Enterprise ontology theory and methodology

Lý thuyết về ontology

Enterprise Ontology

Jan L.G Dietz

Enterprise Ontology

Theory and Methodology

With 99 Figures and 9 Tables

123

Jan L.G Dietz

Department of Software Technology

Faculty of Electrical Engineering, Mathematics,

and Computer Science

Delft University of Technology

Mekelweg 4

2628 CD Delft, The Netherlands

J.L.G.Dietz@tudelft.nl

Library of Congress Control Number: 2006920521

ACM Computing Classification (1998): H.1, H.2, I.6, J.1

ISBN-10 3-540-29169-5 Springer Berlin Heidelberg New York

ISBN-13 978-3-540-29169-5 Springer Berlin Heidelberg New York

This work is subject to copyright All rights are reserved, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilm or in any other way, and storage in data banks Duplication of this publication

or parts thereof is permitted only under the provisions of the German Copyright Law of September 9,

1965, in its current version, and permission for use must always be obtained from Springer Violations are liable for prosecution under the German Copyright Law.

Springer is a part of Springer Science+Business Media

Typeset by the author using a Springer TEX macro package

Production: LE-TEX Jelonek, Schmidt & Vöckler GbR, Leipzig

Cover design: KünkelLopka Werbeagentur, Heidelberg

Printed on acid-free paper 45/3100/YL - 5 4 3 2 1 0

To my CIAO! friends

There is nothing more practical than a good theory

When you study the course of affairs in an arbitrary enterprise as a nạveobserver, you notice that the persons in the enterprise seem busy like ants

in executing a lot of different activities, thereby using a variety of chines and equipment These persons appear to produce and shuffle around

ma-a lot of reports ma-and other documents, while ma-at the sma-ame time communicma-at-ing quite a lot too, either face-to-face or by telephone or by some othermeans Your first impression therefore invariably is one of huge diversityand complexity, and of an apparent lack of structure and logic Moreover,

communicat-if you revisit the enterprise after some time, it could very well be the casethat you hardly recognize it, while from the outside it is still the same, i.e.,the enterprise is still in the same business You may see new persons andyou may see known persons doing other things, you may notice that othermachines and/or equipment are being used, and that the flows of docu-ments and forms have changed The only constant factors seem to be thehuge diversity and complexity, and the difficulty in seeing a structure orlogic in it You may wonder how this enterprise managed to change Whatwas the redesign plan, and what was the reengineering plan, and whywould they be right? How did one proceed to bring about the intendedchanges? How were they justified and evaluated?

Everyone who directed or carried out a substantial organizationalchange would call these questions highly relevant, and, probably, alsorather academic He or she knows that substantial changes get accom-plished generally only with a lot of cheer luck, meaning that they rely onunprecedented, sometimes magical achievements of the people who actu-ally do the work This current practice of many unsuccessful projects, oflucky chances instead of controlled execution of well-designed plans, con-stitutes the motivation for the work that is described in this book, work thathas been both theoretical and practical, and certainly not academic in thecolloquial sense of the word

VIII Prologue

If one thing catches the eye in almost all literature about and proaches to the (re)designing and (re)engineering of enterprises, includingsuch subareas as business process design and engineering, and informationsystems design and engineering, it is the lack of a well-founded theoryabout the operation of enterprises Often even the most basic notions, likeaction and actor and process, are not clearly and precisely defined Obvi-ously, one cannot expect much real help of such approaches

ap-The growing interest in the practical application of the notion of ogy provides a chance to make a fresh start, to bring approaches to the(re)designing and (re)engineering of enterprises to a higher level of qual-ity In order to achieve this, one has to find a way to separate the stable on-tological essence of an enterprise from the variable way in which it is real-ized and implemented That is the only hope for mastering the diversityand complexity in contemporary enterprises What we need is a theory inwhich the concepts of essence and realization and implementation and allother relevant concepts have a well-defined place Such a theory and amatching methodology, which has proven to be effective in numerouspractical projects, constitute the contents of this book

ontol-I am fully aware that ontol-I have profited from the feedback ontol-I received in thepast fifteen years from a large number of people, which continually caused

me to reconsider my thoughts1 Instead of listing the names I have in mind,which would do injustice to the ones I forget, I confine myself to mention-ing that they belong, in a stimulating mixture, to three groups The firstconsists of the MSc and PhD students, formerly at the University of Maas-tricht, and currently at Delft University of Technology The second is con-stituted by the colleagues and assistants, in both universities, I have col-laborated with or still collaborate with The third group consists ofpractitioners, in all lines of business I am particularly happy to notice thatnearly all current PhD students belong also to this group They know fromtheir own experience the problems in enterprises, and they have discoveredthat there is only one way out: improving your way of thinking

I could have waited for another couple of years before having this bookpublished, while continually adding and improving things Instead I de-cided to do it now, for several reasons, of which the most important one isthat I wanted to finish something In this volume, I focus on describing thetheory and the methodology of enterprise ontology, leaving untouchedprobably the most interesting part for practitioners: the overwhelmingamount of practical experience in applying the methodology, which has re-cently been surveyed [24] I can only promise that this will be taken care

1 I refer here to a quote that is attributed to Richard Feynmann: “If you think you understand something, you have not thought about it enough”.

Prologue IX

of; there will appear a second volume containing a selection of the mous variety of practical applications For now, I wish you a joyful learn-ing experience in studying this first volume on enterprise ontology

enor-Delft, January 2006,

Jan L.G Dietz

Part A: Introduction

1 Outline of the Book 3

2 What is Enterprise Ontology? 7

3 An Explanatory Case 15

3.1 The Analysis of the Case Volley 16

3.2 The Ontological Model of the Case Volley 24

Part B: Foundations 4 Factual Knowledge 35

4.1 The Ontological Parallelogram 35

4.2 The Ontology of a World 41

5 A World Ontology Specification Language 45

5.1 The Declaration of Statum Types 46

5.2 The Specification of Existence Laws 49

5.3 The Derivation of Statum Types 51

5.4 Factum Types and Occurrence Laws 55

6 The Notion of System 57

6.1 The Distinct System Notions 57

6.2 Formal Definition Ontological System 60

7 The Notion of Model 63

7.1 Definition of Model 63

7.2 The White-Box Model 65

7.3 The Black-Box Model 67

8 The Role of Ontology in Enterprise Engineering 71

8.1 Design and Engineering 71

8.2 The System Development Process 75

XII Contents

Part C: The Theory

9 The Operation Axiom 81

9.1 Coordination Acts 83

9.2 Production Acts 85

9.3 Actors 87

10 The Transaction Axiom 89

10.1 The Basic Transaction Pattern 90

10.2 The Standard Transaction Pattern 93

10.3 The Cancellation Patterns 95

11 The Composition Axiom 99

12 The Distinction Axiom 105

12.1 Communication 106

12.2 Coordination 109

12.3 Production 113

13 The Organization Theorem 115

13.1 The Realization of an Organization 117

13.2 The Implementation of an Organization 120

14 The CRISP Model 127

14.1 Transaction Time Aspects 127

14.2 Formal Definition of the CRISP Model 130

14.3 The Crispienet 133

Part D: The Methodology 15 The Modeling Method 139

15.1 The Distinct Aspect Models 139

15.2 The Perfoma-Informa-Forma Analysis 144

15.3 The Coordination-Actors-Production Analysis 149

15.4 The Transaction Pattern Synthesis 154

15.5 The Result Structure Analysis 157

15.6 The Construction Synthesis 158

15.7 The Organization Synthesis 158

16 The Interaction Model 159

16.1 The IAM of the Library 160

Contents XIII

16.2 The IAM of the Pizzeria 166

16.3 Practical Relevance of the Interaction Model 170

17 The Process Model 173

17.1 The PM of the Library 174

17.2 The PM of the Pizzeria 180

17.3 Practical Relevance of the Process Model 183

18 The Action Model 185

18.1 The AM of the Library 186

18.2 The AM of the Pizzeria 191

18.3 Practical Relevance of the Action Model 195

19 The State Model 197

19.1 The SM of the Library 200

19.2 The SM of the Pizzeria 203

19.3 Practical Relevance of the State Model 204

20 The Interstriction Model 205

20.1 The ISM of the Library 206

20.2 The ISM of the Pizzeria 209

20.3 Practical Relevance of the Interstriction Model 213

Epilogue 215

Appendix: The Example Cases 217

Bibliography 225

Glossary 229

Index 241

Part A: Introduction

1 Outline of the Book

Part A contains introductory material for the core of the book After theoutline, in Chap 1, the notion of enterprise ontology, in the context of on-tology in general and of the observed societal needs for enterprise ontol-ogy, is discussed in Chap 2 It turns out that the notion of enterprise ontol-ogy could be beneficial not only for managers and designers oforganizations and information systems, but also for the employees and theclients of an enterprise, be it a commercial company or a governmentagency In Chap 3 this notion is exemplified, taking the operational activi-ties of becoming a member of a tennis club as the leading example De-spite the small size of this enterprise, you will acquire a basic knowledge

of enterprise ontology that is appropriate and sufficient to study the mainder of the book

re-Foundation

Before we present the theory that underlies our notion of enterprise ogy in Part C, it is necessary to lay in Part B a sound foundation for it.First, in Chap 4, we investigate meticulously what it means to haveknowledge of some world, and in particular what ontological knowledge

ontol-is A number of basic notions are precisely defined and discussed, amongwhich are state, transition and event The investigation leads to the relevantdistinction between two kinds of facts, which we will call facta and stata.Chapter 5 contains the grammar of a first order logical language that isfully appropriate for describing the ontology of a world For a thoroughunderstanding of our notion of enterprise ontology, in fact, for understand-ing any system ontology, we need to discuss the two kinds of system no-tions that appear to be in use, particularly with respect to enterprises: thefunction- and behavior-oriented system notion and the construction- andoperation-oriented system notion As will become clear, they are not reallytwo distinct system notions but actually two distinct kinds of models ofonly one system notion The two kinds of models are the black-box modeland the white-box model A thorough presentation and discussion of thenotion of system is contained in Chap 6 The notion of model is presentedand discussed in Chap 7 Part B is completed by a study of the role of on-tology in the development of systems in Chap 8, in particular, of enter-

4 1 Outline of the Book

prises and their supporting information systems As in Chap 4, a number

of well-known notions are precisely defined, making use of the new notion

of ontology Also, the need for developing the new discipline of enterpriseengineering is discussed

It states that an enterprise is a layered nesting of three homogeneous aspectsystems: the B-organization (from Business), the I-organization (from In-tellect) and the D-organization (from Documents) In Chap 14 the so-called CRISPmeta model is developed It is the formalization and opera-

2 The Greek letter
is pronounced as PSI, which is taken as an acronym for Performance in Social Interaction, the basic paradigm of the theory.

1 Outline of the Book 5

tionalization of the
-theory, necessary for any methodology for enterpriseontology that has a sound theoretical foundation

Besides the foundations of the
-theory that are discussed in Part B, weneed to mention a scientific source that may be called the most influentialand important It is the Language-Action Perspective (LAP) LAP is thename of a scientific field as well as a scientific community that had itstenth annual conference in 2005 The root ofLAP is constituted of a num-ber of philosophical works, of which the most referenced ones are [4, 32,55] One could say that the exploration of the scientific field of LAPstarted with [26, 27, 66] Some later publications in the field of LAPthatare worth mentioning are [14, 15, 16, 50, 53, 61, 62, 64] All these publica-tions provide relevant background information for studying Part C of thisbook

Methodology

In Part D, a methodology is presented that has proven to be effective innumerous projects in practice First, in Chap 15, a method is presented forproducing an ontological model on the basis of a description of the currentoperations in an enterprise Students are advised to apply this methodologymeticulously; it will be valuable to hold on to Professionals do not like tofollow rigid step-by-step procedures and rightly so! Being a professionalmeans first of all knowing what to do, i.e., what results to achieve Thebest way to it is always contingent on the circumstances, one of them be-ing the expediency of the modeler

A complete enterprise ontology consists of four related aspect models:the construction model, the process model, the action model, and the state

model The construction model is very unlike usual organization models.

Usually, an organization model is understood to be a hierarchical structure

of organizational units, like divisions, business units, and departments,and/or organizational functions, like managers, salesmen, and accountants.However, in the
-theory, these units and functions belong to the realm ofimplementation The essential unit of authority and responsibility is the(elementary) actor role Moreover, the notion of organization is first of allconsidered to be about the interaction of the actor roles Therefore, an in-teraction model shows the ontological construction or composition of anenterprise: the elements (actor roles) and their mutual influences In Chap

16 only the first part of the construction model, called the interactionmodel, is presented

In Chap 17, the ontological process model is presented It shows how

the distinct transactions are interrelated There are two kinds of links tween transaction steps: causal links and conditional links A causal link

be-6 1 Outline of the Book

from a transaction T1 to a transaction T2 means that T2 is initiated fromwithin T1 A conditional link from a transaction T1 to a transaction T2means that T2 has to wait for the completion of T1 before it can proceed.Although actors act autonomously, they follow guidelines or procedures

in order to act responsibly These guidelines or procedures are called tion rules The action rules that pertain exclusively to the enterprise ontol-

ac-ogy are collectively called the ontological action model It is presented in

Chap 18 In order to specify an action model as precisely as possible, wewill make use of a pseudo-algorithmic language

In Chap 19, the ontological state model is presented It specifies the

state space (i.e., the set of allowable states) of both the production worldand the coordination world of the enterprise Stated differently, it containsthe conceptual model of all facts that are produced and all facts that areused State models are expressed inWOSL, the language we introduced inChap 5

In Chap 20, the second part of the construction model, called the striction model, is presented What it adds to the interaction model are allpassive mutual influences between actor roles They are called interstric-tions, as opposed interactions, which are the active influences Interstric-tions are represented as access links from an actor role to an informationbank, expressing that the actor role is allowed to know the contents of thebank For all five aspect models, their practical relevance is discussed atthe end of every chapter

inter-2 What is Enterprise Ontology?

Enterprise ontology is a novel subject, and writing a book on this novelsubject puts the author under the obligation to provide at least two kinds ofexplanation One explanation regards the justification of presenting yet an-other point of view on enterprises Why and how would enterprise ontol-ogy assist in coping with the current and future problems related to enter-prises? The other explanation concerns the particular approach towardsenterprise ontology that the author takes Why would this approach bemore appropriate and more effective than some other one? These are seri-ous questions indeed, and anyone who takes the pain to study this book de-serves satisfying answers You will get the answers; however, not straightaway A first attempt is in this introductory chapter Definite and fully sat-isfying answers can only emerge from a dedicated and thorough study ofthe book The lasting reward of such a study is a novel and powerful in-sight into the essence of the operation of enterprises; by this we mean in-sight that is fully independent of the (current) realization and implementa-tion

Let us start by noting that managing an enterprise, but also getting ices from it as a client or collaborating with it as partner in a network, isnowadays far more complicated than it was in the past Be assured, we willnot elaborate on it, you probably have heard that tune in all pitches andkeys And in case you have not, glance over an arbitrary management bookfrom the past five years and you are informed The problems in current en-terprises, of any kind, are well investigated and well documented Morethan well, in fact, because far less effort is put in thinking about how tocope with them Anyhow, the common denominator of these problems iscomplexity, and complexity can only be mastered if two conditions are ful-filled One condition is that one dispose of a comprehensive theory aboutthe kind of things whose complexity one wants to master The other condi-tion is that one dispose of appropriate analysis methods and techniques,based on that theory

serv-The knowledge that one acquires at management or business schoolsdoes not suffice anymore Actually, it never did; managers were just luckythat the shop floor workers ultimately always managed to really solveproblems and implement desired changes Even the gifted entrepreneur can

8 2 What is Enterprise Ontology?

nowadays not succeed without a basic, systematic, and integral standing of how enterprises work In order to really cope with the currentand the future challenges, a conceptual model of the enterprise is neededthat is coherent, comprehensive, consistent, and concise, and that only

under-shows the essence of the operation of an enterprise model By coherent we

mean that the distinguished aspect models constitute a logical and truly

in-tegral whole By comprehensive we mean that all relevant issues are ered, that the whole is complete By consistent we mean that the aspect models are free from contradictions or irregularities By concise we mean

cov-that no superfluous matters are contained in it, cov-that the whole is compactand succinct The most important property, however, is that this conceptual

model is essential, that it shows only the essence of the enterprise, its deep

structure In particular, we mean that the model abstracts from all tion and implementation issues

realiza-We shall call such a conceptual model an ontological model The

origi-nal Greek word from which the English word “ontology” stems, meansstudy or knowledge of what is or exists, and the philosophical branch withthe same name has taken up the term as referring to the reality around us,regardless our own view on it In other words, ontology requires us tomake a strict distinction between the observing subject and the observedobject This requirement puts the author under another obligation, that ofclarifying the philosophical stance taken with respect to this subject-objectdichotomy We will do it only briefly, without much elaboration The in-terested reader is referred to [25] and [56] There are three philosophicalpositions that are relevant for our discussion: the objectivist, the subjectiv-

ist and the constructivist position Objectivists believe that the world they

live in exists in itself, fully independent of them In other words, they

be-lieve in a true objective reality Subjectivists take the opposite position.

They believe that there is no reality outside the subject (human being) and,

in the extreme, that every subject has its own image of reality Somewhere

in between is the position of the constructivists They agree with the

sub-jectivists that there is no absolute objective reality (as the obsub-jectivists lieve), but they believe that there is instead a kind of semiobjective realitythat they call an intersubjective reality It is built and continuously adaptedthrough negotiating and achieving social consensus among subjects Ourposition is this constructivist one We consider the ontology of a particularpart of reality as the basis for sensible communication about that part ofreality At the same time, we recognize that this ontology is built, rebuilt,and adapted in communication; it cannot be otherwise

be-We like to add to this tripartite philosophical stance two sociologicalparadigms regarding the study of systems, namely the functionalist para-

2 What is Enterprise Ontology? 9

digm and the interpretive paradigm [37] The functionalist paradigm takes

its name from the fact that it wants to ensure that everything in the system

is operating well so as to promote efficiency, adaptation, and survival Anunderstanding can be gained of how systems work by using scientificmethods and techniques to probe the nature of parts of the system, the in-terrelationships between them, and the relationship between the systemand its environment The expertise it provides should put managers more

in control of their operations and organizations, and enable them to

elimi-nate inefficiency and disorder The interpretive paradigm takes its name

from the fact that it believes social systems, such as organizations, resultfrom the purposes people have and that these, in turn, stem from the inter-pretations they make of the situations in which they find themselves Peo-ple act and interact in organizations as a result of their interpretations Thisparadigm wants to understand the different meanings to collaborative ac-tivity and to discover where these meanings overlap, and so give birth toshared, purposeful activity Managers can be guided to seek an appropriatelevel of corporate culture in their organizations They can take decisions,

on the basis of participative involvement, that gain the commitment ofstakeholders Some argue that these paradigms are incommensurable Inour opinion, this is not necessarily the case The notion of enterprise ontol-ogy, as conveyed in this book, is primarily functionalist in nature How-ever, various aspects (e.g., considering an enterprise as a social entity, thefocus on social individuals, Habermas’ theory of communicative action,the autonomy that is basically allowed to actor roles) also reflect an inter-pretive perspective One might argue that a really comprehensive approach

to enterprise engineering should be able to address an enterprise from ferent angles, thus integrating important views from different paradigms.This is what we try to do, as will become clear in Part C Let this be our fi-nal brushstroke in painting the philosophical background for the key no-tion of enterprise ontology

dif-In its modern use, ontology has preserved its original meaning, but it hasalso a definite practical goal It serves to provide a basis for the commonunderstanding of some area of interest among a community of people whomay not know each other at all, and who may have very different culturalbackgrounds If you have ever heard about ontology before, it is mostprobable that it was in the context of the World-Wide Web, particularly inthe context of the Semantic Web [6] There are various definitions of themodern notion of ontology getting around Our main source is the ontol-ogy of Mario Bunge [10, 11], but, as long as there is no conflict, we willalso refer to other sources A widely adopted definition of ontology is theone in [29]: an ontology is a formal, explicit specification of a shared con-

10 2 What is Enterprise Ontology?

ceptualization It states the core properties that our notion of ontology alsowill have First, it regards the conceptualization of (a part of) the world, so

it is something in our mind Because of our constructivist stance, we sider these mental pictures be checked and adapted in communication.Second, this conceptualization is supposed to be shared, which is the prac-tical goal of ontologies This takes also place in communication Third, it

con-is explicit; an ontology must be explicit and clear, there should be no roomfor misunderstandings Fourth, it is specified in a formal way Natural lan-guage is inappropriate for this task, because of its inherent ambiguity andimpreciseness

The notion of ontology as applied in [29], but also in [28], [30], and

[58], is what we will call in Chap 4 a world ontology Common examples

of such an ontology are the world of traveling or the world of cooking anddining The focus is on defining the core elements in such a world andtheir interrelationships in a most clear and extensive way The notion of

ontology as applied in this book is the notion of system ontology Our goal

is to understand the essence of the construction and operation of completesystems; more specifically, of enterprises As will become clear, this no-tion of system ontology includes the notion of world ontology Next, al-though we fully recognize the need for ontologies for the purpose ofworldwide flawless communication among agents over the Internet, ourmotivation for this book is wider In our opinion, the world we live in is,and will remain, in the first place, a world of people, of human beings, de-spite all the technical devices that (can) make our lives much more pleas-ant This is a philosophical stance of course; it is a choice We stronglyoppose, for example, the quite common idea that artificial agents are, or atsome future time will be, equivalent fellow players in human social life.This idea can only be justified by a severe inflation of such notions asauthority and responsibility Throughout the history of mankind, peoplehave used anthropomorphic metaphors for the purpose of understandingand explaining the operation of natural things as well as artifacts Theonly, but at the same time serious, danger is that one forgets that they weremetaphors, that one takes the metaphorical reasoning for real So, for ex-ample, if you think that your computer does not understand you, you aretwofold right First, it is quite okay to use anthropomorphic metaphors inthe interaction with your computer You probably do it sometimes alsowhile driving your car or trying to let your video recorder do what youwant it to do Second, in the most true sense, your computer does not un-derstand you, because understanding in the way human beings have inter-nalized the notion and apply it is applicable only to them We do not be-lieve in a general notion of understanding that human beings would share

2 What is Enterprise Ontology? 11

with artificial intelligent systems There is no evidence for such a belief,except for apparent similar external behavior in some cases To concludefrom these cases that these behaviors are brought about in the same way ismerely speculation

A major motivation for this book and for our work in ontology in eral stems from the conviction that the world is in great need for transpar-ency about the operation of all the systems we daily work with, rangingfrom the domestic appliances to the big societal institutions We are ingreat need already, and this need can only increase if one imagines a futurelife in a cyber culture [5] Our concern is the current lack of an appropriatephilosophical counterbalance to the dominant technocratic and bureau-cratic thinking Let us give some examples to clarify the point First, re-garding technical devices, if you read the user manual of a video recorder

gen-or a computer gen-or a computer program, you become overloaded with vant details You mostly end up with a headache instead of any relevantunderstanding And in case you persevere, there is a high chance that youwill discover so many errors and omissions in the description that readinghas become solving a puzzle As a concrete example, the implementation

irrele-of an ERP(Enterprise Resource Planning) package in an enterprise, even

of only a few modules, may easily take several years and cost the prise a huge amount of money This money is partly spent in having thesupplier of the package (or some intermediary company) explain how touse it, and partly to have the enterprise adapt its current way of workingsuch that it fits the straitjacket of the ERP package Is this societal pro-gress? Do we really need to suffer this? As another example, have you everphoned the help desk of a company or a government agency in order to getthe service they claim you will get in their advertisements? Mostly you end

enter-up not by having what you were looking for, but by being frustrated,maybe to the extent that you think of giving up Why? Because the opera-tion of these institutions is completely opaque to you You do not knowwhat to believe and what not to believe; you are literally lost in cyber-space And, in case you have succeeded in penetrating to the right place,there is a chance that the responsible person does not take on his or her re-sponsibility and concludes your case by blaming the computer or any otherthing that he or she uses as an aid Most probably, he or she acts in thisway not to hamper or frustrate you, but because the institution is alsoopaque to him or her

This situation should stop because it is in no one’s interest that it tinue, as it has been in no one’s interest to have come this far To the best

con-of our knowledge, there has never anywhere been a plan to organize ern society in such a way that nobody is able to understand how it works

mod-12 2 What is Enterprise Ontology?

Likewise, in no manufacturing company has there ever been a plan to sign domestic appliances or professional equipment such that it takes thecurrent incommensurable amount of effort to get to know how to use them.Things have just gone that way But, as was said, there is no reason to let itcontinue Instead, there is abundant ground for stopping it Imagine that it

de-is possible for you to acquire the right amount of the right kind of edge about the operation of the equipment you are working with Imaginethat you are not bothered by incomprehensible and irrelevant things butthat you get the insight you need in a way that you immediately under-stand, because it is about what you want to do with the equipment, not how

knowl-it is designed and assembled In a similar manner, imagine that knowl-it is ble for you to acquire the right amount of the right kind of knowledge ofthe operation of the company from which you bought something you want

possi-to complain about, or of the government agency from which you are trying

to get a license but have not succeeded yet In summary, imagine that thebusiness processes of these enterprises have become transparent to you.Would that not be great? So, this is our goal: to offer a new understanding

of systems of any kind, and of enterprises in particular, such that one isable to look through the distracting and confusing appearance of an enter-prise right into its deep kernel, like an X-ray machine can let you lookthrough the skin and the tissues of the body right into the skeleton As auser of systems, this understanding lets you become master again of youractivities As a designer, it lets you design systems in such a way that theresulting design, in particular, the user dialogue and interface, reflects theessence of the system We will try to achieve this goal through a notion ofontology that includes the dynamic aspects of a system, and that at thesame time does justice to the nature of enterprises This nature is that en-terprises are social systems, of which the operating principle consists ofthe ability of human beings to enter into and comply with commitments

So, that will be our notion of enterprise ontology; and, as a quality rion for evaluating enterprise ontologies, we will apply the five propertiesthat were discussed earlier: coherence, comprehensiveness, consistency,conciseness, and essence, collectively abbreviated as C4E The particularmethodology that we will present lets you develop the ontology of an en-terprise in a systematic way But we do not require you to become a pro-fessional developer of ontologies The explanation of the methodology andthe demonstration of the example cases serve only to have you internalizethese kinds of ontologies, such that, after having studied the book, you areable to understand them and take full advantage of them Of course, werecognize that real great benefit can only be achieved if many people getused to these ontologies, ideally, everybody However, be assured that

crite-2 What is Enterprise Ontology? 13

even if you will be the only one in your working environment who sesses this new understanding of enterprises, you will be rewarded forstudying the book and learning the methodology We can assert this on thebasis of over 15 years practical experience in applying the DEMO3meth-odology, which this book elaborates, to all kinds of enterprises and for allkinds of purposes Even those who only participated inDEMOprojects forone or two days (mostly managers), invariably mentioned that they neverbefore had seen such a coherent, comprehensive, consistent, and concisepicture of what they agreed was the essence of the operation of their enter-prise Concluding, you will learn how to have more control over your pro-fessional life, how to take the lead again, both as a user and as a designer

pos-of enterprises, such that with your help the ideal situation as sketchedabove may be accomplished at some future time The only thing you need

to do in return is to put aside your current way of thinking about prises and to open your mind for new ideas; in short, to be willing to learn

enter-3 DEMO is an acronym that has had several long forms in the course of time, ing with “Dynamic Essential MOdelling” The current one is “Design and Engi- neering Methodology for Organisations” Visit www.demo.nl for more informa- tion.

start-3 An Explanatory Case

In the previous chapter, we defined enterprise ontology as the realizationand implementation independent essence of an enterprise, in short, as thedeep structure behind its observable surface structure In this chapter, wewill demonstrate the notion of enterprise ontology that will be developedmeticulously in the remainder of this book The intention now is to offer aglobal introduction, from which the core essence of enterprise ontologywill become clear We will explain the relevant notions on the basis of asmall example enterprise, namely the activities within a tennis club regard-ing the registration of new members The following exposition applies:

One can become member of the Volley tennis club by sending a letter to the club by postal mail In that letter one has to mention surname and first name, birth date, sex, telephone number, and postal address (street, house number, zip code, and residence) Charles, the administrator of Volley, empties daily the mailbox and checks whether the information provided is complete If not, he makes a telephone call to the sender in order to com- plete the data If a letter is completed, Charles adds an incoming mail number and the date, records the letter in the letter book, and archives it Every Wednesday evening, Charles takes the collected letters to Miranda, the Secretary of Volley He also takes the member register with him If Miranda decides that an applicant will become member of Volley, she stamps ‘new member’ on the letter and writes the date below it This date is the commencement date of the membership She then hands the let- ter to Charles in order to add the new member to the member register This is a book with numbered lines Each new member is entered on a new line The line number is the member number, by which the new member is referenced in the administration.

Next, Miranda calculates the membership fee that the new member has

to pay for the remaining part of the calendar year She finds the amount due for annual fees, as decided at the general meeting, on a piece of paper

in the drawer of her desk Then, she asks Charles to write down the amount in the member register.

If Miranda does not allow an applicant to become a member (e.g., cause he or she is too young or because the maximum number of members

mem-to be performed by bank transfers.

As soon as a payment is received, Charles prints a membership card on which are mentioned the membership number, the commencement date, the name, the birth date, and the postal address The card is sent to the new member by postal mail.

3.1 The Analysis of the Case Volley

As the first step in revealing the deep structure that lies hidden in this scription, we produce a set of flow charts that show the described proc-esses So, we start with taking the process view This is very common inpractice, and very appropriate in this case too The Flow Chart is an ade-quate technique for our purpose, although one may use the Petri Net [39]

de-or the EPC [41] as well What all these techniques basically show is thesequence of actions that are performed, the information that is used as in-put, and the information that is produced as output Fig 3.2, 3.3, and 3.4exhibit the flow charts of Volley Figure 3.1 contains the legend of theFlow Chart technique The sausage-like shape in the flow charts presented

is just a connector from one part of the flow chart to another

choice of paths information set action

flow direction

Fig 3.1 Legend of the Flow Chart

3.1 The Analysis of the Case Volley 17

check info

in letter [administrator]

Is the info complete?

LETTER BOOK

archive letter

archived application letter

begin

additional info

[administrator]

Fig 3.2 Flow Chart 1 of Volley

The distinction axiom that will be presented and discussed in Chap 12,tells us that the actions in the flow charts can be classified as either data-logical, infological, or ontological The notions datalogical and infologicalare the same as proposed originally by Langefors [42] Furthermore, wehave to anticipate the distinguishing of production acts and coordinationacts, as stated by the operation axiom (Chap 9)

A datalogical production act is an act in which one manipulates the

form of information, commonly referred to as data, so without being cerned about its content

con-18 3 An Explanatory Case

archived application letter

deciding about enrollment [secretary]

No

new membership

enter new membership MEMBER

REGISTER

calculate membership fee

GENERAL

MEETING

RESOLUTIONS

enter fee in member register

end A

refusal letter

sending refusal letter [administrator]

3.1 The Analysis of the Case Volley 19

A

make first invoice [administrator]

invoice for new member

sending invoice

make membership card

membership card

sending of membership card

receiving copy of payment [administrator]

Fig 3.4 Flow Chart 3 of Volley

Acts like copying, storing, and transmitting data are typical datalogicalacts For example, the act of recording an application for membership inthe letter book (Fig 3.2) is considered to be a datalogical act Next, speak-ing, listening, writing, and reading are typical datalogical coordination

acts In contrast, an infological production act is an act in which one is not

concerned about the form but, instead, about the content of informationonly Typical infological acts are inquiring, calculating, and reasoning As

an example, calculating the membership fee (Fig 3.3) is considered to be

an infological act Regarding the coordination between people, formulatingthoughts (in written or spoken sentences) and interpreting perceived(through listening or reading) sentences are typical infological coordina-

tion acts An ontological act is an act in which new original things are

brought about Deciding and judging are typical ontological production

20 3 An Explanatory Case

acts For example, the deciding about enrollment in Volley (see Fig 3.3) is

an ontological act Regarding the coordination between people, typical tological acts are requesting and promising Coordination acts appear tooccur in a particular pattern, called the transaction (Chap 10)

result stated

result accepted

result promised

result produced

request

desired result

result requested

promise

state accept

Fig 3.5 The transaction pattern

In Fig 3.5, we present the simplest form of this transaction pattern Itshows that the bringing about of an original new, thus, ontological, pro-duction result (in this case the delivery of a bouquet of flowers) starts withthe requesting of this result by someone in the role of customer fromsomeone in the role of producer The original new thing that is created by

this act, as is the case for every coordination act, is a commitment

Carry-ing through a transaction is a “game” of enterCarry-ing into and complyCarry-ing withcommitments For example, if the customer issues the request, he commitshimself to perform this act and, consequently to create the state “result re-quested” He cannot, after having done this, say that he did not mean it,that he was only joking Likewise, if the producer responds to the requestwith a promise, and, thus brings the transaction process to the state “resultpromised”, he is committed to this act

So, the process starts with the request for the bouquet by the customer,which brings the process to the state “result requested”, the result being theownership by the customer of the desired bouquet The producer responds

to the state “result requested” by promising to bring about the desired sult, which brings the process to the state “result promised” This repre-sents a to-do item for the producer: he has to comply with the promise byactually producing the bouquet of flowers, i.e., executing the production

re-3.1 The Analysis of the Case Volley 21

act In the act of handing over the bouquet to the customer, he states that

he has complied with his promise The process now comes to the state sult stated” The customer responds to this state by accepting the result asproduced This act completes the transaction pattern

“re-There appear to be two (ontological) transactions One is about ing member of the tennis club (let us call it T01), and the other is aboutpaying the first membership fee (T02) The analysis of the case Volley up

becom-to now is exhibited in Fig 3.6, 3.7, and 3.8 In these figures, actions aremarked as datalogical, infological, or ontological In addition, the coordi-nation acts in the two transactions are indicated Comparing the identifiedcoordination acts with the ones shown in Fig 3.5 reveals that the promise

of T01 and the acceptance of T01 are missing This is very often the case

in practice It does not mean that they are not performed (they must be forthe successful carrying through of a transaction), but that they are per-

formed tacitly This means that an aspirant member may assume that

Vol-ley has promised to deal with his or her application if he or she does notget an explicit assertion of the opposite, i.e., the denial to do it Likewise,Volley may assume, after having sent the membership card to the newmember, that everything is okay, unless the member explicitly, and within

a reasonable term, says that something is not right A similar reasoningholds for the absence of the explicit performance of the promise of T02 (bythe new member) and the acceptance of T02 (by Volley)

22 3 An Explanatory Case

check info

in letter [administrator]

Is the info complete?

LETTER BOOK

archive letter

archived application letter

begin

additional info

[administrator]

T01/request (Note: the corresponding T01/promise

3.1 The Analysis of the Case Volley 23

archived application letter

deciding about enrollment [secretary]

No

new membership

enter new membership MEMBER

REGISTER

calculate membership fee

GENERAL

MEETING

RESOLUTIONS

enter fee in member register

end A

refusal letter

sending refusal letter [administrator]

24 3 An Explanatory Case

A

make first invoice [administrator]

invoice for new member

sending invoice

make membership card

membership card

sending of membership card

receiving copy of payment [administrator]

datalogical

T02/request (Note: the corresponding T02/promise is performed tacitly)

(Note: the corresponding T02/accept is performed tacitly)

Fig 3.8 Analysis of Flow Chart 3

3.2 The Ontological Model of the Case Volley

The next step in arriving at the ontological model of Volley is to focus clusively on the ontological production acts and coordination acts in thetwo identified transactions The basic steps in both transaction processes,

ex-as well ex-as their relationships, are exhibited in the Process Model of Fig.

3.9 T01/rq, T01/pm, T01/ex4, T01/st, and T01/ac respectively stand for therequest and the promise of T01, the execution of T01, and the statementand acceptance of T01 A similar reasoning holds for transaction T02 A

4 In the diagram, T01/ex is represented by the gray-colored box with a diamond in

it and the name “T01” A similar explanation holds for T02/ex.

3.2 The Ontological Model of the Case Volley 25

solid arrow is a causal link For example, the being requested of T01

‘causes’ the promise of T01, not as an inevitable cause-effect sequence,like a trigger and its effect, but as a possible act of actor role A01 in deal-ing with the being requested of T01 In dealing with T01/pm, two acts areperformed: the execution of T01 and the request of T02 However, the ac-tual performing of the execution of T01 has to wait for T02 to be accepted;this is the meaning of the dashed arrow from T02/ac to the execution sym-bol of T01 In this way, the two transactions are connected; more specifi-cally, the T02 is enclosed in the T01 Because of the causal and conditional(waiting) relationships between T01 and T02, they constitute together abusiness process, the only business process in (the considered part of) Vol-ley The Process Model shows the process structure of the business proc-esses The actual sequence of steps in a successful instance of this process

is T01/rq, T01/pm, T02/rq, T02/pm, T02/ex, T02/st, T02/ac, T01/ex,T01/st, T01/ac The only unsuccessful kind of instance that we have takeninto account consists of T01/rq followed by T01/dc (decline) (see Fig 3.7).The asking for additional information by the administrator (see Fig 3.6) is

‘contained’ in the performing of T01/rq; every coordination act can logically succeed only if there is full consensus about its content

onto-T01

T01 rq

T01 pm

T01

T02 rq

T02 pm

T02

A01 CA01

membership payment

membership start

CA02

T01 dc

x

T02

Fig 3.9 Ontological Process Model of Volley

There are three distinct actor roles in Volley, denoted by A01, CA01,and CA02 The corresponding gray-lined rectangles enclose the acts for

26 3 An Explanatory Case

which each of them is authorized and responsible So, for example, actorrole CA01 is responsible for T01/rq and T01/ac For every coordinationstep (represented by a disk in a box), there is an action rule that guideshow the performing actor role should respond to the reached status Below,

we present all action rules for actor role A01 The collection of action rules

for an enterprise is called its Action Model The Action Model is another

ontological aspect model of an enterprise, along with the Process Model(discussed above), the State Model, and the Construction Model (the lattertwo to be discussed shortly) A complete and thorough discussion of thefour ontological aspect models is contained in Part D of the book

Every action rule is enclosed by an on-no bracket pair The on clause

specifies the status that is being dealt with Conditional responses (choices)

are represented by an if clause (enclosed in an if-fi bracket pair) If there is

more than one choice, the second and following ones are preceded by thesymbol “” Every choice consists of the condition, which is checked to betrue, followed by the symbol “
”, followed by the action(s) to take Thereader is invited to compare the rules with the description of Volley

on requested T01(M) with member(new M) = P

if age(P) < minimal_age or

#members(Volley)=maximum_number(current_year)
decline T01(M)

age(P) minimal_age and

#members (Volley) < maximum_number(current_year)
promise T01(M)

if the current number of members of Volley equals the maximum numberaccepted All other applications are responded to with a promise

on promised T01(M)

request T02(M) with first_fee(M)

no

3.2 The Ontological Model of the Case Volley 27

This action rule specifies how to deal with the being promised of a T01,regarding a particular membership M The response is requesting payment(T02) of the first fee for this membership The first fee is a derived facttype for which the derivation rule is presented shortly The other actionrules are presented hereafter without further comment

on stated T02(M)

not < payment is acceptible >
reject T02(M)

The next aspect model of the complete ontological model of Volley is

the State Model It shows all information items (object classes and fact

types) that occur in the Action Model The State Model of Volley is ited in Fig 3.10 The diagramming technique is based onORM[34] Read-ers who are not familiar with this technique are directed to the book refer-enced or to Chap 5 Fact types that are pure mathematical functions arenot included in the diagram but listed as properties below the diagram Thediagram must be interpreted as follows In the world of Volley there is anobject or entity typeMEMBERSHIP and an object or entity typePERSON.Instances of the last one are created outside the scope of Volley; therefore,the “roundangle” ofPERSONis colored gray Instances ofMEMBERSHIPare created within Volley With every such instance, an instance of the fact

exhib-type “the member of M is P” is associated; it states that a particular person

is the member in a particular membership For every membership there isexactly one person as the member, however a person may be the member

of more than one membership

There are two production fact types, i.e., results of transactions, namely,PF01 and PF02 Producing a PF01 for a particular membership means thatthe membership then starts to exist Producing a PF02 means that the firstfee for the membership has been paid

28 3 An Explanatory Case

PF02

PF01 membership M has been started

the first fee for membership

M is paid

the member of M is P

SHIP PERSON

maximum_number


VOLLEY

NUMBER

number_of members (*)

VOLLEY

NUMBER

first_fee (*)



MEMBERSHIP
EURO

date_of_birth



PERSON

DATE

Fig 3.10 Ontological State Model of Volley

The object property list below the diagram shows four properties ofVolley itself: the minimum age that a member must have, the annual feethat has to be paid, the maximum number of members that are allowed andthe current number of members Memberships have one property, namely,the first fee that has to be paid Lastly, persons have two properties: thedate of birth and the age The state model contains only those object or en-tity types, fact types, and properties that occur in the Action Model Someproperties are marked by an asterisk (*); these are derived fact types Theirderivation rules are:

#members (Volley) = < number of persons for which there exists rently a membership >

cur-first_fee (M) = ((12 - Current_Month#)/12) * annual_fee(Current_Year)age(P) = < difference in years between birth year of P and current year >

The last, and most comprehensive, aspect model is the Construction Model (Fig 3.11) The two transaction patterns in Fig 3.9 are, so to speak,

‘compressed’ to only one symbol in Fig 3.11, namely, the disk with adiamond in it The two actor roles involved are represented by a box Theyare connected to the transaction symbol by means of a straight line The

3.2 The Ontological Model of the Case Volley 29

producer has, in addition, a small black box on the junction of this line andthe box shape The gray-lined rectangle in the back represents the bound-ary of the enterprise under consideration By convention, all environmentalcomponents are colored gray, and their codes start with a C for “compos-ite” So, the customer of transaction type T01 is the environmental (com-posite) actor role CA01, the aspirant member role Likewise, the producer

of T02 is the environmental (composite) actor role CA02, the payer role.The distinction between these two roles illustrates that an ontologicalmodel is independent of the way in which it is or might be implemented:the person who wants to become member may fulfill both role CA01 andCA02, but it may also be the case that somebody else fulfills role CA02,and it may even be the case that that somebody else fulfills role CA01

CA02 aspirant

CPB02

CPB01

membership start

membership payment

admission rules

personal data

Fig 3.11 Ontological Construction Model of Volley

Assigning actor roles to persons is a matter of implementing the prise; one completely abstracts from it on the ontological level of under-standing the enterprise There is only one internal actor role, namely, A01

enter-It is colored white because it is an elementary actor role An elementaryactor role is an atomic amount of authority and responsibility It is pro-ducer of exactly one transaction, and customer of zero, one, or more trans-actions In our case, A01 is the producer of T01 and the customer of T02

An actor role is fulfilled by a human being If one rereads the description

of Volley, either the expository one or the set of flow charts, then it

ap-30 3 An Explanatory Case

pears that actor role A01 is fulfilled by both Miranda and Charles As wewill see later on in the book, this is a deviation from the ideal implementa-tion of an actor role, which has its price, namely, the need for a continuoustuning in of the norms and values of both persons

The transaction symbol actually has two interpretations In addition torepresenting a transaction type of which instances are carried through inthe enterprise, it is the combination of a coordination bank (the disk) and aproduction bank (the diamond) A coordination bank contains all coordina-tion facts created; they allow one to monitor the progress of all transactioninstances A production bank contains all created production facts Produc-tion bank PB01 (of T01) contains instances of the typeMEMBERSHIPaswell as instances of the fact type “the member of M is P” and instances ofthe fact type “membership M has been started” Production bank PB02 (ofT02) contains instances of the fact type “the first fee for membership Mhas been paid” There are two composite production banks (CPB01 andCPB02) They contain facts that are needed by A01 but that are producedoutside the enterprise under consideration CPB01 contains instances ofthe property types “minimum age”, “annual fee”, and “maximum number”.CPB02 contains instances of the typePERSONas well as instances of theproperty type “date of birth” The dashed lines indicate that actor role A01(the only internal actor role) has access to the contents of these banks Asthe producer of T01 and the customer of T02, it has also access to the pro-duction bank and the coordination bank of T01 and to those of T02

It is time to assess what we have done so far What is exactly the ence between the ontological model of Volley, expressed in the four aspectmodels, and any other kind of model, like the flow chart? What are thebenefits of the ontological model, and for what purposes? Without trying

differ-to be exhaustive, let us look at some striking matters and discuss theirpractical consequences

First, the ontological model of Volley is really and fully abstracted fromthe current way in which it operates, according to the description at the be-ginning of this chapter It does not contain organizational functions, likesecretary and administrator and members, or references to persons(Miranda or Charles) It also does not contain any infological or datalogi-cal things: no computing, no inquiring, no letter book, no register, etc.Moreover, it completely abstracts from the way in which the fulfillers ofthe distinguished actor roles communicate: no letters, no telephone calls,etc These properties, of course, make the ontological model very stable Anew organizational structure will not change the ontology; nor will replac-ing letters by e-mails; nor will replacing the register book by a database,and so on

3.2 The Ontological Model of the Case Volley 31

Second, the ontological model of Volley shows things that have no plicit implementation, like the promise and the acceptance of T01 As youmay tell from your own experience, these omissions are potential break-downs in business processes as soon as changes are implemented, e.g., thereplacement of an employee by somebody new (who does not know all the

ex-“no news is good news” rules that are in place) Put differently, the logical model is the only right starting point for proposing, analyzing, andimplementing such changes because there is only one thing that one has tosee to while bringing about changes in an enterprise It is that all parts ofthe ontological model are implemented, nothing more and nothing less.The ontological model also provides the right amount of (re)design and(re)engineering freedom For example, the state model of Volley rightlydoes not contain the postal mail address of members Such information isdependent on the particular communication channel one chooses, in thiscase the postal mail service However, should we choose e-mail, we wouldneed e-mail addresses instead The ontological model only states in this re-spect that the (fulfillers of the) two actor roles must be able to communi-cate

onto-Third, the four aspect models constitute a truly coherent set of models,based on a common theoretical foundation (that will be explained in thebook) This is rather unique There are hardly any competing methodolo-gies for enterprise ontology because the models produced with them arenot guaranteed to be implementation independent In addition, the aspectmodels in most approaches do not constitute a complete set and are not in-trinsically linked to each other For example, one may add a data model to

a process model, but they cannot be linked together intrinsically because ofthe missing common foundation; they can only be “talked” together andthat is what often happens in practice

Fourth, the Process Model (Fig 3.9) shows the structure of the (single)business process in Volley: transaction type T02 enclosed in transactiontype T01 This structure is lost when using current process modeling ap-proaches, such as the Flow Chart, the Petri Net and the Event Driven Proc-ess Chain They can exhibit only the sequences of actions This sequencehowever can easily be derived from a Process Model For example, the se-quence of steps in the Process Model of Volley is:

T01/ac (which ends the process successfully)

Reengineering a business process is something that is done for every tological process step Starting not from the ontological model but from animplementation dependent model, like the ones we mentioned, may be-come tedious but will normally not raise real problems Redesigning abusiness process, however, means removing or adding complete transac-tions One can imagine that without insight into the transactional structure,such a redesign cannot possibly be done in a guaranteed proper and correctway The only guarantee that can be given is that the new process is for-mally correct, as can be checked, for example, for Petri Nets Obviously,that is not enough; it is also not the first thing to worry about

on-This concludes the introduction to the notion of enterprise ontology Aswas said at the beginning, we inevitably had to go through it in large steps,which means that you may find yourself now left with a lot of unansweredquestions Don not get discouraged by these questions at the moment Itmay even be an advantage in studying the remainder of the book becauseyour study will be more goal-directed than otherwise We only hope that

by now you recognize and appreciate that the current practice of changingorganizations, in particular of redesigning and reengineering businessprocesses, and of supporting them with new ICT applications, can at best

be suboptimal and at worst lead to the failures we all know of