Challenges to Support a Ppi Management Lifecycle

Challenges To Suppo A PPI Managemen Li ecycle∗ Adela del-Río-O ega Manuel Resinas An onio Ruiz-Co és ETS Ingenie ía In o má ica ETS Ingenie ía In o má ica ETS Ingenie ía In o má ica Uni . de Se illa Uni . de Se illa Uni . de Se illa [email p o ec ed] [email p o ec ed] [email p o ec ed] Abs ac An impo an aspec in he business p ocess li ecycle is he e alua ion o business p ocesses pe o mance, since i helps o ganisa ions o de- ine and measu e p og ess owa ds hei goals. Pe o mance equi emen s on business p o- cesses can be speci ied by means o P ocess Pe o mance Indica o s (PPIs), as sugges ed in many me hodologies and amewo ks like, o ins ance, COBIT, ITIL o EFQM. As a consequence, i is con enien o in eg a e he managemen o PPIs in o he whole business p ocess li ecycle om i s design o i s e alu- a ion, enabling hus a mo e effec i e an effi- cien au oma ed suppo o ex ac in o ma- ion om such indica o s. In his pape , we analyse some app oaches ela ed o his issue and iden i y he challenges ha need o be aced. 1 In oduc ion I is inc easingly impo an o e alua e he pe o mance o business p ocesses. A key ins umen o ca y ou his e alua ion is by means o P ocess Pe o mance Indica o s (PPIs). A PPI is a measu e ha e lec s he c i ical success ac o s o a business p ocess de ined wi hin an o ganisa ion, in which i s a ge alue mus be eached in a ce ain pe- iod and e lec s he objec i es pu sued by he o ganisa ion wi h ha business p ocess. No e ha we use PPI as a kind o Key Pe o mance ∗This wo k has been pa ially suppo ed by he Eu opean Commission (FEDER), and SETI (TIN2009-07366); and p ojec P07-TIC-2533 unded by he Andalusian local Go e nmen . ! "#$%&'()*+$%'#$$) ,-./#$$) 01#'2%3%/42%.')4'1) 5.1#66%'& ! 7'468$%$() 946%142%.':) ;%<+642%.')4'1) 9#-%3%/42%.' ! ,-./#$$)5%'%'&) 4'1)*75 =>46+42%.' "#$%&')4'1) 7'468$%$ ! ;8$2#<)$#6#/2%.':) 0<?6#<#'242%.') 4'1)@#$2)4'1) "#?6.8<#'2 ! A?#-42%.':) 5.'%2.-%'&)4'1) 54%'2#'4'/# =>46+42%.' 7'468$%$ B.'3%&+-42%.'='4/2<#'2 Figu e 1: Business P ocess Li ecycle as desc ibed in [4] Indica o (KPI) ha ocuses exclusi ely on he indica o s de ined on he business p o- cesses. Nowadays, many me hodologies and amewo ks like, o ins ance, COBIT, ITIL o he EFQM excellence model [1, 2, 3], con i m his impo ance by including he de ini ion o hese PPIs wi hin hei ecommenda ions as a means o e alua e he pe o mance o he exis ing business p ocesses. In o de o make his e alua ion o business p ocesses easie , i is con enien o in eg a e he managemen o PPIs in o he whole business p ocess li ecy- cle. In [4], he ou -phase business p ocess li ecycle o Figu e 1 is p oposed. This li ecycle s a s wi h he Design and analysis phase, in which business p ocesses a e iden i ied, mod- elled using a pa icula no a ion, alida ed o check whe he all alid p ocess ins ances a e e lec ed by i s co esponding business p ocess model, simula ed o de ec possible undesi ed Ac as de los Talle es de las Jo nadas de Ingenie ía del So wa e y Bases de Da os, Vol. 4, No. 4, 2010 ISSN 1988-3455 SISTEDES, 2010 8 execu ion sequences, and e i ied. Once he business p ocess model is designed and e i- ied, i needs o be implemen ed. This is done du ing he con igu a ion phase, whe e, in case a dedica ed so wa e sys em is needed, i mus be selec ed and con igu ed in o de o ake in o accoun he in e ac ions o he employees wi h he sys em and he in eg a ion wi h exis ing so wa e sys ems. Then, he implemen a ion o business p ocess needs o be es ed and i- nally deployed in i s a ge en i onmen . Nex phase is he enac men , ha encompasses he un ime o he business p ocess. On he one hand, a co ec o ches a ion is necessa y o he business ac i i ies o be pe o med acco d- ing o he business p ocess’s execu ion con- s ain s. On he o he hand, p ocess moni- o ing is a good mechanism o p o iding in- o ma ion abou he s a us o business p ocess ins ances. Finally, he e alua ion phase uses in o ma ion a ailable o e alua e and imp o e business p ocess models and hei implemen- a ions. No e ha he e no exis s a s ic empo al o de ing in which hese phases need o be execu ed; inc emen al and e olu iona y app oaches in ol ing concu en ac i i ies in mul iple phases a e, hus, common. In his pa- pe , we a e in e es ed in desc ibing wha has been done in he li e a u e ega ding he in- eg a ion o PPIs as i s -class ci izens in his business p ocess li ecycle, and in ou lining he open challenges his issue p esen s. The emainde o his pape is o ganised as ollows. In Sec ion 2 we explain how o in e- g a e PPIs in o he business p ocess manage- men li ecycle and analyse how diffe en ap- p oaches deal wi h his issue. F om his sec- ion, a numbe o open esea ch ques ions a e iden i ied and p esen ed in Sec ion 3. Finally, Sec ion 4 d aws he conclusions om ou wo k, summa izes he pape and ou lines ou u u e wo k. 2 In eg a ing PPIs In o The Busi- ness P ocess Managemen Li ecy- cle In o de o make he e alua ion o he pe o - mance o business p ocesses easie , i is con- enien o in eg a e he managemen o PPIs in o he whole business p ocess li ecycle. In his sec ion, we ake he business p ocess li e- cycle p esen ed by Weske in [4, 5] as a e e - ence and we desc ibe how managemen o PPIs can be in eg a ed in o his li ecycle. Fu he - mo e, we also epo on se e al p oposals ha can be used o suppo his in eg a ion. 2.1 Design and Analysis In he design and analysis phase, PPIs should be modelled oge he wi h he business p o- cess. Fu he mo e, his model o PPIs should also enable hei analysis by de ec ing he de- pendencies amongs hem a design ime and also using hem as pa o he business p o- cess analysis, o ins ance in business p ocess simula ion echniques. Wi h espec o his phase, se e al au ho s ha e p esen ed some app oaches: Ped inaci e al. p esen in [9] a me ic on ology o al- low he de ini ion and compu a ion o me - ics. Popo a e al. [13] p esen a amewo k o modeling pe o mance indica o s wi hin a gen- e al o ganisa ion modeling amewo k. They de ine indica o s by assigning alues o a se o a ibu es. In [23], hey poin ou he way hese indica o s a e calcula ed and de ine e- la ions be ween PPIs and he p ocesses. Rela- ionships be ween PPIs (causali y, co ela ion and agg ega ion) a e b ie ly in oduced in [13] and explained in de ail in [23]. They also de ine empo al p ope ies o e PPIs (called PI exp essions) in [24]. We zs ein e al. de- sc ibe in [20] a KPI on ology using WSML o speci y KPIs o e seman ic business p ocesses. In [15] Maye l e al. discuss how o de i e me - ic dependency de ini ions om unc ional de- pendencies by applying dependency pa e ns. Cas ellanos e al. p esen in [18] he IBOM pla o m, ha allows, among o he hings, o de ine business measu es. The use can de ine business measu es ( h ough a GUI) o measu e cha ac e is ics o p ocess ins ances, p ocesses, esou ces o o he o e all business ope a ions. Speci ically, hey cha ac e ize me ics h ough ou a ibu es: name (unique), a ge en i y (obje o be measu ed), da a ype (nume ic, boolean, axonomy o SLA) and desi able me - Ac as de los Talle es de las Jo nadas de Ingenie ía del So wa e y Bases de Da os, Vol. 4, No. 4, 2010 ISSN 1988-3455 SISTEDES, 2010 9 ic alues. This app oach is no ocused on business p ocesses bu on he whole o ganisa- ion. 2.2 Con igu a ion Du ing he con igu a ion phase, he ins u- men a ion o he p ocesses ha a e necessa y o ake he measu es mus be de ined. Momm e al. p esen in [19] a me a-model o he speci ica ion o he PPI moni o ing along wi h an ex ension o he BPMN me a- model o modeling he equi ed ins umen- a ion o he moni o ing and an ou line o me hodology o an au oma ed gene a ion o his ins umen a ion.This app oach explains how o de ine a KPI and how o ake he mea- su es. 2.3 Enac men Du ing he business p ocess enac men , when aluable execu ion da a is ga he ed, he PPIs’ alues ha e o be calcula ed and he moni o - ing o hese PPIs should be ca ied ou . Fo ins ance, his can be done based on execu ion logs ha s o e in o ma ion abou he p ocess such as he s a o he end o ac i i ies. To o e come his issue, Ped inaci e al. p esen in [9] a Seman ic Business P ocess Moni o ing Tool called SENTINEL. This ool can suppo au oma ed easoning. Popo a e al. p opose in [23] o mal echniques o anal- ysis o execu ions o o ganiza ional scena ios. We zs ein e al. in oduce in [20] a amewo k o BAM as pa o he seman ic business p o- cess managemen , whe e PPI models a e au- oma ically con e ed o IT-le el e en -based models and used o eal- ime moni o ing us- ing easoning echnology. In his amewo k, he e is a moni o model ha speci ies how e en s a e p ocessed o calcula e KPIs, whose alues a e hen published in dashboa ds 2.4 E alua ion Finally, du ing he e alua ion phase, he mon- i o ing in o ma ion ob ained in he p e ious phase will help o iden i y co ela ions and p e- dic u u e beha iou . Cas ellanos e al. [18] add ess his issue, since hei IBOM pla o m also allows o pe - o m in elligen analysis on business measu es o unde s and causes o undesi ed alues and p edic u u e alues. 3 Challenges As can be deduced om Sec ion 2, business p ocess li ecycle can bene i om he de ini ion o PPIs o e p ocesses and om a se o anal- ysis ope a ions o hem, enabling a mo e effec- i e and efficien au oma ed suppo o ex ac in o ma ion om such indica o s. In o de o ob ain such bene i , he e exis s a numbe o esea ch ques ions ha a e no answe ed ye . In he ollowing we s a e some o hem: •Which in o ma ion is necessa y in o de o de ine P ocess Pe o mance Indica o s o e business p ocesses? How do hese PPIs ela e o he business p ocesses o which hey we e de ined? An app op ia e de ini ion o PPIs is key o enable he au oma ed suppo o he a o emen ioned PPIs li ecycle. Mo eo e , in o de o be able o ge in o ma ion, no only om he indica o s, bu also om he business p ocesses hey we e de- ined o , i is impo an o de ine an ex- plici connec ion be ween indica o s and he p ocess elemen s hey a e measu ing. The e a e se e al esea ch p oposals o de ine PPIs, as p esen ed in Sec ion 2, bu none o hem a e well-sui ed because ei- he he e a e commonly used PPIs ha canno be de ined wi h hem o hey a e no eady o enable a design- ime analysis o PPIs. In Ped ina i e al.’s app oach [9] i is no clea whe he he de ined me - ics a e explici ly connec ed wi h he el- emen s o a business p ocess. Popo a e al. [13] do no conside de i ed measu es when de ining indica o s. We zs ein e al. [20] do no ake in o accoun indica- o s ela ed o da a. Maye l e al. [15] do no del e in o he de ini ion o measu es, Ac as de los Talle es de las Jo nadas de Ingenie ía del So wa e y Bases de Da os, Vol. 4, No. 4, 2010 ISSN 1988-3455 SISTEDES, 2010 10 hey only se he seman ics o some ele- men s o conside when de ining measu es based on concep s o he CIM me ics model [16] (e.g. Uni O Wo k, which e- la es me ics wi h unc ional en i ies) and he QoS UML p o ile desc ibed in [17]. Momm e al.’s me amodel o he speci i- ca ion o PPIs [19] lacks some p ope ies iden i ied like, o ins ance, he analysis pe iod o he PPI). Finally, Cas ellanos e al. [18], as a as we can deduce om he pape , do no ake in o accoun du ing he de ini ion o me ics some aspec s such as he analysis pe iod, he uni o measu e o he dimension o be measu ed. •Which analysis ope a ions can be applied o PPIs in o de o ge ele an in o ma ion abou hem? Which in o ma ion can be ga he ed om hese PPIs’ de ini ion? Making que ies and ex ac ing in o ma- ion om PPIs may be use ul when de in- ing dashboa ds o moni o p ocesses. Fu - he mo e, he analysis o hese PPIs a design- ime can help o assu e he de i- ni ion o achie able goals o e en iden- i y con lic ing goals. E.g. I could be de ec ed i wo PPIs in e sely dependen a e being ied o be op imised. Ano he bene i esul ing om his analysis is he possibili y o ind ou how ce ain ac i i y in luences PPIs o ice e sa. Wi h espec o his challenge, he e ex- is some app oaches [9, 23, 20, 18] ha add ess he analysis o PPIs, bu hey do i in un ime, losing he a o emen ioned bene i s o his design- ime analysis. •Which a e he issues o conside when de eloping a sui able ool o suppo his au oma ic analysis o p ocess pe o mance indica o s? Explici business p ocess models ex- p essed in a g aphical no a ion (such as, o ins ance, BPMN) ease communica ion abou hese p ocesses, so ha s akehold- e s can communica e efficien ly, and e- ine and imp o e hem. In he same way, i is con enien o de elop a g aphical no- a ion o depic hese PPIs oge he wi h he co esponding business p ocess mod- els. Thus, i is necessa y o de elop a ool o suppo such a g aphical no a ion o PPIs, as well as o enable hei au o- ma ed analysis. Fu he mo e, i would be e y use ul o in eg a e such a ool wi h an edi o o business p ocess diag ams. To he bes o ou knowledge, he e no exis s such a g aphical no a ion no he desc ibed ool. 4 Conclusions and Fu u e Wo k In his pape , we a gue he impo ance o in- eg a ing he managemen o PPIs in o he whole business p ocess li ecycle. Speci ically, in he design and analysis phase, PPIs mus be modelled oge he wi h he business p ocess. This model should enable (a design- ime) an au oma ed o semi-au oma ed analysis o, o ins ance, de ec he dependencies and po en- ial con lic s amongs hem o o use hem oge he wi h o he business p ocess analysis echniques such as simula ion echniques. We ha e analysed how diffe en app oaches ad- d ess his issue and hei sho comings. We ha e also iden i ied he challenges ha need o be aced o his espec . Ou u u e wo k ocuses on inding answe s o he open esea ch ques ions p esen ed in Sec ion 3. In o de o do so we ha e al eady s a ed by de ining an on ology o he de ini- ion o PPIs (desc ibed in a pape submi ed o he 18 h In e na ional Con e ence on Co- ope a i e In o ma ion Sys ems-CoopIS 2010), whose main bene i s can be summa ised as ol- lows: 1. The ela ion be ween PPIs and he busi- ness p ocess is explici ly es ablished. 2. I suppo s he de ini ion o a wide a- ie y o PPIs, including hose associa ed wi h da a objec s. I also suppo s he de ini ion o an exp essi e analysis pe iod o a PPI Ac as de los Talle es de las Jo nadas de Ingenie ía del So wa e y Bases de Da os, Vol. 4, No. 4, 2010 ISSN 1988-3455 SISTEDES, 2010 11 3. Dependencies be ween measu es can be au oma ically ob ained om he on ology, which enables he analysis o PPIs a de- sign ime. Fu he mo e, since he on ol- ogy has been de ined in OWL DL, au- oma ed easone s can be used o make que ies abou he PPI model Mo eo e , we a e ex ending his on ological de ini ion wi h a ex ual language as close o he na u al language as possible, in o de o ease his de ini ion. Wi h espec o he second poin , we plan o de ine a mechanism o make que ies on PPIs (PPI-Q) simila o he one p esen ed in [21] o make que ies on BPMN (BPMN-Q). We a e also wo king on desc ibing mechanisms o de- i e es ic ions ( aking ad an age om Con- s ain Sa is ac ion P oblems) be ween PPIs, in o de o ex ac knowledge in an au oma ic way om he PPIs de ined o e a pa icula business p ocess. Finally, o o e come he hi d challenge, we a e de eloping a g aphical no a ion in o de o depic hese on ological concep s o PPIs o e business p ocesses and we a e also in e- g a ing his no a ion in o he web-based edi o ORYX [22] as a esul o a collabo a ion s ay wi h he BPT g oup a he HPI Po sdam. Re e ences [1] ISACA: COBIT 4.1 (2009) [2] O ice o Go e nmen Comme ce: In o - ma ion echnology in as uc u e lib a y (ITIL) 3. Collec ion o books (2007) [3] Eu opean Founda ion o Quali y Man- agemen : EFQM excellence model 2010 [4] Weske, M.: Business P ocess Manage- men : Concep s, Languages, A chi ec- u es. Sp inge (2007) [5] del Río-O ega, A., Resinas, M., Ruiz- Co és, A.: Towa ds modelling and ac- ing key pe o mance indica o s in busi- ness p ocesses. In: II Talle sob e P o- cesos de Negocio e Ingenie ía de Se icios (PNIS 2009). (2009) [6] OMG: Business p ocess modeling no a- ion (BPMN) 2.0 (2009) [7] W3C OWL Wo king G oup: OWL 2 Web On ology Language (2009) [8] B ockmans, S., Volz, R., Ebe ha , A., Löffle , P.: Visual modeling o OWL DL on ologies using UML. In: ISWC 2004. Volume 3298., Sp inge (2004) 198–213 [9] Ped inaci, C., Lambe , D., We zs ein, B., an Lessen, T., Ceko , L., Dim- i o , M.: Sen inel: a seman ic business p ocess moni o ing ool. In:P oceedings o he Fi s In e na ional Wo kshop on On ology-suppo ed Business In elligence (OBI). (2008) [10] Ab amowicz, W., Filipowska, A., Kacz- ma ek, M., Ped inaci, C., S a zecka, M., A., W.: O ganiza ion s uc u e de- sc ip ion o he needs o seman ic busi- ness p ocess managemen . In: 3 d in- e na ional Wo kshop on Seman ic Busi- ness P ocess Managemen coloca ed wi h 5 h Eu opean Seman ic Web Con e ence. (2008) [11] Ab amowicz, W., Filipowska, A., Kacz- ma ek, M., Kaczma ek, T.: Seman ically enhanced business p ocess modelling no- a ion. In: P oceedings o SBPM 2007 Seman ic P ocess and P oduc Li ecy- cle Managemen in conjunc ion wi h he 3 d Eu opean Seman ic Web Con e ence (ESWC 2007). (2007) [12] Ped inaci, C., Domingue, J., Al es de Medei os, A.K.: A co e on ology o busi- ness p ocess analysis. In: 5 h Eu opean Seman ic Web Con e ence. (2008) [13] Popo a, V., Sha panskykh, A.: Mod- eling o ganiza ional pe o mance indica- o s. In o ma ion Sys ems (2009) [14] Ga cia, F., Be oa,M.F., Cale o, C., Val- lecillo, A., Ruiz, F., pia ini, M., Gene o, M.: Towa ds a consis en e minology o so wa e measu emen In: In o ma ion & So wa e Technology (48). 2006 631-644 Ac as de los Talle es de las Jo nadas de Ingenie ía del So wa e y Bases de Da os, Vol. 4, No. 4, 2010 ISSN 1988-3455 SISTEDES, 2010 12 [15] Maye l, C., Hüne , K., Gaspa , J.U., Momm, C., Abeck, S.: De ini ion o me - ic dependencies o moni o ing he im- pac o quali y o se ices on quali y o p ocesses. In: Second IEEE/IFIP In e - na ional Wo kshop on Business-d i en IT Managemen (Munich). (2007) 1–10 [16] Dis ibu ed Managemen Task Fo ce: Common in o ma ion model (CIM)(2003) [17] OMG: UML TM p o ile o modeling quali y o se ice and aul ole ance cha - ac e is ics and mechanisms speci ica ion (2006) [18] Cas ellanos, M., Casa i, F., Shan, M.C., Dayal, U.: IBOM: A pla o m o in elligen business ope a ion manage- men . In: P oceedings. 21s In e - na ional Con e ence on Da a Enginee - ing, 2005., Hewle -Packa d Labo a o ies (2005) 1084– 1095 [19] Momm, C., Malec, R., Abeck, S.: To- wa ds a model-d i en de elopmen o moni o ed p ocesses. In: Wi scha sin- o ma ik (2). (2007) 319–336 [20] We zs ein, B., Ma, Z., Leymann, F.: To- wa ds Measu ing Key Pe o mance Indi- ca o s o Seman ic Business P ocesses. In: 11 h In e na ional Con e ence on Busi- ness In o ma ion Sys ems (BIS’08). Inns- b uck, Aus ia (2008) 227–238 [21] Awad, A., Decke , G., Weske, M.: Effi- cien compliance checking using bpmn-q and empo al logic. In: BPM. Volume 5240 o LNCS, Sp inge (2008) 326–341 [22] Decke , G., O e dick, H., Weske, M.: O yx - An Open Modeling Pla o m o he BPM Communi y. In: 6 h In e - na ional Con e ence on Business P ocess Managemen (BPM). Volume 5240 o LNCS. Sp inge (2008) 382–385 [23] Popo a V. and Sha panskykh, A. Fo mal analysis o execu ions o o ganiza ional scena ios based on p ocess-o ien ed spec- i ica ions In: Applied In elligence. (2009) [24] Popo a V. and Sha panskykh, A. Fo - mal Goal-based Modeling o O ganiza- ionss In: MSVVEIS. (2008) 19–28 Ac as de los Talle es de las Jo nadas de Ingenie ía del So wa e y Bases de Da os, Vol. 4, No. 4, 2010 ISSN 1988-3455 SISTEDES, 2010 13