Klinkmülle , Ch is ophe ; Banda a, H. M. N. Dilum; an de Aals , Wil; Hobeck,
Richa d; Webe , Ingo
A icle — Published Ve sion
On he Sui abili y o P ocess Mining o Enhancing
T anspa ency o Blockchain Applica ions
Business & In o ma ion Sys ems Enginee ing
Sugges ed Ci a ion: Klinkmülle , Ch is ophe ; Banda a, H. M. N. Dilum; an de Aals , Wil; Hobeck,
Richa d; Webe , Ingo (2024) : On he Sui abili y o P ocess Mining o Enhancing T anspa ency o
Blockchain Applica ions, Business & In o ma ion Sys ems Enginee ing, ISSN 1867-0202, Sp inge
Fachmedien Wiesbaden, Wiesbaden, Vol. 67, Iss. 6, pp. 777-796,
h ps://doi.o g/10.1007/s12599-024-00903-5
This Ve sion is a ailable a :
h ps://hdl.handle.ne /10419/333368
S anda d-Nu zungsbedingungen:
Die Dokumen e au EconS o dü en zu eigenen wissenscha lichen
Zwecken und zum P i a geb auch gespeiche und kopie we den.
Sie dü en die Dokumen e nich ü ö en liche ode komme zielle
Zwecke e iel äl igen, ö en lich auss ellen, ö en lich zugänglich
machen, e eiben ode ande wei ig nu zen.
So e n die Ve asse die Dokumen e un e Open-Con en -Lizenzen
(insbesonde e CC-Lizenzen) zu Ve ügung ges ell haben soll en,
gel en abweichend on diesen Nu zungsbedingungen die in de do
genann en Lizenz gewäh en Nu zungs ech e.
Te ms o use:
Documen s in EconS o may be sa ed and copied o you pe sonal
and schola ly pu poses.
You a e no o copy documen s o public o comme cial pu poses, o
exhibi he documen s publicly, o make hem publicly a ailable on he
in e ne , o o dis ibu e o o he wise use he documen s in public.
I he documen s ha e been made a ailable unde an Open Con en
Licence (especially C ea i e Commons Licences), you may exe cise
u he usage igh s as speci ied in he indica ed licence.
h p://c ea i ecommons.o g/licenses/by/4.0/
RESEARCH PAPER
On he Sui abili y o P ocess Mining o Enhancing T anspa ency
o Blockchain Applica ions
Richa d Hobeck •Ch is ophe Klinkmu
¨lle •H. M. N. Dilum Banda a •
Ingo Webe •Wil an de Aals
Recei ed: 18 Decembe 2022 / Accep ed: 28 June 2024 / Published online: 25 Oc obe 2024
ÓThe Au ho (s) 2024
Abs ac Blockchain echnology is known o i s ans-
pa ency p ope ies due o i s publicly a ailable,
immu able da a. Ye , as da a a ailabili y does no inhe -
en ly ensu e anspa ency, u he analy ical me hods may
be equi ed o human in e p e a ion o da a aces. P ocess
mining has eme ged as a popula oolbox o unde s anding
p ocesses and how hey a e execu ed in p ac ice. The pape
s udies p ocess mining as a me hod o enhance he ans-
pa ency o blockchain da a. To his end, wo popula
E he eum applica ions we e analyzed using p ocess min-
ing: he p edic ion and be ing ma ke place Augu and he
ne wo k ma ke ing pla o m Fo sage. Obse a ions om
he p ocess-mining analyses a e used o discuss i p ocess
mining can se e as a me hod o es ablish anspa ency o a
blockchain. Fo bo h applica ions, new insigh s a e gene -
a ed o usage scena ios such as applica ion edesign,
secu i y analysis, use beha io analysis, and e ealing
blind spo s in Augu ’s and Fo sage’s documen a ion. The
pape concludes ha he e is e idence ha p ocess mining
can se e as a me hod o enhance anspa ency in block-
chains a he cos o echnical se up and knowledge
acquisi ion.
Keywo ds Blockchain P ocess mining T anspa ency
P ocess disco e y Con o mance checking P ocess
enhancemen E he eum
1 In oduc ion
A blockchain can be cha ac e ized as a dis ibu ed, append-
only da a s o e o ansac ions (Xu e al. 2019). Second-
gene a ion blockchains allow o deploying and execu ing
use -de ined p og ams called sma con ac s. On his
basis, blockchain has eme ged as a echnology ha enables
he au oma ion o c oss-o ganiza ional p ocesses on a
neu al pla o m (Mendling e al. 2018; Webe e al. 2016),
and mo e gene ally he design, de elopmen , and ope a ion
o decen alized applica ions (DApps) (Xu e al. 2019). In
heo y, public blockchains o e anspa ency p ope ies
(Xu e al. 2019, p.19 ). Tha is, public blockchains con-
s i u e a dis ibu ed compu ing en i onmen in which use s
can hos , execu e, and s o e applica ions and da a. A co e
a gumen o anspa ency p ope ies o public blockchains
is he accessibili y o his en i onmen , including access o
he deployed p og ams, s o ed da a, and, i made a ailable,
he sou ce code o applica ions. Following Leona di and
T eem (2020), we a gue ha in o ma ion a ailabili y alone
is insu icien o es ablish anspa ency. Addi ional ana-
ly ical s eps a e equi ed o p esen he da a in easily
accessible o ma s and he eby achie e anspa ency.
Hence, examining he beha io o blockchain applica ions
Accep ed a e 3 e isions by Daniel Be e ungen.
R. Hobeck (&)
Chai o Se ice-cen ic Ne wo king, Technische Uni e si ae
Be lin, Be lin, Ge many
e-mail: [email p o ec ed]
C. Klinkmu
¨lle
BPMo ion, Sydney, Aus alia
H. M. N. D. Banda a
Da a61, CSIRO, Sydney, Aus alia
I. Webe (&)
School o CIT & F aunho e -Gesellscha , Technical Uni e si y
o Munich, Munich, Ge many
e-mail: [email p o ec ed]
W. an de Aals
RWTH Aachen Uni e si y, Aachen, Ge many
123
Bus In Sys Eng 67(6):777–796 (2025)
h ps://doi.o g/10.1007/s12599-024-00903-5
and use s equi es e o o u n he in o ma ion om he
blockchain in o clea insigh s.
P ocess mining ( an de Aals 2016) p o ides a se o
ools o ex ac knowledge om da a, e.g., h ough he
disco e y o p ocess models om da a wi hou p io
in o ma ion abou he p ocess (IEEE Task Fo ce on P o-
cess Mining 2011). P ocess mining has become popula as
a oolbox o unde s anding p ocesses and how hey a e
execu ed in p ac ice. Fo example, many case s udies
anging om heal hca e (And ews e al. 2018; Mans e al.
2009; Ro ani e al. 2015; Su iadi e al. 2014), inance (De
Wee d e al. 2013; Jans e al. 2011), manu ac u -
ing (Rozina e al. 2009), and public se ices ( an de
Aals e al. 2007; Leemans e al. 2019) o so wa e de el-
opmen (Lemos e al. 2011) ha e applied p ocess mining
o analyze p ocesses om di e en pe spec i es such as
con ol low, con o mance, d i s, and pe o -
mance (Reinkemeye 2020). Ne e heless, p ocess mining
on blockchain da a has u ned ou o be a challenging
ask (Di Ciccio e al. 2018). Hence, ecen ly esea che s
ha e c ea ed echniques o ex ac au ho i a i e da a om
blockchains (Klinkmu
¨lle e al. 2019,2020).
In his a icle, we complemen hose p oposi ions by
s udying he u ili y o p ocess mining on blockchain da a in
he con ex o eal-wo ld use cases. To his end, we iew
p ocess mining om a me hodological pe spec i e ( an
Eck e al. 2015; Klinkmu
¨lle e al. 2019). While p ocess
mining p o ides aluable ools o da a-d i en analysis o
all ypes o p ocesses, we do no only wan o unde s and
he u ili y o hese ools bu also he easibili y o applying
hem. He e, we ocus on DApp anspa ency and speci i-
cally on wo p ominen esea ch goals om he blockchain
domain (see Sec . 2.3):
G1: Code Valida ion & Ve i ica ion: De e mine o wha
ex en p ocess mining can con ibu e o making
DApps mo e anspa en by suppo ing he alida ion
and e i ica ion o hei sou ce code.
G2: Use Beha io Analysis: De e mine o wha ex en
p ocess mining can con ibu e o making DApps
mo e anspa en by suppo ing he analysis o hei
use s’ beha io .
To his end, we conduc wo in-dep h DApp analyses wi h
p ocess mining on popula E he eum applica ions: Augu
1
and Fo sage.
2
Augu is a p edic ion and be ing ma ke -
place, whe e use s (1) aise ques ions abou u u e e en s,
(2) p edic and be on answe s, and (3) se le he be s a e
he e en occu ed and he answe is known. Fo sage is a
ne wo k ma ke ing (aka mul i-le el ma ke ing) pla o m
ha was e y popula un il being decla ed a Ponzi
scheme by ma ke egula o s in se e al coun ies
3,
.
4
No e
ha we p esen ed he Augu analysis in a p e ious
pape (Hobeck e al. 2021) in which we ou lined ou gen-
e al expe ience and indings om using p ocess mining o
explo e his applica ion in an open-ended analysis. By
con as , in his pape , we examine he u ili y o p ocess
mining o speci ic applica ion scena ios and syn hesize
ou ini ial insigh s wi h hose o he second DApp analysis.
In his way, we p o ide e idence ha p ocess mining can
enhance he anspa ency o blockchains and, as a esul ,
gene a e aluable insigh s o code alida ion & e i ica-
ion and use beha io analysis.
The emainde o he pape is s uc u ed as ollows.
Fi s , we mo i a e ou esea ch goals by (1) summa izing
blockchain esea ch challenges, (2) ou lining why p ocess
mining can help add ess some o hem in p inciple, and (3)
e iewing ela ed wo k. In Sec . 3, we p esen ou
me hodology, including jus i ica ions o choosing Augu
and Fo sage. Sec ion 4ou lines he da a ex ac ion and p e-
p ocessing p ocedu es unde lying bo h DApp analyses. The
pape ’s ocal poin a e he DApp analyses o Augu
(Sec . 5) and Fo sage (Sec . 6). In each sec ion, we
desc ibe he espec i e applica ion and da a analyses,
co e ing insigh s om da a explo a ion, p ocess disco e y,
con o mance checking, and pe o mance analysis. In
Sec . 7, we hen discuss and syn hesize he esul s o assess
he con ibu ion o p ocess mining in enhancing he
anspa ency in a blockchain en i onmen in alignmen
wi h ou esea ch goals. Finally, we conclude in Sec . 8.
2 Mo i a ion
This sec ion mo i a es ou wo k. Fi s , we in oduce he
pape ’s unde s anding o anspa ency in Sec . 2.1 and
summa ize he p ocess-mining discipline in Sec . 2.2.We
hen in oduce basic blockchain concep s and discuss open
esea ch challenges in Sec . 2.3, highligh ing code alida-
ion & e i ica ion and use beha io analysis as challenges
ha could bene i om p ocess mining. Las ly, we e iew
wo k ela ed o p ocess mining in he con ex o blockchain
in Sec . 2.4.
2.1 So wa e T anspa ency
Acco ding o Lei e and Cappelli (2010), anspa ency in
so wa e conce ns he disclosu e o in o ma ion, i.e., exe-
cu ion da a and so wa e unc ions ha ans o m inpu o
1
h ps://augu .ne /, accessed 12 June 2022.
2
h ps:// o sage.io/, accessed 12 June 2022.
3
h ps://www.sec.go .ph/cdo-2020/ o sage-and- o sage-philippines/,
accessed 28 June 2022.
4
h ps://csim .go /2021/04/07/us-s a e-issues-cease-and-desis -
o de -agains -dapp- o sage/, accessed 28 Aug 2023.
123
778 R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025)
ou pu da a. They no e ha al hough disclosed, so wa e
unc ions can emain ob usca ed i he so wa e code is
p esen ed in a way ha is ha d o ead. Leona di and T eem
(2020) ha e a simila s andpoin wi h espec o ans-
pa ency and da a a ailabili y. They a gue ha e o s o
achie e g ea e anspa ency by disclosing mo e da a can
ob usca e he isibili y o in o ma ion in ha da a. They
call ha e ec he anspa ency pa adox and a gue ha i
akes addi ional analy ical s eps o e ie e in o ma ion
om he da a. In his pape , we use he anspa ency no ion
desc ibed by Lei e and Cappelli (2010) and ollow he
a gumen o Leona di and T eem (2020) ha in o ma ion
a ailabili y equi es app op ia e analy ical s eps on he
disclosed da a.
2.2 P ocess Mining
P ocess mining ( an de Aals 2016) is inc easingly used
o moni o and imp o e ope a ional p ocesses. I o e s a
ich ool se o analyzing e en da a, i.e., da a ha con ain
in o ma ion abou e en s occu ing du ing he execu ion o
p ocesses ( an de Aals 2016). Each e en mus a leas
ha e h ee a ibu es: (1) a case iden i ie o he case ha
he e en belongs o, (2) an ac i i y name ha ep esen s
he ac i i y whose execu ion led o he e en , and (3) a
imes amp a which he e en occu ed. Addi ional a i-
bu es may e e o loca ions, esou ces, cos s, ansac ion
da a, and on he E he eum blockchain, he consumed gas
(i.e., a measu e o he compu a ional e o o an ope a-
ion). A ace is he sequence o all e en s belonging o he
same case and so ed by imes amps. Focusing on he
ac i i y names only, aces can be ans o med in o a i-
an s, i.e., unique sequences o ac i i ies. Las ly, an e en
log is a collec ion o e en s s o ed in a o ma like XES
(Ex ensible E en S eam) (Acampo a e al. 2017).
Figu e 1illus a es an o e iew o p ocess-mining ca -
ego ies ( an de Aals 2016). No e ha he ca ego ies a e
shown as ec angles. Fi s , p ocess disco e y algo i hms
in e p ocess models, hus isualizing he con ol low o
he p ocess ha gene a ed he da a. Second, con o mance
checking analyzes he deg ee o which he beha io o
indi idual aces and/o en i e e en logs adhe es o a
no ma i e p ocess model ha desc ibes he expec ed con-
ol low. Thi d, pe o mance analysis can be used o
ob ain insigh s in o he key pe o mance indica o s like
cycle ime o s a p oduc i i y, and o de ec d i e s o
hose indica o s. Fou h, compa a i e p ocess mining and
d i de ec ion enable he in es iga ion o p ocess beha io
unde di e en condi ions, allowing analys s o, e.g.,
inspec beha io al di e ences be ween use g oups o
changes o e ime. Fi h, explo a ion enables analys s o
unde s and da a cha ac e is ics using isual analy ics and o
de ec ou lie s.
Addi ionally, p ocess-mining esea ch add esses
me hodological aspec s, see e.g., an Eck e al. (2015) and
Klinkmu
¨lle e al. (2019). Such esea ch aims o unde -
s and and p o ide guidelines o conduc ing p ocess-min-
ing p ojec s, including aspec s like planning, da a
ex ac ion, da a p epa a ion, and insigh alida ion. Sec . 3
p esen s he p ocess-mining me hodology applied in he
con ex o his wo k.
In essence, p ocess mining p o ides means o analyze
di e en p ocess pe spec i es, including con ol low,
con o mance, d i s, and pe o mance (Reinkemeye
2020). The e o e, we a gue ha hose analyses can e eal
meaning ul insigh s om blockchain ansac ion execu ion
da a. Tha is, we expec insigh s in o he blockchain
applica ion beha io o assis so wa e de elope s wi h
code alida ion & e i ica ion in gene al, and speci ically
wi h ensu ing ha applica ions co ec ly implemen p o-
cess equi emen s. Conside ing ha blockchain e en s a e
he esul s o ansac ions ini ia ed by use s, we also expec
ha insigh s in o he applica ion beha io e eal insigh s
in o use beha io s. No e ha blockchains eco d he use
add ess ha igge ed a speci ic ansac ion, enabling us o
en ich ex ac ed e en logs wi h his in o ma ion. Nex , we
ou line ou a ionale o in es iga ing p ocess mining o
blockchain analysis, and in pa icula o code alida ion &
e i ica ion and use beha io analysis.
2.3 Blockchain Resea ch Challenges
Ablockchain is an append-only s o e o ansac ions, dis-
ibu ed ac oss a pee - o-pee ne wo k and s uc u ed as a
linked lis o blocks (Xu e al. 2019). Second-gene a ion
blockchains also p o ide a neu al execu ion in as uc u e
o unning use -de ined p og ams, called sma con ac s.
Applica ions ha ope a e au onomously h ough sma
con ac s and un on op o a blockchain a e called de-
cen alized applica ions (DApps). Blockchain p o ides
immu abili y, anspa ency, and da a in eg i y o DApps.
Simila o adi ional en e p ise in o ma ion sys ems, many
DApps suppo ope a ional p ocesses. Fo example, he e
a e decen alized au onomous o ganiza ions (DAOs) o
which blockchain applica ions de ine a se o anspa en
p ocesses and ules ha allow he o ganiza ions’ membe s
o con ol he o ganiza ions wi hou equi ing cen alized
leade ship (P us y 2017). Simila ly, c oss-o ganiza ional
p ocesses, such as supply chains, can be acili a ed by
blockchain echnology ha en o ces business ules and
exchanges business in o ma ion (Mendling e al. 2018;
Webe e al. 2016).
The e a e a ious e iews and esea ch agendas ha
map ou challenges and di ec ions o blockchain esea ch.
Table 1ca ego izes challenges iden i ied in hose publi-
ca ions. O e all, he e a e six ca ego ies. Fi s , applica ion
123
R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025) 779
design conce ns he de elopmen o DApps, gi en he
limi a ions o a blockchain en i onmen . While his ca e-
go y subsumes aspec s like p og amming language ea-
u es, applica ion op imiza ion, and immu abili y o sma
con ac s, a key a ea o conce n is code alida ion & e -
i ica ion. Second, da a p i acy co e s e hical and legal da a
p i acy conce ns, as well as da a con iden iali y challenges.
Thi d, he socio- echnical ca ego y ocuses on in e ac ions
be ween use s and blockchain echnology. In his ega d, a
cen al challenge is o unde s and use beha io and how i
is in luenced by echnological ea u es. Fou h, scalabili y
e ol es a ound issues conce ning la ency, and limi ed
ansac ion and s o age olumes o blockchains. Fi h,
blockchain in e ope abili y includes in eg a ion wi h
exis ing sys ems and o -chain componen s. Las ly, he
de elopmen o consensus mechanisms ocuses on p o o-
cols o blockchain nodes o ag ee on he blockchain s a e
in a c yp og aphically secu e and en i onmen ally sus-
ainable way.
In his wo k, we s udy p ocess mining as a means o
enhance blockchain anspa ency. Conside ing he analy -
ical na u e o p ocess mining, we belie e ha p ocess
mining can help use s in il e ing, analyzing, and isual-
izing blockchain da a so ha i is easie o unde -
s and (Leona di and T eem 2020, p. 1611). Hence, ou
in e es speci ically lies in explo ing p ocess mining in he
con ex o challenges ha can bene i om he analysis o
blockchain da a.
Fi s , his includes sma con ac code alida ion &
e i ica ion challenges om he applica ion design ca e-
go y. He e, es ing and e alua ing DApps a e deploymen
is mos ele an o ou wo k, as i is o in e es o use s who
wan o moni o i he DApp and any upda es comply wi h
expec ed beha io (Casino e al. 2019; Rossi e al. 2019;
Zheng e al. 2020), e.g., o de ec audulen sche-
mes (Risius and Spoh e 2017; Casino e al. 2019; Zheng
e al. 2020). No e ha while he deployed code o indi-
idual sma con ac s is immu able, DApp beha io can be
al e ed a un- ime by dynamically changing pa ame e s
and he binding o sma con ac s. In his ega d, e o s by
ex e nal use s a e equen ly exace ba ed by he una ail-
abili y o eadabili y o sma con ac code (Zheng e al.
2020; Sha ma e al. 2023). We a gue ha analyzing e en
da a gene a ed du ing sma con ac execu ion using p o-
cess mining can enhance anspa ency and assis use s in
sma con ac code alida ion & e i ica ion (G1). In
pa icula , we belie e ha p ocess-disco e y and con o -
mance-checking capabili ies suppo he de ec ion o
de ia ions om he expec ed beha io . In p ac ical e ms,
use s can access he public blockchain da a o a DApp bu
Fig. 1 P ocess-mining
ca ego ies o e en da a
analysis
Table 1 Ca ego ies o challenges om ela ed wo k (– gene ally men ion ca ego y; CV – men ion challenges ela ed o code alida ion &
e i ica ion; UB – men ion use beha io analysis)
Applica ion design Da a p i acy Socio- echnical Scalabili y In e ope abili y Consensus mechanisms
Risius and Spoh e (2017)CV UB
Zheng e al. (2018)
Casino e al. (2019)CV
Rossi e al. (2019)CV UB
Zheng e al. (2020)CV
Vacca e al. (2021)CV
Sha ma e al. (2023)CV
123
780 R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025)
equi e me hods o check i he DApp implemen s se ices
as ad e ised o no , and he e p ocess mining could be he
me hod o choice.
Second, he e is consensus ha om a socio- echnical
pe spec i e, blockchain esea ch mus become mul i-dis-
ciplina y and conside he pe spec i es o all in ol ed
s akeholde s. While he e is a gene al call o heo ies ha
explain how echnological blockchain ea u es in luence
use beha io (Rossi e al. 2019; Risius and Spoh e 2017),
we ocus on a pa icula challenge o indi idual s ake-
holde s. Tha is, many DApps, such as hose implemen ed
o DAOs, can be iewed as p ocess coo dina ion o
o ches a ion mechanisms ha allow indi iduals o ans-
ac . This c ea es he need o DApp use s o decide whe he
hey wan o us o he use s. Howe e , insigh s in o he
mo i a ions and beha io s o o he use s a e no eadily
a ailable. Simila o code alida ion & e i ica ion,
mechanisms ha can help make use beha io mo e
anspa en a e hus equi ed (Rossi e al. 2019). He e, we
conside p ocess mining o examine use beha io as a
po en ial means o o e coming his challenge (G2). Fo
ins ance, while da a migh cap u e he use s ha a e
in ol ed in a case, hei gene al beha io and po en ially
hei mo i a ion only become appa en when analyzing
hei ac ions ac oss cases using, e.g., compa a i e p ocess
mining.
2.4 Rela ed Wo k
Resea che s ha e explo ed p ocess mining on blockchain
da a be o e. Fi s , esea ch has ocused on ex ac ing e en
logs in he XES o ma (Mu
¨hlbe ge e al. 2019; Klink-
mu
¨lle e al. 2019,2020; Koschmide and Duchmann
2021) o a specialized, objec -cen ic log o ma (Moc a
e al. 2023; Hobeck and Webe 2023). The co esponding
publica ions ocus on ou lining echnical implemen a ion
de ails and p esen illus a i e applica ions o he app oa-
ches, in pa elying on eal-wo ld applica ions.
Second, building on da a ex ac ion capabili ies,
esea che s ha e sugges ed echnical concep s ha use
p ocess mining o audi ing and moni o ing. Di Ciccio
e al. (2020) p opose an app oach o DApp moni o ing bu
do no apply hei app oach o blockchain da a. Co adini
e al. (2019) in oduce a me hodology o DApp audi ing
ha is based on ace clus e ing and p ocess disco e y.
While hey e alua e hei me hodology on a small eal-
wo ld applica ion, hey solely ocus on measu ing he
quali y o he disco e ed models and do no p o ide
insigh s in o how hei me hodology helps o con ibu e o
anspa ency. Mu
¨lle and Ruppel (2019) demons a e an
app oach ha elies on p ocess mining o moni o he
blockchain ne wo k, bu no a single p ocess o DApp.
Finally, only wo publica ions epo insigh s om
applying p ocess mining on eal-wo ld DApp da a. Hobeck
e al. (2021) gene a e alue-adding indings by applying
p ocess mining o da a ex ac ed om a DApp deployed on
E he eum. Lamgha i (2023) applies p ocess mining o
sugges ideas o ex ending an E he eum-based game. Bo h
publica ions in e p e he abili y o de i e insigh s as a
gene al indica o o he sui abili y o p ocess mining.
Conside ing hese publica ions, he con ibu ion o his
pape is wo old. Fi s , i s udies p ocess mining as a means
o enhance he anspa ency o DApp use s in he con ex
o code alida ion & e i ica ion and use beha io anal-
ysis, aspec s ha ha e no been add essed by p io wo k.
Second, we aim o unde s and he u ili y o p ocess mining
om a me hodological pe spec i e. Tha is, we do no only
conside he applica ion o a echnique o da a, bu co e all
s eps equi ed in a p ocess-mining p ojec including e.g.,
amilia iza ion wi h a DApp, da a ex ac ion, and i e a i e
in es iga ion.
3 Resea ch Me hodology
In his pape , we adop he esea ch me hodology depic ed
in Fig. 2. As a i s s ep, we de ined ou esea ch goals. To
his end, we iden i ied cu en challenges commonly dis-
cussed in he blockchain li e a u e and a gued how p ocess
mining could in p inciple con ibu e o sol ing hem
(see Sec . 2). Consequen ly, ou goal is o p o ide deepe
insigh s in o he ac ual sui abili y o p ocess mining o
code alida ion & e i ica ion and use beha io analysis in
eal-wo ld se ings. As such, his pape can be classi ied as
p ocess-mining esea ch on he indi idual le el (i.e.,
examining speci ic asks like code e i ica ion & e i ica-
ion and use beha io analysis) and he ecosys em le el
(i.e., ocused on in e -o ganiza ional se ings such as
decen alized blockchain applica ions) ( om B ocke e al.
2021).
To in es iga e he esea ch goals, we hen de ined ou
esea ch app oach. Conside ing ha he u ili y o p ocess
mining in he con ex o blockchain anspa ency has no
been s udied ye (see Sec . 2.4), we chose an explo a o y
app oach sui ed o examine such no el phenomena (Recke
2021). We decided o analyze wo DApps using p ocess
mining, in pa icula ocusing on so wa e alida ion &
e i ica ion and use beha io analysis. In bo h analyses,
we ollowed bes p ac ices o p ocess-mining p ojec s (see
below and Fig. 3) and conduc ed all s eps equi ed o in e
insigh s om blockchain da a. These s eps include DApp
amilia iza ion, da a ex ac ion, and insigh e alua ion. The
goal is o sha e and discuss obse a ions om hese anal-
yses. The obse a ions do no only e e o he insigh s ha
we ob ained. They also co e me hodological aspec s ha
123
R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025) 781
mus be conside ed when applying p ocess mining. By
elying on such obse a ions, we go beyond a demons a-
ion o he po en ial alue o applying p ocess mining.
Ins ead, we mo e b oadly discuss con ex ual ac o s ha
impac , e.g., he eliabili y o insigh s o he e o o
gene a e hem, and hus he sui abili y o p ocess mining
o es ablishing blockchain anspa ency. No e ha we
decided o ely on ou own obse a ions due o lacking
ini ia i es and expe s ha applied p ocess mining on
blockchain da a. The implica ions o his decision o he
alidi y o ou indings a e discussed in Sec . 7.3.
A e de ining he esea ch app oach, we selec ed he
DApps, choosing Augu and Fo sage as sui able, inde-
penden DApps o he ollowing easons. Fi s , bo h
DApps ope a e on public E he eum, he mos popula
pla o m o decen alized applica ions (Wu e al. 2021;
Qasse e al. 2020). Hence, ime-s amped e en da a is
publicly accessible o bo h DApps. Second, bo h DApps
we e among he mos popula E he eum DApps a imes
5,6
,
esul ing in he a ailabili y o subs an ial olume o da a o
analyze. Thi d, Augu and Fo sage we e designed so ha
log en ies a e acked and s o ed by a cen al logging
con ac o e lec majo e en s du ing con ac execu ion,
which enables meaning ul analyses and simpli ies da a
ex ac ion. Fou h, in o ma ion on bo h DApps is widely
a ailable, such as in Augu ’s whi e pape (Pe e son e al.
2018) and Fo sage’s ma ke ing ma e ial. Fi h, because we
wan ed o s udy DApps ha we e no de eloped wi h a
p ocess-cen ic ocus, we e iewed he publicly a ailable
sou ce code and documen a ion o bo h DApps. We did
no ind any indica ion ha hey we e de eloped in a p o-
cess-d i en ashion o ha hei execu ion was adminis-
e ed by any business p ocess managemen sys em, e.g., as
p oposed by Lo
´pez-Pin ado e al. (2017). Hence, con i -
ma o y e idence ega ding ou esea ch goals implies ha
p ocess mining can be a use ul ool o blockchain appli-
ca ions, independen o whe he he applica ion is based on
a p ocess-cen ic design o no . Las ly, anspa ency is
impo an o use s o bo h DApps, as hey in es c yp-
ocu ency o ecei e some kind o e u n on in es men .
Hence, insigh s in o sou ce code alidi y and use beha io
bea he po en ial o p o ide enhanced anspa ency in o
he DApps’ us wo hiness.
We hen analyzed bo h DApps ollowing an Eck e al.
(2015)’s widely adop ed me hodology o p ocess-mining
p ojec s (see Fig. 3). We adop ed his me hodology o
ensu e ha ou analyses a e aligned wi h common p ac ices
and hence a e ep esen a i e o how analys s gene ally
conduc such p ojec s. To ha end, we collec ed da a o
bo h DApps. Fo emos , we ex ac ed log en ies om
public E he eum o gene a e e en da a. Addi ionally, we
collec ed supplemen a y da a o econs uc p ocess models
om, e.g., Augu ’s whi e pape (Pe e son e al. 2018) and
Fo sage ma ke ing ma e ial. These p epa a ion s eps a e
desc ibed in Sec . 4. As pa o he da a p epa a ion o he
p ocess-mining analyses, we il e ed he e en da a and
c ea ed iews, in pa by al e ing he case no ion o sam-
pling he da a (see also Sec . 4). In he case o Fo sage, we
en iched he e en log wi h da a on E he ans e s. Du ing
he mining and analyses, we applied p ocess-disco e y and
con o mance-checking algo i hms, as well as o he o ms
o p ocess analysis. We also documen ed ou obse a ions.
The analyses a e desc ibed in de ail in Sec . 5and Sec . 6.
We e alua ed he esul s o ou p ocess-mining p ojec s in
Fig. 2 O e iew o he esea ch app oach
Fig. 3 The p ocess-mining me hodology used o he DApp analyses (adap ed om an Eck e al. (2015))
5
Augu : h ps://dapp ada .com/dapp/augu ? ange-ha=all, accessed
04 Aug 2023.
6
Fo sage: h ps://dapp ada .com/dapp/ o sage? ange-ha=all, acces-
sed 04 Aug 2023.
123
782 R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025)
an in e iew wi h he lead a chi ec o Augu and using
li e a u e on Fo sage (Kell e al. 2021).
While we ollowed he same me hodology o bo h
DApps, he plans o he wo analyses di e ed sligh ly. By
analyzing Augu (Hobeck e al. 2021), we ini ially aimed
o unde s and he gene al u ili y o p ocess mining o
blockchain applica ions. We hence conduc ed an open-
ended analysis and among o he s ob ained insigh s ele an
o code alida ion & e i ica ion and use beha io analy-
sis. Conside ing he ele ance o hese wo opics o
blockchain esea ch (see Sec . 2.3), in his pape , we aim
o e ine he espec i e obse a ions by conduc ing he
Fo sage analysis. The esul ing i e a i e na u e o ou
esea ch app oach is e lec ed in he loop in Fig. 2.
A e ha , we consolida ed he obse a ions om he
DApp analyses and discuss he indings wi h espec o G1
and G2 in Sec . 7. Finally, we epo he esul s o ou
esea ch in his a icle.
4 Analysis P epa a ion
As depic ed in Fig. 3, be o e ex ac ing he da a we had o
plan he analysis. Besides selec ing sui able DApps, his
included becoming amilia wi h he inne wo kings o he
DApps by examining he sma con ac code and s udying
o he ma e ials, such as he documen a ion, Augu ’s whi e
pape (Pe e son e al. 2018), blog en ies, pos s on social
media, and ideos. Fo Augu , his s ep was exace ba ed by
he complexi y o he sou ce code which consis ed o 95
Solidi y sou ce code iles.
7
While he e is a cen al logging
con ac ha emi s ele an execu ion in o ma ion and
se es as an en y poin in o unde s anding he sou ce code,
he la ge numbe o sma con ac s and hea y use o
dynamic binding impeded in e p e a ion o he DApp’s
inne wo kings. Wi h Fo sage being a Ponzi scheme, we
aced a di e en challenge. As also no ed by Kell e al.
(2021), Fo sage’s implemen a ion is opaque and no all
pa s o he sou ce code a e a ailable, making i ha d o
ollow he p og am’s logic in code and in e p e he log
en ies. This was u he exace ba ed by he una ailabili y
o de elope documen a ion, o cing us o ely on second-
hand documen a ion such as epo s and blog pos s.
Obse a ion 1 Ob aining a ounda ional unde s anding o
he wo DApps and hei inne wo kings was a c i ical s ep
o da a ex ac ion and analysis, bu was exace ba ed by he
una ailabili y and complexi y o sou ce code and
documen a ion.
On a second-gene a ion blockchain, such as E he eum,
log en ies a e equen ly used by de elope s o commu-
nica e in o ma ion ela ed o he esul s o sma con ac
execu ion o o -chain componen s. We ex ac ed he log
en y da a using he open-sou ce E he eum Logging
F amewo k (ELF) (Klinkmu
¨lle e al. 2020). I allowed us
o de ine decla a i e que ies o ex ac , ans o m, and
o ma da a om E he eum-based applica ions. ELF
abs ac s many echnical de ails, such as es ablishing a
connec ion o an E he eum node o o ches a ing API calls.
Howe e , de eloping ELF que ies ha ex ac high-quali y
da a equi ed us o e ine and es he que ies in mul iple
i e a ions. In each i e a ion, we (1) inspec ed sample que y
esul s, (2) compa ed hem o o he da a sou ces, e.g., da a
om e he scan.io, and (3) e alua ed basic da a cha ac e -
is ics h ough explo a i e analysis.
Obse a ion 2 While so wa e such as ELF educed
echnical implemen a ion e o o da a ex ac ion, we s ill
needed mul iple i e a ions o diligen ly de elop que ies and
ensu e high da a quali y
Fo each DApp, ELF que y execu ion esul ed in one
e en log in he XES o ma , con aining one e en o each
log en y eco ded by he DApp.
8
Fo Augu , we ex ac ed
in o ma ion ela ed o 2,897 ma ke s om 9 July 2018 o
10 No embe 2020. The o me da e ma ks he i s exe-
cu ion o Augu 1.0, and he la e e e s o he las e en
we ex ac ed when unning he ELF mani es on 16
No embe 2020. Howe e , as ou lined in Sec . 5, he
launch o Augu 2.0 in July 2020 ende ed Augu 1.0
‘‘economically insecu e’’ and unsu p isingly caused a
decline in use in e es , which al eady s a ed a e he
announcemen o Augu 2.0 in Ap il 2020.
To accoun o his dec ease, we emo ed 162 cases ha
we e ei he c ea ed a e 2.0 was announced on 2 Ap il
2020 o we e no inalized be o e i s ac ual launch,
esul ing in 2,735 cases and 22,772 e en s. This log can be
unde s ood o co e he comple e li ecycle o Augu 1.0.
Fo Fo sage, he log is 6 GB in size and con ains
1,055,931 cases wi h 13,368,052 e en s be ween 31 Jan-
ua y 2020 and 15 Ap il 2021. The o me da e ma ks he
i s DApp execu ion, while he la e ma ks he day on
which we ex ac ed he log. Du ing ha ime DApp ac i i y
was phasing ou and a he ime o w i ing Fo sage does no
pe mi new use s o join i s E he eum e sion. Hence, ou
7
The sou ce code o Augu 1 is a ailable on: h ps://gi hub.com/
Augu P ojec /augu -co e, accessed 25 June 2024.
8
In he spi i o open science, we made a i ac s c ea ed o da a
ex ac ion and analysis a ailable online ia ou blockchain da a
collec ion (Banda a e al. 2021): h ps://ingo-webe .gi hub.io/dapp-
da a/index.h ml. Fo Augu and Fo sage, his includes sc ip s o da a
ex ac ion and high esolu ion igu es. Fo Fo sage, we also p o ide
Jupy e no ebooks o analysis. No e ha we could no sha e sc ip s
o analyses pe o med in UI-based ools such as P oM o Disco.
123
R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025) 783
log cap u es e en s om he main pe iod o ac i i y o
Fo sage.
Fo Augu , we concluded ha he ex ac ed e en log
co e s ele an p ocess s eps. In con as , o an analysis o
Fo sage, we no iced a need o addi ional in o ma ion. Tha
is because, he e en a ibu es did no p o ide a mean-
ing ul me ic o obse e paymen s o and om he sma
con ac s di ec ly, which is cen al o unde s anding he
economic wo kings and e ec s o Fo sage ac i i ies. We
hence used a second, complemen ing da a sou ce o deal
wi h he challenges. We que ied he E he scan API
9
o
e ie e egula and in e nal ansac ions o and om Fo -
sage’s cen al con ac add ess. We also ex ac ed he
co esponding ansac ion ees o all use add esses ac i e
in he log, o he ime ame o he analysis. Based on ha ,
we c ea ed a balance shee , allowing us o sum up each
use ’s o al income and spending h ough in e ac ions wi h
Fo sage. The challenges o in e p e ing Fo sage a e bes
illus a ed by a compa ison o he o al use income and
spending which we e app oxima ely 728k ETH and 768k
ETH, espec i ely. O he emaining 40.1k ETH, 39.6k
ETH can be a ibu ed o ansac ion ees, i.e., use
add esses paying he E he eum ne wo k o including
ansac ions and execu ing Fo sage sma con ac unc-
ions. A he ime o w i ing, we canno explain he
whe eabou s o he emaining 500 ETH (0.06% o he
u no e ) despi e ou e o s o ha end.
Obse a ion 3 The da a a ailable in blockchain log
en ies o Fo sage did no co e all aspec s ele an o he
analysis. We we e able o enhance he log-en y-based da a
s uc u e wi h on-chain da a on oken mo emen o es ab-
lish a da abase o ou analysis
In gene al, he ex ac ion and analysis o blockchain
da a impose speci ic demands on he compu a ional
in as uc u e. Fi s , o ha e access o his o ical da a, we
needed o se up an E he eum a chi e node o which he
cu en ha dwa e equi emen s include up o 12TB SSD
disk space
10
. Second, he size o he Fo sage e en log
posed a p oblem when applying common p ocess-mining
ools and pla o ms (P oM, Disco, and Py hon) on an o -
he-sel no ebook wi h a 1.80 GHz CPU and 16 GB RAM.
In pa icula , he used so wa e packages ailed o load he
e en log o ailed o alloca e su icien memo y when
applying p ocess-disco e y o con o mance-checking
algo i hms (mo e de ails in Sec . 6). To cope wi h he e en
log’s size and con as beha io o di e en g oups, we
c ea ed subse s o he o iginal log based on use success, as
ollows: G oup A – success ul use add esses: he 1000
add esses wi h he highes p o i s, G oup B – a e age use
add esses: 1000 add esses andomly sampled om use
add esses wi h a balance be ween median and he 75%-
pe cen ile o he o al, and G oup C – unsuccess ul use
add esses: he 1000 add esses wi h he highes losses. Fo
he analysis, e en names we e ex ended wi h he Fo sage
ma ix upg ade le el when applicable.
Obse a ion 4 Da a ex ac ion and analysis equi ed
access o su icien compu ing in as uc u e. Fo da a
analysis, we had o c ea e subse s o he Fo sage e en log
due o limi a ions o a ailable ha dwa e and so wa e.
No e ha mos da a p epa a ion, a c ucial s ep in any
p ocess-mining p ojec (see Fig. 3), was implemen ed as
pa o he ELF que ies. Howe e , depending on he
speci ic analysis, we il e ed he e en da a be o e applying
p ocess mining. Whe e ele an , de ails a e ou lined
in Sec . 5and 6.
5 P ocess-Mining Analysis o Augu
Augu ’s whi e pape (Pe e son e al. 2018) cha ac e izes
he mechanics o a p edic ion and be ing ma ke : ‘‘indi-
iduals can specula e on he ou comes o u u e e en s;
hose who o ecas he ou come co ec ly win money, and
hose who o ecas inco ec ly lose money.’’ As a be ing
ma ke o ganized on E he eum, he de elope s claim ha
Augu bypasses he disad an ages o adi ional be ing
ma ke s, such as us ed ma ke ope a o s and limi ed
pa icipa ion (Pe e son e al. 2018). Cu en ly, wo e -
sions o Augu a e a ailable in pa allel: Augu 1.0
(launched 9 July 2018) and Augu 2.0 (de ails announced
Ap il 2020
11
, launched 28 July 2020
12
). To gain use us ,
he Augu de elope s open-sou ced he sma con ac s and
deployed bo h e sions wi hou any op ion o upda e o
s op hem – as gi ing hemsel es he p i ilege o do ei he
migh esul in he loss o use s’ c yp ocu ency, so omi -
ing ha possibili y s eng hens us wo hiness. Hence, he
new e sion is deployed in pa allel o he old e sion, as
such no comp ising an upda e in any adi ional sense.
Howe e , once he new e sion was deployed and use s
mig a ed o i , he old e sion became ‘‘economically
insecu e’’ acco ding o he de elope eam, and he e o e
should no longe be used. Because p edic ion ma ke s a e
long- unning and hence ex ended obse a ion ime ames
a e c ucial o hei analysis, we ne e heless ocused on
9
h ps://docs.e he scan.io/api-endpoin s/accoun s, accessed 1 Jun
2022.
10
h ps://e he eum.o g/en/de elope s/docs/nodes-and-clien s/
a chi e-nodes/, accessed: 12 Dec 2023.
11
h ps:// wi e .com/Augu P ojec /s a us/1245715269042888706,
accessed 14 Ma 2021.
12
h ps://www.augu .ne /blog/augu - 2-launch/, accessed 14 Ma
2021.
123
784 R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025)
Obse a ion 17 Pe o mance analysis deli e ed clues o
con i m claims abou empo al on-chain ac i i y-ou come
ela ions o Fo sage.
We also ound social media pos s wi h Fo sage e e al
links and could associa e hem wi h some o he G oup A
accoun s.
18
This indica es ha accoun s wi h ac i e
ec ui ing e o s may ha e gene a ed highe income,
suppo ing he ou h claim. Howe e , sea ching o he
mos success ul accoun s o hei e e al links did no
always u n up esul s.
Obse a ion 18 Blockchain accoun pseudonymi y does
no pe mi linking indi iduals o blockchain accoun s,
hampe ing he analysis o ac o s o use success based on
o -chain componen s.
7 Discussion
We s a ed his pape by posing wo esea ch goals
ega ding he con ibu ions o p ocess mining o in o ma-
ion anspa ency in a blockchain en i onmen . In pa ic-
ula , we se a hema ic ocus on he suppo p ocess mining
can o e o add ess wo widely ecognized challenges o
he blockchain communi y: code alida ion & e i ica ion
and use beha io analysis. In he p eceding sec ions, we
applied p ocess mining o blockchain applica ions and
highligh ed ou obse a ions. These obse a ions a e
summa ized in Table 2. In his sec ion, we discuss hese
obse a ions and pu hem in con ex o he wo esea ch
goals. Subsequen ly, we discuss h ea s o alidi y.
7.1 Code Valida ion and Ve i ica ion
G1: Code Valida ion & Ve i ica ion: De e mine o wha
ex en p ocess mining can con ibu e o making
DApps mo e anspa en by suppo ing he alida ion
and e i ica ion o hei sou ce code.
In Sec . 2.3, we a gued ha p ocess mining can con ibu e
o sol ing code alida ion & e i ica ion challenges in he
blockchain domain. In he wo DApp analyses, we
obse ed ha applying p ocess mining ackles hese chal-
lenges o some deg ee. P ocess disco e y helped wi h he
coa se-g ained alida ion o a DApp-based se ice by
p o iding a con ol- low isualiza ion (O5). P ocess dis-
co e y also o e ed mo e ine-g ained insigh s when he
equency and o de o e en s we e he subjec o analysis,
helping o alida e and in alida e claims abou expec ed
beha io (O13). In his way, i deli e ed clues o iden i y a
audulen scheme implemen ed in sma con ac s (O13), a
speci ic challenge highligh ed in he blockchain li e a-
u e (Risius and Spoh e 2017; Casino e al. 2019; Zheng
e al. 2020). Con o mance checking yielded de ailed
compa isons be ween he expec ed beha io o sma
con ac s and hei ac ual execu ion. I showed con o ming
beha io (O9) and de ia ions om design speci ica ions
(O16), alida ing and in alida ing di e en unc ionali ies
o he sma con ac code. Con o mance checking also
helped o de ec a so wa e bug in sma con ac code
(O10) ha can be add essed in new o upda ed e sions o
he DApp be o e deploymen , a blockchain challenge
highligh ed by Rossi e al. (2019); Zheng e al. (2020).
Using pe o mance analysis, empo al aspec s o code
alida ion & e i ica ion can be checked (O17).
To alida e he co ec ness o ou indings and assess he
use ulness o he insigh s gene a ed by ou analyses, we
in e iewed Paul Gebheim, he chie a chi ec o Augu .
Gi en ha we only in e iewed one pe son, we classi y
esul s om his in e iew as anecdo al e idence; howe e ,
gi en his posi ion, we belie e his e idence is aluable. We
asked him o check ou assump ions – all o which he
con i med – and p esen ed in e media e esul s om ou
analyses o him. F om his pe spec i e, using p ocess
mining o analyzing DApps gene ally, and Augu , in
pa icula , p o ides alue in h ee ways. Fi s , i helps o
e i y he design mechanisms and check o unin ended
beha io and bugs in he (immu able) code; immu abili y
poses a challenge om a BPM pe spec i e (Mendling e al.
2018) and so wa e enginee ing in gene al (Webe and
S aples 2021). Second, p ocess mining p o ides a clea
iew o how an applica ion is used, which is also help ul
o designing upda ed e sions o an applica ion. Thi d, i
has g ea po en ial o echnical and economic secu i y
analysis, e.g., an audi o could c ea e a model and con-
o mance-check i agains ac ual use beha io . Also, e en
hough a sma con ac ypically implemen s a ixed se o
ules, analyses o p ocess a iabili y may e eal aluable
insigh s ha could help e ol e u u e e sions o he sma
con ac , e.g., o align hem be e wi h changed use
expec a ions.
Based on hese indings, we in e ha p ocess mining
can con ibu e o code alida ion & e i ica ion e o s
in ended o s eng hen DApp anspa ency. Like o he
means o isualiza ion and analysis ( an Wijk 2005), he
u ili y o p ocess mining o code alida ion & e i ica ion
depends on whe he use s can expec he alue o he
insigh s o exceed he cos o gene a e hem. In his ega d,
we encoun e ed a ew issues ha migh impac he cos o
alue o p ocess-mining insigh s.
Familia iza ion wi h he DApp and i s implemen a ion is
a c ucial s ep o p epa ing da a ex ac ion and analysis.
Cos s o execu e his s ep can be a ec ed by he complexi y
18
h ps://www.you ube.com/c/Se geyMaslako p o i biz, accessed 22
Jun 2022.
123
R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025) 791
and una ailabili y o in o ma ion and code (O1). The cos s
migh u he be a ec ed when he log en ies emi ed by
he DApp do no su icien ly co e ele an p ocess s eps
(O3). In he case o he Ponzi scheme Fo sage, his was
p esumably due o he de elope s ying o obscu e hei
in en ions. Also, emi ing log en ies (on E he eum) gen-
e ally incu s ansac ion ees (Wood e al. 2014) ha a e
ul ima ely paid by DApp use s and migh hence be a oided
o no jeopa dize DApp usage. Mo eo e , ensu ing da a
quali y equi ed mul iple i e a ions o de eloping and
es ing ex ac ion sc ip s (O2). Tes ing o he da a ex ac-
ion is a common s ep in any p ocess-mining p ojec .
Howe e , in he con ex o blockchain applica ions he
complexi y (and hence he cos s) o da a ex ac ion is
highe when log en ies a e insu icien and use s need o
eso o o he on-chain da a sou ces (e.g., blocks, ans-
ac ions, and ansac ion eplay), o o -chain da abases
(O3). Las ly, cos s a e po en ially incu ed by ha dwa e and
capaci y equi emen s ha mus be sa is ied o synch onize
and access he da a o a blockchain node o o pe o m da a
analysis (O4).
Rega ding he alue gene a ed by p ocess mining, he e
a e a ew cons ain s. In gene al, p ocess mining ocuses on
beha io al aspec s and migh hence no be app op ia e
when use s wan o alida e & e i y aspec s beyond DApp
beha io , such as sa e y o cos s.
Simila ly, we did no apply p ocess mining as a design
ime o p e-deploymen es o so wa e ulne abili ies.
Tha is, we elied on e en logs ha consis o blockchain
e en s ha we e emi ed a un ime, i.e., when code seg-
men s we e in oked by use s o he deployed DApps.
Consequen ly, we could only analyze code sec ions and
beha io when hey we e execu ed du ing un ime. Code
ha was no execu ed emained hidden om ou analysis.
Ne e heless, ou DApp analyses show ha p ocess mining
can se e as a ool o de ec bugs and pe o mance issues
o blockchain applica ions pos -deploymen (based on
ac ual code execu ion). No e ha use s could, in p inciple,
also apply p ocess mining o s a ic analysis be o e code
deploymen – howe e , his is ou side he scope o his
a icle.
Las ly, some p ocess-mining analyses (in pa icula
con o mance checking) hinge on he documen a ion and
Table 2 Obse a ions (O) om he p ocess-mining p ojec s on Augu and Fo sage
Augu Fo sage
O1: Ob aining a ounda ional unde s anding o he wo DApps and hei inne wo kings was a c i ical s ep o da a ex ac ion and analysis, bu
was exace ba ed by he una ailabili y and complexi y o sou ce code and documen a ion
O2: While so wa e such as ELF educed echnical implemen a ion e o o da a ex ac ion, we s ill needed mul iple i e a ions o diligen ly
de elop que ies and ensu e high da a quali y
O4: Da a ex ac ion and analysis equi ed access o su icien compu ing in as uc u e. Fo da a analysis, we had o c ea e subse s o he Fo sage
e en log due o limi a ions o a ailable ha dwa e and so wa e
O5: P ocess disco e y helped o isualize Augu ’s sma con ac
execu ion and hus o conduc a i s coa se-g ained alida ion o he
DApp’s implemen a ion O6: P ocess-explo a ion helped plo an
o e iew o DApp execu ion da a, showing use ac i i y g adien s
h oughou Augu ’s li e cycle, incl. seasonal peak ac i i y in e als O7:
Con o mance checking equi ed in o ma ion abou a no ma i e p ocess
ha we had o ans e in o a p ocess model o ma sui able o he
con o mance-checking algo i hm O8: Blockchain accoun
pseudonymi y allowed us o ela e a beha io al pa e n o a use ,
al hough pe sonal in o ma ion abou he indi idual was no a ailable
O9: Using con o mance checking, we ound ha in mos ins ances
Augu ’s execu ion could be explained wi h a no ma i e p ocess
desc ip ion. We also ound hin s owa ds use -implemen ed au oma ion
p o ocols in e ac ing wi h Augu ’s applica ion in e ace. O10: Using
con o mance checking, we ound a de ia ion om he expec ed DApp
beha io ha u ned ou o be a bug in Augu ’s sma con ac code
O11: Wi h he help o pe o mance analysis, we obse ed use s a egy
adjus men s and ma u ing e ec s in he communi y beha io ,
s eamlining ma ke p ocesses o e ime
O3: The da a a ailable in blockchain log en ies o Fo sage did no
co e all aspec s ele an o he analysis. We we e able o enhance he
log-en y-based da a s uc u e wi h on-chain da a on oken mo emen
o es ablish a da abase o ou analysis O12: Blockchain accoun
add esses p o ide a no ion o assign e en s o ac o s on he blockchain.
The pseudonymi y o he accoun s, howe e , did no pe mi ela ing
use accoun s o indi iduals in e ac ing wi h he applica ion O13: Wi h
p ocess explo a ion and p ocess disco e y, we espec i ely con i med
o alsi ied claims abou ou comes o usage scena ios o Fo sage, based
on e en o de and equency. We ound indica o s o a audulen
scheme using ime-sensi i e p o i /loss analysis O14: Wi h p ocess
explo a ion and p ocess disco e y, we analyzed use engagemen wi h
Fo sage and use s a egies ac oss use g oups, including s a egy
adjus men s o e ime O15: Con o mance checking equi ed
in o ma ion abou a no ma i e p ocess, which we cap u ed in
p ecedence ules ha se ed as inpu o he con o mance-checking
algo i hm O16: Using con o mance checking, we es ed he mechanics
o Fo sage and we e able o poin ou de ia ions be ween he DApp’s
communica ed design speci ica ions and he execu ion o he sma
con ac code O17: Pe o mance analysis deli e ed clues o con i m
claims abou empo al on-chain ac i i y-ou come ela ions o Fo sage
O18: Blockchain accoun pseudonymi y does no pe mi linking
indi iduals o blockchain accoun s, hampe ing he analysis o ac o s
o use success based on o -chain componen s
123
792 R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025)
desc ip ions o he DApps’ o-be p ocesses (O7 and O15),
which may no be exhaus i e o a ailable in ull. Fo
beha io s ha a e no documen ed and/o ha use s migh
no be awa e o , hey will hus no be able o alida e &
e i y hem based on he ac ual code execu ion.
7.2 Use Beha io Analysis
G2: Use Beha io Analysis: De e mine o wha ex en
p ocess mining can con ibu e o making DApps
mo e anspa en by suppo ing he analysis o hei
use s’ beha io .
As poin ed ou in Sec . 2.3, he blockchain li e a u e ec-
ognizes use beha io analysis as a esea ch challenge. Ou
obse a ions om he DApp analyses imply ha p ocess
mining can con ibu e o unde s anding use beha io . The
isualiza ions om p ocess explo a ion p o ided insigh s
abou use beha io on a mac o le el, including waxing
and waning use ac i i y in he applica ion’s li e ime and
peak ac i i y phases (O6). Gaining mo e ine-g ained
insigh s in o use beha io equi ed combining se e al
indings om p ocess explo a ion and p ocess disco e y.
They included compa isons o beha io be ween use
g oups and in o ma ion on use s a egies adjus ing o a
changing applica ion en i onmen (O14). Addi ional
beha io al adjus men s o e ime, e.g., ma u ing e ec s in
use beha io , can also be obse ed wi h pe o mance
analysis (O11). Pe o mance analysis also helped o
examine beha io ac oss use g oups in mo e dep h, wi h
in o ma ion on use eac ion ime (O17). A di e en aspec
o use beha io became e iden om con o mance
checking, whe e esul s sugges ha use s au oma ed pa
o hei in e ac ion wi h a DApp and he eby, by p oxy,
hei in e ac ion wi h o he use s (O9).
Gi en he insigh s p esen ed abo e, we in e ha p ocess
mining can indeed con ibu e o DApp anspa ency by
suppo ing he analysis o use beha io . Simila o code
alida ion & e i ica ion, conside a ions ela ed o he cos
and alue associa ed wi h applying p ocess mining migh
limi he u ili y o p ocess mining in a speci ic con ex .
Tha is, amilia iza ion can be exace ba ed by he com-
plexi y and una ailabili y o in o ma ion and code (O1),
da a ex ac ion equi es ho ough es ing (O2), and com-
pu e & s o age capaci y equi emen s mus be me (O4).
Rega ding he deg ee o which log en ies p o ide ele an
da a (O3), we no e ha essen ial use ac i i y migh no be
o ganized on-chain, bu o -chain, e.g., use ec ui ing on
social media pla o ms in Fo sage. Accoun pseudonymi y
allowed assigning e en s o accoun s in he i s place (O8,
O12). Howe e , in pa due o he accoun pseudonymi y,
es ablishing da a connec i i y be ween o - and on-chain
da a is challenging (O12, O18). In ac , du ing ou
analyses, we lea ned abou o -chain ac i i y bu could no
eliably ela e such da a wi h p ocess pa icipan s in ou
e en da a on a la ge scale (O18), apa om a ew
excep ions whe e pseudonyms we e disclosed.
The pseudonymi y o blockchains does no only impac
analysis cos s bu also limi s he alue o applying p ocess
mining o analyze use beha io . In ou analyses (O11,
O17), we in e p e ed esul s unde he assump ion ha
e e y E he eum accoun add ess ep esen s a unique use .
The pseudonymi y o accoun s on E he eum does no
gua an ee ha assump ion, so ha mul iple accoun s may
belong o he same use (s) (O12). Fo example, o Fo -
sage, ou loss-p o i calcula ions may no ep esen some
use ’s ne gains h ough Fo sage. We also canno dis in-
guish mul iple use s copying each o he ’s beha io , o a
single use applying he same (possibly au oma ed)
beha io mul iple imes. Finally, one (o mul iple) accoun s
migh be con olled by a g oup o eam o use s. We
add essed he issue by di e en ia ing use s om use
add esses in ou w i ing (see Sec . 6), bu he limi a ions
on he possible insigh s emain.
7.3 Th ea s o Validi y
The e a e se e al h ea s o he alidi y o ou indings
(Wohlin e al. 2012, p. 68). Fo ou analyses, we ook on
he ole o conduc o s o p ocess-mining ac i i ies and
obse e s examining he p ocess-mining ac i i ies. Ou
aking on he ole o conduc o s does no a ec he ex e nal
alidi y o ou analyses. On he one hand, we we e no
in ol ed in he de elopmen o he DApps and hence ook
on he ole o DApp use s whose goal is o make hese
DApps mo e anspa en . On he o he hand, we ollowed a
commonly applied me hodology o applying p ocess
mining (see Sec . 3). Howe e , we migh ha e in oduced a
con i ma ion bias o ou indings. The bias is mi iga ed o a
ce ain deg ee as we ini ially conduc ed he Augu analysis
open-ended, be o e inalizing his pape ’s esea ch goals.
Fu he mo e, ou analyses esul s a e cons ained by a
ew h ea s o in e nal alidi y. We migh ha e in oduced a
bias in ou con o mance-checking app oach o Augu . As
a basis o con o mance checking, we used he en i e
no ma i e p ocess model (see Fig. 4) and hus he o e all
con ol low wi hou checking he ga e condi ions o
indi idual cases. Tha migh ha e led o o e ly gene alized
esul s, igno ing non-con o ming cases. Simila ly, we
migh no ha e included all possible combina ions o ules
in he ule-based con o mance-checking app oach and
hence migh ha e missed non-con o ming cases in Fo sage.
The e is also a chance ha G oup B use add esses we e
no ep esen a i e due o he andom sampling. We migh
ha e in oduced a o m o selec ion bias by he choice o
con o mance-checking echniques, applying an alignmen -
123
R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025) 793
based echnique o Augu ’s well-s uc u ed ma ke p ocess
and ule checking o Fo sage’s lexible in es men p ocess.
Gi en he na u e o he p ocesses, hese choices a e sen-
sible bu migh ha e in luenced he quali y o he analysis
esul s.
An ex e nal h ea conce ns ou selec ion o DApps. We
in ended o s eng hen he gene alizabili y by conduc ing
wo DApp analyses bu migh ha e in oduced a bias by
ou choice o use cases – di e en DApps migh no ha e
had de ia ions be ween hei speci ica ions and ac ual code
execu ion. Simila ly, ano he DApp’s log en ies migh no
ha e co e ed ele an sec ions o he code execu ion
leading o less insigh s wi h espec o he esea ch goals.
Ano he ex e nal h ea o he s udy may be ha he da a we
pe o med ou analysis on we e incomple e o i s quali y
was co up ed. We did, howe e , ake p ecau ions in
educing hese h ea s by alida ing in e media e esul s
and indings wi h Augu ’s use in e ace and hei chie
a chi ec and c oss-checking ou esul s o Fo sage wi h
indings by Kell e al. (2021).
8 Conclusion and Fu u e Wo k
In o ma ion anspa ency is a ibu ed o blockchain en i-
onmen s. Achie ing such anspa ency in p ac ice, how-
e e , depends on he a ailabili y o in o ma ion and
adequa e da a p ocessing. We sugges ed ha p ocess
mining could p o ide a da a p ocessing oolki i o con-
ibu e o blockchain anspa ency by add essing wo
blockchain esea ch challenges: (1) code alida ion &
e i ica ion and (2) use beha io analysis. We conduc ed
wo p ocess-mining analyses o da a ex ac ed om he
blockchain applica ions Augu and Fo sage. To his end,
we used ELF o ex ac da a o e essen ially he en i e
li ecycle o Augu 1.0 and Fo sage in i s wo-ma ix se -
up. We used p ocess-mining me hods and ools o explo e
he da a, disco e p ocess models, and conduc con o -
mance-checking and pe o mance analyses. We we e able
o show de ia ions om he expec ed code execu ion, we
de ec ed a bug in Augu ’s sma con ac code, and we
alsi ied h ee o he ou main claims abou he sys em
om Fo sage’s p omo ional ma e ial. P ocess mining was
also help ul o examining use beha io , including indi-
idual use ac i i ies and use s a egies and hei adjus -
men s o e ime. Finally, we in e iewed he chie a chi ec
o Augu o alida e ou insigh s and unde s and hei
use ulness, and c oss-checked ou Fo sage esul s wi h
academic publica ions. In summa y, we iden i ied pa e ns
and mo i es in he blockchain da a using p ocess mining.
The e o e, we conclude ha he e is s ong suppo ha
p ocess mining con ibu es o es ablishing anspa ency in
blockchain en i onmen s in e ms o (1) code alida ion &
e i ica ion and (2) use beha io analysis.
A c i ical di ec ion o u u e esea ch is he de elop-
men o me hods and ools ha help use s wi h DApp
amilia iza ion, da a ex ac ion, and insigh alida ion –
h ee a eas ha incu ed cos s o limi ed he u ili y o
p ocess-mining insigh s in ou analyses. Mo eo e , he da a
basis can be ex ended beyond log en ies and oken ans-
ac ions, e.g., o comp ise eplayed ansac ion aces in
o de o ensu e comple eness o he execu ion da a. Also,
ou use cases elied on his o ic da a gene a ed pa ially
yea s be o e he ex ac ion. Real- ime moni o ing could
p o ide imely insigh s in o blockchain ope a ions enabling
swi esponses o i egula i ies. In eg a ing p ocess mining
in o he de elopmen li ecycle o sma con ac s could
p o ide p e-deploymen and pos -deploymen moni o ing,
ensu ing ha sma con ac s pe o m as in ended. This
in eg a ion could also acili a e con inuous imp o emen o
blockchain applica ions, aligning hem wi h use expec a-
ions and equi emen s.
Acknowledgemen s We a e e y g a e ul o he inpu o Paul
Gebheim, chie a chi ec a he Augu P ojec . We would also like o
hank Ma in Rebesky and Hend ik Bock a h o w i ing he i s
e sions o he ELF mani es s o ex ac Augu and Fo sage e en
logs.
Funding Open Access unding enabled and o ganized by P ojek
DEAL.
Open Access This a icle is licensed unde a C ea i e Commons
A ibu ion 4.0 In e na ional License, which pe mi s use, sha ing,
adap a ion, dis ibu ion and ep oduc ion in any medium o o ma , as
long as you gi e app op ia e c edi o he o iginal au ho (s) and he
sou ce, p o ide a link o he C ea i e Commons licence, and indica e
i changes we e made. The images o o he hi d pa y ma e ial in his
a icle a e included in he a icle’s C ea i e Commons licence, unless
indica ed o he wise in a c edi line o he ma e ial. I ma e ial is no
included in he a icle’s C ea i e Commons licence and you in ended
use is no pe mi ed by s a u o y egula ion o exceeds he pe mi ed
use, you will need o ob ain pe mission di ec ly om he copy igh
holde . To iew a copy o his licence, isi h p://c ea i ecommons.
o g/licenses/by/4.0/.
Re e ences
Acampo a G, Vi iello A, S e ano B, an de Aals W, Gu
¨n he C,
Ve beek E (2017) IEEE 1849: he XES s anda d. IEEE Compu
In ell Mag 12(2):4–8
And ews R, Su iadi S, Wynn M, e Ho s ede AHM, Ro hwell S
(2018) Imp o ing pa ien lows a S . And ew’s Wa Memo ial
Hospi al’s eme gency depa men h ough p ocess mining. In:
Business p ocess managemen cases: digi al inno a ion and
business ans o ma ion in p ac ice, pp 311–333
Banda a HD, Bock a h H, Hobeck R, Klinkmu
¨lle C, Pu ahl L,
Rebesky M, an de Aals W, Webe I (2021) E en logs o
e he eum-based applica ions. In: BPM’21: in e na ional con e -
ence on business p ocess managemen , Rome, I aly
123
794 R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025)
Ba ole i M, Ca a S, Cimoli T, Saia R (2019) Dissec ing ponzi
schemes on E he eum: iden i ica ion, analysis, and impac .
a Xi :1703.03779. Accessed 22 Jun 2022
Ca mona J, Dongen B, Sol i A, Weidlich M (2018) Con o mance
checking: ela ing p ocesses and models. Sp inge , Heidelbe g
Casino F, Dasaklis TK, Pa sakis C (2019) A sys ema ic li e a u e
e iew o blockchain-based applica ions: cu en s a us, classi-
ica ion and open issues. Telema In o m 36:55–81
Co adini F, Ma can oni F, Mo iche a A, Polini A, Re B, Sampaolo
M (2019) Enabling audi ing o sma con ac s h ough p ocess
mining. In: F om so wa e enginee ing o o mal me hods and
ools, and back, pp 467–480
De Wee d J, Schupp A, Vande loock A, Baesens B (2013) P ocess
mining o he mul i- ace ed analysis o business p ocesses—a
case s udy in a inancial se ices o ganiza ion. Compu Ind
64(1):57–67
Di Ciccio C e al (2018) Blockchain-based aceabili y o in e -
o ganisa ional business p ocesses. In: Business modeling and
so wa e design, pp 56–68
Di Ciccio C, Me oni G, Plebani P (2020) Business p ocess moni o -
ing on blockchains: po en ials and challenges. In: En e p ise,
business-p ocess and in o ma ion sys ems modeling, pp 36–51
Hobeck R, Webe I (2023) Towa ds objec -cen ic p ocess mining o
blockchain applica ions. In: Ko
¨pke J, Lo
´pez-Pin ado O, Pla au
R, Rehse JR, Gdowska K, Gonzalez-Lopez F, Munoz-Gama J,
Smi K, an de We JMEM (eds) Business p ocess manage-
men : blockchain, obo ic p ocess au oma ion and educa o s
o um. Sp inge , Cham, pp 51–65
Hobeck R, Klinkmu
¨lle C, Banda a HD, Webe I, an de Aals W
(2021) P ocess mining on blockchain da a: a case s udy o
Augu . In: BPM’21: In e na ional con e ence on business
p ocess managemen , I aly, Rome, pp 306–323
IEEE Task Fo ce on P ocess Mining (2011) P ocess mining
mani es o. In: Bpm wo kshops, LNBIP, ol 99. Sp inge ,
Heidelbe g
Jans M, an de We JM, Lybae N, Vanhoo K (2011) A business
p ocess mining applica ion o in e nal ansac ion aud mi iga-
ion. Expe Sys Appl 38(10):13351–13359
Kell T, Yousa H, Allen S, Meiklejohn S, Juels A (2021) Fo sage:
ana omy o a sma -con ac py amid scheme. CoRR abs/
2105.04380. Accessed 22 Jun 2022
Klinkmu
¨lle C, Mu
¨lle R, Webe I (2019) Mining p ocess mining
p ac ices: an explo a o y cha ac e iza ion o in o ma ion needs
in p ocess analy ics. In: Hildeb and T, an Dongen BF,
Ro
¨glinge M, Mendling J (eds) Business p ocess managemen .
Sp inge , Cham, pp 322–337
Klinkmu
¨lle C, Ponoma e A, T an AB, Webe I, an de Aals WMP
(2019) Mining blockchain p ocesses: ex ac ing p ocess mining
da a om blockchain applica ions. In: BPM blockchain o um,
pp 71–86
Klinkmu
¨lle C, Webe I, Ponoma e A, T an AB, an de Aals W
(2020) E icien logging o blockchain applica ions. CoRR abs/
2001.10281. Accessed 21 Ma 2021
Koschmide A, Duchmann F (2021) Ex ac ion o meaning ul e en s
o p ocess mining om blockchain. Sp inge , Cham, pp 13–29
Lamgha i Z (2023) Towa ds he p ocess mining applicabili y in he
chickenhun blockchain game. In J Compu Digi al Sys
13(1):1–1
Leemans SJ, Poppe E, Wynn MT (2019) Di ec ly ollows-based
p ocess mining: explo a ion and a case s udy. In: 2019 in e na-
ional con e ence on p ocess mining (ICPM), pp 25–32. h ps://
doi.o g/10.1109/ICPM.2019.00015
Lei e JCSP, Cappelli C (2010) So wa e anspa ency. Bus In Sys
Eng 2(3):127–139. h ps://doi.o g/10.1007/s12599-010-0102-z
Lemos AM, Sabino CC, Lima RMF, Oli ei a CAL (2011) Using
p ocess mining in so wa e de elopmen p ocess managemen : a
case s udy. In: 2011 IEEE in e na ional con e ence on sys ems,
man, and cybe ne ics, pp 1181–1186
Leona di PM, T eem JW (2020) Beha io al isibili y: a new
pa adigm o o ganiza ion s udies in he age o digi iza ion,
digi aliza ion, and da a ica ion. O gan S ud 41(12):1601–1625
Le ia IA, Go on A (2015) Model checking as suppo o inspec ing
compliance o ules in lexible p ocesses. J Vis Lang Compu
28:100–121
Lo
´pez-Pin ado O, Ga cı
´a-Ban
˜uelos L, Dumas M, Webe I (2017)
Ca e pilla : a blockchain-based business p ocess managemen
sys em. BPM (Demos) 172
Mans R, Schonenbe g MH, Song M, an de Aals WMP, Bakke P
(2009) Applica ion o p ocess mining in heal hca e: a case s udy
in a Du ch hospi al. Biomed Eng Sys Technol 25:425–438
Mendling J e al (2018) Blockchains o business p ocess manage-
men –challenges and oppo uni ies. ACM T ansac Manag In
Sys (TMIS) 9(1):4:1-4:16
Moc a M’Baba L, Assy N, Sellami M, Gaaloul W, Fa ouk Nanne M
(2023) P ocess mining o a i ac -cen ic blockchain applica-
ions. Sim Model P ac Theo 127(102):779. h ps://doi.o g/10.
1016/j.simpa .2023.102779
Mu
¨hlbe ge R, Bachho ne S, Di Ciccio C, Ga cı
´a-Ban
˜uelos L,
Lo
´pez-Pin ado O (2019) Ex ac ing e en logs o p ocess
mining om da a s o ed on he blockchain. In: Business p ocess
managemen wo kshops, pp 690–703
Mu
¨lle M, Ruppel P (2019) P ocess mining o decen alized
applica ions. In: IEEE in e na ional con e ence on decen alized
applica ions and in as uc u es, pp 164–169
Pe e son J, K ug J, Zol u M, Williams AK, Alexande S (2018)
Augu : a decen alized o acle and p edic ion ma ke pla o m.
Technical epo , Fo ecas Founda ion. h ps://gi hub.com/Augu
P ojec /whi epape /blob/mas e / 1/english/whi epape .pd .
Accessed 05 Jan 2021
P us y N (2017) Building blockchain p ojec s. Pack , Bi mingham
Qasse IA, Spillne J, Talib MA, Nasi Q (2020) A s udy on Ðapps
cha ac e is ics. In: 2020 IEEE in e na ional con e ence on
decen alized applica ions and in as uc u es (DAPPS),
pp 88–93
Recke J (2021) Scien i ic esea ch in in o ma ion sys ems: a
beginne ’s guide. Sp inge , Heidelbe g
Reinkemeye L (2020) P ocess mining in ac ion: p inciples, use cases
and ou look. Sp inge , Heidelbe g
Risius M, Spoh e K (2017) A blockchain esea ch amewo k: wha
we (don’ ) know, whe e we go om he e, and how we will ge
he e. Bus In Sys Eng 59:385–409
Rossi M, Muelle -Bloch C, Tha che JB, Beck R (2019) Blockchain
esea ch in in o ma ion sys ems: cu en ends and an inclusi e
u u e esea ch agenda. J Assoc In Sys 20(9):14
Ro ani M, Maggi FM, Leoni M, an de Aals WMP (2015)
Decla a i e p ocess mining in heal hca e. Expe Sys Appl
42(23):9236–9251
Rozina A, de Jong ISM, Gu
¨n he CW, an de Aals WMP (2009)
P ocess mining applied o he es p ocess o wa e scanne s in
ASML. IEEE T ans Sys Man Cybe n Pa C 39(4):474–479
Sha ma P, Jindal R, Bo ah MD (2023) A e iew o sma con ac -
based pla o ms, applica ions, and challenges. Clus e Compu
26(1):395–421
Su iadi S, Mans RS, Wynn MT, Pa ing on A, Ka non J (2014)
Measu ing pa ien low a ia ions: a c oss-o ganisa ional p ocess
mining app oach. In: Asia paci ic business p ocess managemen ,
pp 43–58
Vacca A, Di So bo A, Visaggio CA, Can o a G (2021) A sys ema ic
li e a u e e iew o blockchain and sma con ac de elopmen :
echniques, ools, and open challenges. J Sys So w
174(110):891
123
R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025) 795
an de Aals WMP e al (2007) Business p ocess mining: an
indus ial applica ion. In Sys 32(5):713–732
an de Aals WMP (2016) P ocess mining: da a science in ac ion.
Sp inge , Heidelbe g
an Eck ML, Lu X, Leemans SJJ, an de Aals WMP (2015) PM2: a
p ocess mining p ojec me hodology. In: Zd a ko ic J, Ki iko a
M, Johannesson P (eds) Ad anced in o ma ion sys ems engi-
nee ing. Sp inge In e na ional, Cham, pp 297–313
an Wijk J (2005) The alue o isualiza ion. In: Vis 05. IEEE
isualiza ion, pp 79–86. h ps://doi.o g/10.1109/VISUAL.2005.
1532781
om B ocke J, Jans M, Mendling J, Reije s HA (2021) A i e-le el
amewo k o esea ch on p ocess mining. Bus In Sys Eng 1–8
Webe I, S aples M (2021) P og ammable money: nex -gene a ion
condi ional paymen s using blockchain—keyno e pape . In:
In e na ional con e ence on cloud compu ing and se ices
science (CLOSER)
Webe I, Xu X, Ri e e R, Go e na o i G, Ponoma e A, Mendling J
(2016) Un us ed business p ocess moni o ing and execu ion
using blockchain. In: In e na ional con e ence on business
p ocess managemen , Rio de Janei o, B azil
Webe I e al (2017) On a ailabili y o blockchain-based sys ems. In:
IEEE in e na ional symposium on eliable dis ibu ed sys ems
(SRDS), pp 64–73
Wohlin C, Runeson P, Ho
¨s M, Ohlsson MC, Regnell B, Wessle
´nA
(2012) Expe imen a ion in so wa e enginee ing. Sp inge ,
Heidelbe g
Wood G e al (2014) E he eum: a secu e decen alised gene alised
ansac ion ledge . E he eum p ojec yellow pape 151:1–32
Wu K, Ma Y, Huang G, Liu X (2021) A i s look a blockchain-based
decen alized applica ions. So w P ac Exp 51(10):2033–2050
Xu X, Webe I, S aples M (2019) A chi ec u e o blockchain
applica ions. Sp inge , Heidelbe g
Zheng Z, Xie S, Dai HN, Chen X, Wang H (2018) Blockchain
challenges and oppo uni ies: a su ey. In J Web G id Se
14(4):352–375
Zheng Z, Xie S, Dai HN, Chen W, Chen X, Weng J, Im an M (2020)
An o e iew on sma con ac s: challenges, ad ances and
pla o ms. Fu u Gene Compu Sys 105:475–491
123
796 R. Hobeck e al.: On he Sui abili y o P ocess Mining o Enhancing. . ., Bus In Sys Eng 67(6):777–796 (2025)
