collaborative building of an ontology of key performance ......collaborative building of an ontology...

55
Collaborative Building of an Ontology of Key Performance Indicators Claudia Diamantini, Laura Genga, Domenico Potena, Emanuele Storti DII, Universit ´ a Politecnica delle Marche, Ancona, Italy CoopIS 2014, OTM 2014, Amantea, Italy, Oct 27-31 This work is licensed under the Creative Commons Attribution-ShareAlike 4.0 License. Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 1 / 29

Upload: others

Post on 05-Aug-2021

10 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Collaborative Building of an Ontologyof Key Performance Indicators

Claudia Diamantini, Laura Genga,Domenico Potena, Emanuele Storti

DII, Universita Politecnica delle Marche, Ancona, Italy

CoopIS 2014, OTM 2014, Amantea, Italy, Oct 27-31

This work is licensed under the Creative Commons Attribution-ShareAlike 4.0 License.

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 1 / 29

Page 2: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Outline

1 Introduction and Motivation2 Related Work3 Goal and Approach4 Methodology

Ontological model for KPIsReasoning over KPI formulasOntology editor

5 Experimentation6 Conclusion

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 2 / 29

Page 3: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Introduction

Scenario: performance measurement in collaborative environments(e.g., virtual organizations)

Problems:KPIs

complex data with an aggregate/compound naturebusiness view vs. technical view of KPIslack of shared understanding of the meaning of KPIs

shared library of KPIscollaborative managementmaintaining consistency of the repository

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 3 / 29

Page 4: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Motivation

Example: compare performances of two enterprises of a virtualenterprise

Requirements for the repository:compositional semantics for KPIs should be made explicitcollaboratively built: keep the repository consistent

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 4 / 29

Page 5: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Motivation

Example: compare performances of two enterprises of a virtualenterprise

Requirements for the repository:compositional semantics for KPIs should be made explicitcollaboratively built: keep the repository consistent

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 4 / 29

Page 6: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Motivation

Example: compare performances of two enterprises of a virtualenterprise

Requirements for the repository:compositional semantics for KPIs should be made explicitcollaboratively built: keep the repository consistent

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 4 / 29

Page 7: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Motivation

Example: compare performances of two enterprises of a virtualenterprise

Requirements for the repository:compositional semantics for KPIs should be made explicitcollaboratively built: keep the repository consistent

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 4 / 29

Page 8: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Motivation

Example: compare performances of two enterprises of a virtualenterprise

Requirements for the repository:compositional semantics for KPIs should be made explicitcollaboratively built: keep the repository consistent

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 4 / 29

Page 9: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Motivation

Example: compare performances of two enterprises of a virtualenterprise

Requirements for the repository:compositional semantics for KPIs should be made explicitcollaboratively built: keep the repository consistent

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 4 / 29

Page 10: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Related work

The complex nature of PI is not fully captured in existing models:semantic representations of the multidimensional model [Priebe et al2003, Niemi et al 2007, Neumar et al 2012]

to simplify and automatize design and analysis(non-logical) modeling of indicator formulas

to support calculation [Pedrinaci et al 2009],[Horkoff et al 2012]for interoperability [Golfarelli et al 2012]

Few work dealing with logical models for indicators:inference of dependencies among indicators [Popova et al 2010], [DelRıo Ortega et al 2010]

ontological representation of formulas [Kehlenbeck et al 2009]

Collaborative framework for management of KPIs:KPIshare, a social knowledge base [Resinas et al 2014]

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 5 / 29

Page 11: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Goal

Collaborative management of a shared, formal and valid referencemodel for KPIs providing support to:

introduction of new valid KPIs in the repositorysafe update/deletion of a KPIbrowsing and searching of KPIsconsensus management & versioning

Requirements:global view of KPIs and their formulasmapping of enterprise KPIs to global definitionsfunctionalities to manipulate and reason about KPIs

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 6 / 29

Page 12: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Goal

Collaborative management of a shared, formal and valid referencemodel for KPIs providing support to:

introduction of new valid KPIs in the repositorysafe update/deletion of a KPIbrowsing and searching of KPIsconsensus management & versioning

Requirements:global view of KPIs and their formulasmapping of enterprise KPIs to global definitionsfunctionalities to manipulate and reason about KPIs

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 6 / 29

Page 13: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Approach

Ontological representation of KPIsdescriptive and compositional semantics for KPIsgoal: annotation of local data with global definitions

Reasoning and manipulation services for KPIsfacts to represent KPIspredicates to formalize mathematical axioms for formulamanipulation and check of consistency

Ontology editorflexible management of the ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 7 / 29

Page 14: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Approach

Ontological representation of KPIsdescriptive and compositional semantics for KPIsgoal: annotation of local data with global definitions

Reasoning and manipulation services for KPIsfacts to represent KPIspredicates to formalize mathematical axioms for formulamanipulation and check of consistency

Ontology editorflexible management of the ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 7 / 29

Page 15: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Approach

Ontological representation of KPIsdescriptive and compositional semantics for KPIsgoal: annotation of local data with global definitions

Reasoning and manipulation services for KPIsfacts to represent KPIspredicates to formalize mathematical axioms for formulamanipulation and check of consistency

Ontology editorflexible management of the ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 7 / 29

Page 16: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Approach

Ontological representation of KPIsdescriptive and compositional semantics for KPIsgoal: annotation of local data with global definitions

Reasoning and manipulation services for KPIsfacts to represent KPIspredicates to formalize mathematical axioms for formulamanipulation and check of consistency

Ontology editorflexible management of the ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 7 / 29

Page 17: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

KPI Ontology

Inspiring principles: VRM, SCOR, Six-Sigma, multidimensional models

Indicator ≡ ∀ hasDimension.Dimension u∀ hasFormula.Formula u (=1 hasFormula) u∀ hasUnitOfMeasure.UoM u (=1 hasUnitOfMeasure) u∀ hasBusObj.BusinessObject u (=1 hasBusObj) u∀ hasAggrFunction.AggrFun u (=1 hasAggrFunction)

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 8 / 29

Page 18: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

KPI Ontology

Dimension: coordinate along which an indicator can be computed(e.g., time, place and product for Total Costs)

Dimensions are organized in levels, e.g.:Time: Day�Week�Month�YearPlace: City�Region�Country

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 9 / 29

Page 19: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

KPI Ontology

TBoxVE v OrganizationDimension

Enterprise v OrganizationDimension u ∀partOf.VE u(=1 partOf)

Department v OrganizationDimension u ∀partOf.Enterpriseu(=1 partOf)

ProjectTeam v OrganizationDimension u ∀partOf.Departmentu(=1 partOf)

ABoxVirEnt1:VE, ACME:Enterprise, R&D:Department, Team A:ProjectTeam

partOf(Team A,R&D),partOf(R&D,ACME), partOf(ACME,VirtEnt1)

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 10 / 29

Page 20: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

KPI Ontology

Formula: mathematical expression stating how an indicator iscomputed(e.g., PersonnelTrainingCost=HourlyCost*PersonnelTrainingTime)

Definition (Well-formed formula)Given a set {f1, ..., fn} of symbol of indicators and a set {op1, ...,opm} ofalgebraic operators,

fi is a well-formed formula;

opj(f1, ..., fk ) is a well-formed formula.

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 11 / 29

Page 21: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

KPI Ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 12 / 29

Page 22: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

KPI Ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 12 / 29

Page 23: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

KPI Ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 12 / 29

Page 24: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

KPI Ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 12 / 29

Page 25: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Approach

Ontological representation of KPIsdescriptive and compositional semantics for KPIsgoal: annotation of local data with global definitions

Reasoning and manipulation services for KPIsfacts to represent KPIspredicates to formalize mathematical axioms for formulamanipulation and check of consistency

Ontology editorflexible management of the ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 13 / 29

Page 26: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning functionalities

Logic programming as unifying logic layer:representation of formulas as Prolog facts, e.g.:A = B + C → formula(A,B+C,’branch node’)

Prolog predicates for basic reasoning tasks:mathematical theory for formula manipulation (PRESS: PRologEquation Solving System)

basic math operationequation solving and formula rewriting

definition of specific predicates for KPI management

XSB as Prolog reasoner

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 14 / 29

Page 27: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning functionalities

Logic programming as unifying logic layer:representation of formulas as Prolog facts, e.g.:A = B + C → formula(A,B+C,’branch node’)

Prolog predicates for basic reasoning tasks:mathematical theory for formula manipulation (PRESS: PRologEquation Solving System)

basic math operationequation solving and formula rewriting

definition of specific predicates for KPI management

XSB as Prolog reasoner

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 14 / 29

Page 28: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning functionalities

Logic programming as unifying logic layer:representation of formulas as Prolog facts, e.g.:A = B + C → formula(A,B+C,’branch node’)

Prolog predicates for basic reasoning tasks:mathematical theory for formula manipulation (PRESS: PRologEquation Solving System)

basic math operationequation solving and formula rewriting

definition of specific predicates for KPI management

XSB as Prolog reasoner

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 14 / 29

Page 29: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning functionalities

Logic programming as unifying logic layer:representation of formulas as Prolog facts, e.g.:A = B + C → formula(A,B+C,’branch node’)

Prolog predicates for basic reasoning tasks:mathematical theory for formula manipulation (PRESS: PRologEquation Solving System)

basic math operationequation solving and formula rewriting

definition of specific predicates for KPI management

XSB as Prolog reasoner

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 14 / 29

Page 30: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: mathematical predicates

Mathematical functions for rewriting:

solve equation for resolution of equations:Costs = TravelCosts+PersonnelCostsPersonnelCosts= Costs − TravelCosts

simplify solution (simplification and substitution):

ROI IG = ExpectedMarketImpactTravelCosts+PersonnelCosts

ROI IG = ExpectedMarketImpactCosts

Note: useful to infer new formulas not explicitly written in the ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 15 / 29

Page 31: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: mathematical predicates

Mathematical functions for rewriting:

solve equation for resolution of equations:Costs = TravelCosts+PersonnelCostsPersonnelCosts= Costs − TravelCosts

simplify solution (simplification and substitution):

ROI IG = ExpectedMarketImpactTravelCosts+PersonnelCosts

ROI IG = ExpectedMarketImpactCosts

Note: useful to infer new formulas not explicitly written in the ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 15 / 29

Page 32: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Case study

KPIs chosen for monitoring of a Virtual Enterprise:

InvestmentInEmplDevelopment = PersonnelTrainingCosts + TeachCost

PersonnelTrainingCosts = HourlyCost ∗ PersonnelTrainingTime

TeachCosts = PersonnelTrainingTime ∗ HourRate

PersonnelCosts = NumHours ∗ HourlyCost ∗ (Overhead + 1)

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 16 / 29

Page 33: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Case study

KPIs chosen for monitoring of a Virtual Enterprise:

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 16 / 29

Page 34: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: consistency check

Let KB be a repository of KPIs, a new indicator with formula fnew isconsistent with KB iif:

@fi ∈ KB/fi = fnew(no formula is identical to fnew )

KB 1 fnew(no formula is mathematically equivalent to fnew )

KB ∪ {fnew} 1 ⊥(the formula is coherent with the others in the KB)

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 17 / 29

Page 35: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: consistency check

Let KB be a repository of KPIs, a new indicator with formula fnew isconsistent with KB iif:

@fi ∈ KB/fi = fnew(no formula is identical to fnew )→ identical

KB 1 fnew(no formula is mathematically equivalent to fnew )

KB ∪ {fnew} 1 ⊥(the formula is coherent with the others in the KB)

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 17 / 29

Page 36: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: consistency check

Let KB be a repository of KPIs, a new indicator with formula fnew isconsistent with KB iif:

@fi ∈ KB/fi = fnew(no formula is identical to fnew )→ identical

KB 1 fnew(no formula is mathematically equivalent to fnew )→ equivalence

KB ∪ {fnew} 1 ⊥(the formula is coherent with the others in the KB)

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 17 / 29

Page 37: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: consistency check

Let KB be a repository of KPIs, a new indicator with formula fnew isconsistent with KB iif:

@fi ∈ KB/fi = fnew(no formula is identical to fnew )→ identical

KB 1 fnew(no formula is mathematically equivalent to fnew )→ equivalence

KB ∪ {fnew} 1 ⊥(the formula is coherent with the others in the KB)→ incoherence

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 17 / 29

Page 38: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: consistency check

Evaluation of equivalent formulas:

equivalence(F): given a formula G of the ontology, a formula F isequivalent to G if can be rewritten as G (by rewriting functionalities)

Policies:remove the duplicatekeep duplicates and declare that the corresponding KPIs are thesame indicator

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 18 / 29

Page 39: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: consistency check

Scenario: insertion of a new indicatorTotCostEmpTrain = TeachCosts+PersonnelTrainingTime ∗HourlyCost

Result: equivalent to InvestmentInEmployeeDevelopment

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 19 / 29

Page 40: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: consistency check

Scenario: insertion of a new indicatorTotCostEmpTrain = TeachCosts+PersonnelTrainingTime ∗HourlyCost

Result: equivalent to InvestmentInEmployeeDevelopment

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 19 / 29

Page 41: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: consistency check

Predicate to maintain a consistent ontology:

incoherence(F): given a new formula F = expression

for each formula G in KB containing a reference to F

solve G for F, obtaining F = expression2if {indicators in expression} ⊆ {indicators in expression2} then F isinconsistent

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 20 / 29

Page 42: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: consistency check

Scenario: update of an indicatorPersonnelTrainingTime = PersonnelTrainingCosts + HourlyCost

Result: incoherent with PersonnelTrainingCost

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 21 / 29

Page 43: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Reasoning: consistency check

Scenario: update of an indicatorPersonnelTrainingTime = PersonnelTrainingCosts + HourlyCost

Result: incoherent with PersonnelTrainingCost

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 21 / 29

Page 44: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Approach

Ontological representation of KPIsdescriptive and compositional semantics for KPIsgoal: annotation of local data with global definitions

Reasoning and manipulation services for KPIsfacts to represent KPIspredicates to formalize mathematical axioms for formulamanipulation and check of consistency

Ontology editorflexible management of the ontology

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 22 / 29

Page 45: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Ontology editor

Creation of a new indicator

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 23 / 29

Page 46: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Ontology editor

Search/Browse indicators

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 24 / 29

Page 47: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Ontology editor

Creation of a new member

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 25 / 29

Page 48: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Experimentation

BIVEE ontology:356 production and innovation KPIs (281 connected, 75disconnected)lattices: 3.14 levels and 2.67 operands per indicator

Experimental procedure (for identical, equivalence, incoherence):

for each connected indicator

compute execution time (x10 times)average results

average on all indicators

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 26 / 29

Page 49: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Experimentation

BIVEE ontology:356 production and innovation KPIs (281 connected, 75disconnected)lattices: 3.14 levels and 2.67 operands per indicator

Experimental procedure (for identical, equivalence, incoherence):

for each connected indicator

compute execution time (x10 times)average results

average on all indicators

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 26 / 29

Page 50: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Experimentation: results

Average execution time[ms]identical 4 (±1.1)

equivalence 197 (±11.5)incoherence 201 (±10.5)

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 27 / 29

Page 51: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Experimentation: results

Experiments with a set of synthetic ontologies with increasing sizeoperands:2..4, levels:2..5 = 12 ontologies73980 unique tests

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 27 / 29

Page 52: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Conclusion

Collaborative building of a ontology of Key Performance Indicatorslogical model for indicators (KPIOnto)reasoning services (mathematical model + set of predicates)

Current status of the workmanagement of KPIOnto in BIVEE projectonly linear equations (not a limitation)

Future workreason with both descriptive properties and formulasrecognition of similar (not equivalent) formulas

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 28 / 29

Page 53: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Conclusion

Collaborative building of a ontology of Key Performance Indicatorslogical model for indicators (KPIOnto)reasoning services (mathematical model + set of predicates)

Current status of the workmanagement of KPIOnto in BIVEE projectonly linear equations (not a limitation)

Future workreason with both descriptive properties and formulasrecognition of similar (not equivalent) formulas

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 28 / 29

Page 54: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Conclusion

Collaborative building of a ontology of Key Performance Indicatorslogical model for indicators (KPIOnto)reasoning services (mathematical model + set of predicates)

Current status of the workmanagement of KPIOnto in BIVEE projectonly linear equations (not a limitation)

Future workreason with both descriptive properties and formulasrecognition of similar (not equivalent) formulas

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 28 / 29

Page 55: Collaborative Building of an Ontology of Key Performance ......Collaborative Building of an Ontology of Key Performance Indicators. Collaborative Building of an Ontology of Key Performance

Acknowledgements

This work has been partly funded by the EC through the ICT ProjectBIVEE: Business Innovation and Virtual Enterprise Environment(FoF-ICT-2011.7.3-285746)

Emanuele Storti (UNIVPM) Building of an Ontology of KPIs CoopIS 2014 29 / 29