some o he isual elemen s om VOWL wi h he ac ha
i is a good no a ion o display on ologies and ha i is al eady being used by se e al
applica ions. Tools ha allow he isualiza ion o he di e ence be ween on ologies and hei
main s eng hs and weaknesses we e also discussed – such su ey suppo ed ou decisions in
he c ea ion o On oDi G aph. To sol e his p oblem we p opose a solu ion ha consis s o
a isualiza ion echnique ha will be implemen ed in an Ja a applica ion. The a chi ec u e,
ea u es and de elopmen decisions o his applica ion we e de ailed in his pape .
The applica ion ha was p oposed is s ill unde de elopmen , howe e he isualiza ion o
an on ology is almos comple e. The e is s ill wo k ha needs o be done in he de ec ion o
di e ences in wo e sions o an on ology and he ep esen a ion o his in o ma ion howe e
i is al eady possible o de e mine he on ology elemen s ha we e added and emo ed om
he di e en e sions. A e he implemen a ion o hese ea u es in he applica ion i will be
e alua ed o e i y i i needs o be modi ied o imp o ed.
SLATE 2017
14:8 Visualiza ion o On ology E olu ion Using On oDi G aph
Re e ences
1Sean Bechho e , F ank an Ha melen, Jim Hendle , Ian Ho ocks, Debo ah L. McGuin-
ness, Pe e F. Pa el-Schneide , and Lynn And ea S ein. OWL Web On ology Lan-
guage Re e ence. Wo ld Wide Web Conso ium, 2004. h p://www.w3.o g/TR/2004/
REC-owl- e -20040210/.
2Da e Becke . RDF/XML Syn ax Speci ica ion (Re ised). Wo ld Wide Web Conso ium,
2004. URL: h p://www.w3.o g/TR/2004/REC- d -syn ax-g amma -20040210/.
3Giuseppe Di Ba is a, Pe e Eades, Robe o Tamassia, and Ioannis G. Tollis. Algo i hms
o d awing g aphs: an anno a ed bibliog aphy. Compu a ional Geome y: heo y and
applica ions, 4(5):235–282, 1994.
4Ramana han Guha and Dan B ickley. RDF Schema 1.1 Recommenda ion. Wo ld Wide Web
Conso ium, 2014. URL: h p://www.w3.o g/TR/2014/REC- d -schema-20140225/.
5Michael Ha ung, Anika G oss, and E ha d Rahm. CODEX: explo a ion o seman ic
changes be ween on ology e sions. Bioin o ma ics, 28(6):895–896, 2012.
6Ma hew Ho idge and Sean Bechho e . The OWL API: A Ja a API o OWL On ologies.
Seman ic Web, 2(1):11–21, Janua y 2011.
7Pe K emen, Ma ek Smid, and Zdenek Kouba. OWLDi : a p ac ical ool o compa ison
and me ge o OWL on ologies. In 22nd In e na ional Wo kshop on Da abase and Expe
Sys ems Applica ions, pages 229–233, 2011.
8And é La a. Visualiza ion o On ology E olu ion using On oDi G aph. Mas e ’s hesis,
Uni e si y o Minho, 2017.
9S e en Lohmann, S e an Neg u, Flo ian Haag, and Thomas E l. VOWL 2: use -o ien ed
isualiza ion o on ologies. In K zysz o Janowicz, S e an Schlobach, Pa ick Lamb ix, and
Ee o Hy önen, edi o s, In e na ional Con e ence on Knowledge Enginee ing and Knowledge
Managemen , pages 266–281, 2014.
10 Na alya F idman Noy, Ma k A. Musen, e al. P omp Di : A ixed-poin algo i hm o
compa ing on ology e sions. In Associa ion o he Ad ancemen o A i icial In elligence
(AAAI) Con e ence, pages 744–750, 2002.
11 Da id S ephen John Pe in. P omp - iz: On ology e sion compa ison isualiza ions wi h
eemaps. PhD hesis, Uni e si y o Vic o ia, Canada, 2004.
12 Miguel Ángel Rod ıguez-Ga cıa, Luke Sla e , Kei on O’Shea, Paul N. Scho ield, Geo gios V.
Gkou os, and Robe Hoehndo . Visualizing on ologies wi h Abe OWL. In James Malone,
Robe S e ens, Ke s in Fo sbe g, and And ea Splendiani, edi o s, Seman ic Web Applica-
ions and Tools o he Li e Sciences, pages 183–192, Decembe 2015.