The REA-DSL : A Domain Specific Modeling Language for Business Models
Advanced Information Systems Engineering Proceedings of the 23rd International Conference CAiSE 2011 (2011)
- ISBN: 9783642216398
Available from
Alessio Maria Braccini's profile on Mendeley.
or
Page 1
The REA-DSL : A Domain Specific Modeling Language for Business Models
The REA-DSL: A Domain Speci c Modeling
Language for Business Models
C. Sonnenberg2, C. Huemer1, B. Hofreiter2, D. Mayrhofer1, A. Braccini3
1TU Vienna, 2University of Liechtenstein,3LUISS University
1flastg@big.tuwien.ac.at, 2ffirst.lastg@hochschule.li, 3abraccini@luiss.it
Abstract. In the discipline of accounting, the resource-event-agent (REA)
ontology is a well accepted conceptual accounting framework to analyze
the economic phenomena within and across enterprises. Accordingly, it
seems to be appropriate to use REA in the requirements elicitation to
develop an information architecture of accounting and enterprise infor-
mation systems. However, REA has received comparatively less attention
in the eld of business informatics and computer science. Some of the
reasons may be that the REA ontology despite of its well grounded core
concepts is (1) sometimes vague in the de nition of the relationships
between these core concepts, (2) misses a precise language to describe
the models, and (3) does not come with an easy to understand graphical
notation. Accordingly, we have started developing a domain speci c mod-
eling language speci cally dedicated to REA models and corresponding
tool support to overcome these limitations. In this paper we present our
REA DSL which supports the basic set of REA concepts.
1 Introduction
Analyzing the economic phenomena on which companies base their business
may serve as a good starting point in the requirements elicitation phase when
developing enterprise information systems. Business models specify - amongst
other things - the main actors, their relationships and the values exchanged
between them (cf. [1]).
We see three main ontologies to conceptualize business models: the Business
Model Ontology (BMO) [2], the e3-value ontology [3], and the Resource-Event-
Agent ontology (REA) [4]. BMO is easy to use by the domain expert because
it focuses mainly on the categorization of aspects relevant for the delivery of
products and services to ful ll customers' requests. It helps the domain expert
to ask herself the right questions when developing a business model, but has
a limited focus on conceptualizing the elements of the business model. In con-
trary, e3-value de nes a conceptual model to describe the exchanges of value
among actors in a network. e3-value comes with a graphical syntax that is easy
understood by the domain expert. Furthermore, it allows the domain expert to
perform nancial assessment of the value exchanges.
The Resource-Event-Agent (REA) ontology has its roots in the accounting
discipline and was originally developed as a reference framework to conceptualize
Language for Business Models
C. Sonnenberg2, C. Huemer1, B. Hofreiter2, D. Mayrhofer1, A. Braccini3
1TU Vienna, 2University of Liechtenstein,3LUISS University
1flastg@big.tuwien.ac.at, 2ffirst.lastg@hochschule.li, 3abraccini@luiss.it
Abstract. In the discipline of accounting, the resource-event-agent (REA)
ontology is a well accepted conceptual accounting framework to analyze
the economic phenomena within and across enterprises. Accordingly, it
seems to be appropriate to use REA in the requirements elicitation to
develop an information architecture of accounting and enterprise infor-
mation systems. However, REA has received comparatively less attention
in the eld of business informatics and computer science. Some of the
reasons may be that the REA ontology despite of its well grounded core
concepts is (1) sometimes vague in the de nition of the relationships
between these core concepts, (2) misses a precise language to describe
the models, and (3) does not come with an easy to understand graphical
notation. Accordingly, we have started developing a domain speci c mod-
eling language speci cally dedicated to REA models and corresponding
tool support to overcome these limitations. In this paper we present our
REA DSL which supports the basic set of REA concepts.
1 Introduction
Analyzing the economic phenomena on which companies base their business
may serve as a good starting point in the requirements elicitation phase when
developing enterprise information systems. Business models specify - amongst
other things - the main actors, their relationships and the values exchanged
between them (cf. [1]).
We see three main ontologies to conceptualize business models: the Business
Model Ontology (BMO) [2], the e3-value ontology [3], and the Resource-Event-
Agent ontology (REA) [4]. BMO is easy to use by the domain expert because
it focuses mainly on the categorization of aspects relevant for the delivery of
products and services to ful ll customers' requests. It helps the domain expert
to ask herself the right questions when developing a business model, but has
a limited focus on conceptualizing the elements of the business model. In con-
trary, e3-value de nes a conceptual model to describe the exchanges of value
among actors in a network. e3-value comes with a graphical syntax that is easy
understood by the domain expert. Furthermore, it allows the domain expert to
perform nancial assessment of the value exchanges.
The Resource-Event-Agent (REA) ontology has its roots in the accounting
discipline and was originally developed as a reference framework to conceptualize
Page 2
economic phenomena in an enterprise. In its proposal in 1982, McCarthy already
had the vision to facilitate the design of data structures of accounting information
systems by means of REA [4]. Since this time the REA model has been further
extended and evolved into a domain ontology [5]. All REA concepts are based on
well established concepts of the literature in economic theory - which is certainly
one of the strengths of REA. However, REA has no dedicated representation
format and, consequently, no graphical syntax. Thus, users may struggle when
describing the REA models leading to the impression that REA is a rather
heavy-weight approach. McCarthy feels (as he expressed during our joint work
in standardization activities) that a dedicated graphical syntax - such as it exists
for e3-value - may help in overcoming this problem and may lead to a much
more signi cant adoption of REA. Accordingly, we have started the endeavor of
developing a domain speci c modeling language for REA.
A domain-speci c language (DSL) is a small, usually declarative language. It
o ers expressive power through appropriate notations and abstractions focused
on { and usually restricted to { a particular problem domain [6]. Besides textual
DSLs, we see an increasing interest in domain-speci c modeling languages [7,
8]. According to [6] the bene ts of a DSL approach are manifold: They allow
solutions to be expressed at the level of abstraction of the problem domain.
As a matter of fact, domain experts themselves can understand, validate and
often modify DSL programs/models. The DSL programs/models are concise and
self documenting to a large extent. They enhance productivity, reliability and
maintainability. DSLs allow for validation and optimization at the domain level.
When developing our REA-DSL we followed methodological steps that have
been suggested by Strembeck and Zdun for the systematic development of do-
main speci c languages [9]. Amongst other variants, they describe the develop-
ment process for extracting a DSL from an existing system, which is appropriate
for our needs, because we extract the DSL from the existing REA ontology. Ac-
cordingly, we started with (1) the identi cation of elements in the REA ontology.
Next, we underwent a number of revision cycles of (2) deriving the abstract syn-
tax of the REA model including the core language model and the language
model constraints and (3) de ning the DSL behavior, i.e. determining how the
language elements of the DSL interact to produce the intended behavior. Once
we had reached a stable state, we de ned the DSL concrete syntax (4). Finally,
we implemented a modeling tool support for the DSL (5), but we skipped the
last step described in [9], i.e. integrating the DSL into a software platform, since
REA stays at the platform independent level.
The remainder of this paper is structured as follows: In Section 2 we give a
basic introduction into the REA ontology including an illustrative example. The
core of the paper in Section 3 presents our DSL for REA models. We elaborate on
the meta model of our REA DSL for describing duality models and value chains.
Furthermore, we use the same example as in Section 2 to illustrate our DSL.
Section 4 reports on our evaluation by means of a REA DSL tool. A summary
and remarks on future work conclude the paper in Section 5.
had the vision to facilitate the design of data structures of accounting information
systems by means of REA [4]. Since this time the REA model has been further
extended and evolved into a domain ontology [5]. All REA concepts are based on
well established concepts of the literature in economic theory - which is certainly
one of the strengths of REA. However, REA has no dedicated representation
format and, consequently, no graphical syntax. Thus, users may struggle when
describing the REA models leading to the impression that REA is a rather
heavy-weight approach. McCarthy feels (as he expressed during our joint work
in standardization activities) that a dedicated graphical syntax - such as it exists
for e3-value - may help in overcoming this problem and may lead to a much
more signi cant adoption of REA. Accordingly, we have started the endeavor of
developing a domain speci c modeling language for REA.
A domain-speci c language (DSL) is a small, usually declarative language. It
o ers expressive power through appropriate notations and abstractions focused
on { and usually restricted to { a particular problem domain [6]. Besides textual
DSLs, we see an increasing interest in domain-speci c modeling languages [7,
8]. According to [6] the bene ts of a DSL approach are manifold: They allow
solutions to be expressed at the level of abstraction of the problem domain.
As a matter of fact, domain experts themselves can understand, validate and
often modify DSL programs/models. The DSL programs/models are concise and
self documenting to a large extent. They enhance productivity, reliability and
maintainability. DSLs allow for validation and optimization at the domain level.
When developing our REA-DSL we followed methodological steps that have
been suggested by Strembeck and Zdun for the systematic development of do-
main speci c languages [9]. Amongst other variants, they describe the develop-
ment process for extracting a DSL from an existing system, which is appropriate
for our needs, because we extract the DSL from the existing REA ontology. Ac-
cordingly, we started with (1) the identi cation of elements in the REA ontology.
Next, we underwent a number of revision cycles of (2) deriving the abstract syn-
tax of the REA model including the core language model and the language
model constraints and (3) de ning the DSL behavior, i.e. determining how the
language elements of the DSL interact to produce the intended behavior. Once
we had reached a stable state, we de ned the DSL concrete syntax (4). Finally,
we implemented a modeling tool support for the DSL (5), but we skipped the
last step described in [9], i.e. integrating the DSL into a software platform, since
REA stays at the platform independent level.
The remainder of this paper is structured as follows: In Section 2 we give a
basic introduction into the REA ontology including an illustrative example. The
core of the paper in Section 3 presents our DSL for REA models. We elaborate on
the meta model of our REA DSL for describing duality models and value chains.
Furthermore, we use the same example as in Section 2 to illustrate our DSL.
Section 4 reports on our evaluation by means of a REA DSL tool. A summary
and remarks on future work conclude the paper in Section 5.
Page 3
2 Resource - Event - Agent (REA)
2.1 The REA Ontology
The objective of the REA ontology is the conceptualization of common eco-
nomic phenomena of a rm independent of application-speci c demands. REA
accounting information systems focus on economic exchanges as the central unit
of analysis. Instead of representing these exchanges with double-entry bookkeep-
ing artifacts (e.g. debits, credits, accounts), REA proposes concepts and patterns
to derive semantic models of economic exchanges and transformations. The un-
derlying assumption of REA is that all business enterprises operate in the same
manner [10] according to an entrepreneurial script: acquiring nancial resources,
engaging in a chain of economic exchanges with other parties, each time giving
up an economic resource in return for another resource of greater value [10].
After executing this script, the business generates a justi able pro t after hav-
ing paid interests and creditors. This entrepreneurial script essentially discloses
the entrepreneurial rationale of a business. Hence, the REA ontology is not only
used to facilitate the design of accounting information systems but in particular
for business modeling.
The basic REA ontology is a stereotypical representation of an economic ex-
change as a core economic phenomenon [5]. This exchange is executed between
parties inside and outside of a rm's boundaries and follows a particular object
pattern (cf. [5], see also Figure 1(a) and 1(b)). In order to conceptualize this
pattern, the REA ontology suggests three concepts that constitute an exchange:
resource, event, and agent. Resources are things being exchanged between partic-
ipating agents. In an exchange, an agent (inside agent) usually gives up control
of a resource to an outside agent in order to gain control over another resource.
Events occur in the course of executing economic activities. In REA basically
two types of events are distinguished: increment and decrement events. Exten-
sions of the REA ontology [11] also distinguish between transfer (exchanges
with external actors) and transformation (concerns value creation within the
rm) events.
Furthermore, the following economic primitives (relationship types) are spec-
i ed by the REA: duality, stock-
ow, and participation. A duality relationship
connects decrement events with corresponding increment events and thus pro-
vides the rationale of individual economic activities. Stock-
ow relationships con-
nect economic resources with economic events (decrement or increment events).
Depending on the connected event type, the following stock-
ow relationship
types are distinguished: give and take (transfer events), use, consume, and pro-
duce (transformation events). Participation relationships describe the involve-
ment of an agent in an economic event. As such, the basic REA ontology not
only conceptualizes economic exchanges but also relates to business process con-
cepts. It captures Who is involved in an exchange (economic agents), What is
being exchanged (economic resources), When (and under what conditions) do
the components of an exchange occur (economic events), Why are the partici-
pants engaged in an exchange (duality relationships, stock-
ows), and How do
2.1 The REA Ontology
The objective of the REA ontology is the conceptualization of common eco-
nomic phenomena of a rm independent of application-speci c demands. REA
accounting information systems focus on economic exchanges as the central unit
of analysis. Instead of representing these exchanges with double-entry bookkeep-
ing artifacts (e.g. debits, credits, accounts), REA proposes concepts and patterns
to derive semantic models of economic exchanges and transformations. The un-
derlying assumption of REA is that all business enterprises operate in the same
manner [10] according to an entrepreneurial script: acquiring nancial resources,
engaging in a chain of economic exchanges with other parties, each time giving
up an economic resource in return for another resource of greater value [10].
After executing this script, the business generates a justi able pro t after hav-
ing paid interests and creditors. This entrepreneurial script essentially discloses
the entrepreneurial rationale of a business. Hence, the REA ontology is not only
used to facilitate the design of accounting information systems but in particular
for business modeling.
The basic REA ontology is a stereotypical representation of an economic ex-
change as a core economic phenomenon [5]. This exchange is executed between
parties inside and outside of a rm's boundaries and follows a particular object
pattern (cf. [5], see also Figure 1(a) and 1(b)). In order to conceptualize this
pattern, the REA ontology suggests three concepts that constitute an exchange:
resource, event, and agent. Resources are things being exchanged between partic-
ipating agents. In an exchange, an agent (inside agent) usually gives up control
of a resource to an outside agent in order to gain control over another resource.
Events occur in the course of executing economic activities. In REA basically
two types of events are distinguished: increment and decrement events. Exten-
sions of the REA ontology [11] also distinguish between transfer (exchanges
with external actors) and transformation (concerns value creation within the
rm) events.
Furthermore, the following economic primitives (relationship types) are spec-
i ed by the REA: duality, stock-
ow, and participation. A duality relationship
connects decrement events with corresponding increment events and thus pro-
vides the rationale of individual economic activities. Stock-
ow relationships con-
nect economic resources with economic events (decrement or increment events).
Depending on the connected event type, the following stock-
ow relationship
types are distinguished: give and take (transfer events), use, consume, and pro-
duce (transformation events). Participation relationships describe the involve-
ment of an agent in an economic event. As such, the basic REA ontology not
only conceptualizes economic exchanges but also relates to business process con-
cepts. It captures Who is involved in an exchange (economic agents), What is
being exchanged (economic resources), When (and under what conditions) do
the components of an exchange occur (economic events), Why are the partici-
pants engaged in an exchange (duality relationships, stock-
ows), and How do
Page 4
the exchanges materialize as economic activities or business processes (series of
small events that move business process through to completion) (cf. [12]).
EconomicResource
EconomicEvent
EconomicAgent
stock-flow
inside outside
<<duality>> +initiating +terminating
(a) REA Meta Model
<<EconomicResource>> Fish <<stock-flow>>
<<duality>> +initiating +terminating
<<EconomicResource>> Cash
<<EconomicEvent>> FishSale <<EconomicEvent>> Payment
<<EconomicAgent>> Salesperson <<EconomicAgent>> Customer <<EconomicAgent>> Cashier
<<stock-flow>>
- GIVE
+TAKE provide <<inside>> receive <<inside>> receive <<outside>> provide <<outside>>
(b) REA Object Constellation
Fig. 1. Basic REA Ontology
The basic REA ontology with its economic primitives is illustrated in Figure
1(a) by means of a UML class diagram. Figure 1(b) shows an instantiation of
the REA ontology, a so called object constellation. Furthermore, the ontology
de nes three axioms that restrict the use of the concepts and primitives for
conceptualizing economic exchanges (cf. [11]):
{ Axiom 1: At least one take event and one give event exist for each eco-
nomic resource (guarantees modeling of economic activities as a sequence of
exchanges).
{ Axiom 2: All events e ecting a resource decrement must be eventually
paired in duality relationships with events e ecting an increment and vice
versa (ensures correct enumerations of exchanges).
{ Axiom 3: Each exchange needs an instance of both the "inside" and the
"outside" agents(ensures presence of exchanges between parties with com-
peting economic interests).
Axioms one to three apply for transfer events. Axiom two also holds for
transformation events.
Since its proposal in 1982, some extensions to the basic REA ontology have
been proposed (e.g. [5]). The extended REA ontology envisions a vertical and
horizontal layering of economic exchanges. With regard to the vertical layering, a
hierarchy consisting of three levels is proposed: (1) the value chain speci cation
level, (2) the duality speci cation level, and (3) the task or work
ow level ([5]).
In this paper we focus on the upper two speci cation layers: the value chain and
the duality . Thereby, we base our work on the diagram style used by Geerts
and McCarty in [13], which has never been formalized.
The horizontal layering in REA enables the analysis of economic exchanges
at di erent points in time on all three vertical layers described above. There-
fore, the REA ontology considers an accountability infrastructure and a policy
infrastructure [5]. The REA accountability infrastructure conceptualizes actual
business events and captures "what has occurred" or "what is or has been". The
small events that move business process through to completion) (cf. [12]).
EconomicResource
EconomicEvent
EconomicAgent
stock-flow
inside outside
<<duality>> +initiating +terminating
(a) REA Meta Model
<<EconomicResource>> Fish <<stock-flow>>
<<duality>> +initiating +terminating
<<EconomicResource>> Cash
<<EconomicEvent>> FishSale <<EconomicEvent>> Payment
<<EconomicAgent>> Salesperson <<EconomicAgent>> Customer <<EconomicAgent>> Cashier
<<stock-flow>>
- GIVE
+TAKE provide <<inside>> receive <<inside>> receive <<outside>> provide <<outside>>
(b) REA Object Constellation
Fig. 1. Basic REA Ontology
The basic REA ontology with its economic primitives is illustrated in Figure
1(a) by means of a UML class diagram. Figure 1(b) shows an instantiation of
the REA ontology, a so called object constellation. Furthermore, the ontology
de nes three axioms that restrict the use of the concepts and primitives for
conceptualizing economic exchanges (cf. [11]):
{ Axiom 1: At least one take event and one give event exist for each eco-
nomic resource (guarantees modeling of economic activities as a sequence of
exchanges).
{ Axiom 2: All events e ecting a resource decrement must be eventually
paired in duality relationships with events e ecting an increment and vice
versa (ensures correct enumerations of exchanges).
{ Axiom 3: Each exchange needs an instance of both the "inside" and the
"outside" agents(ensures presence of exchanges between parties with com-
peting economic interests).
Axioms one to three apply for transfer events. Axiom two also holds for
transformation events.
Since its proposal in 1982, some extensions to the basic REA ontology have
been proposed (e.g. [5]). The extended REA ontology envisions a vertical and
horizontal layering of economic exchanges. With regard to the vertical layering, a
hierarchy consisting of three levels is proposed: (1) the value chain speci cation
level, (2) the duality speci cation level, and (3) the task or work
ow level ([5]).
In this paper we focus on the upper two speci cation layers: the value chain and
the duality . Thereby, we base our work on the diagram style used by Geerts
and McCarty in [13], which has never been formalized.
The horizontal layering in REA enables the analysis of economic exchanges
at di erent points in time on all three vertical layers described above. There-
fore, the REA ontology considers an accountability infrastructure and a policy
infrastructure [5]. The REA accountability infrastructure conceptualizes actual
business events and captures "what has occurred" or "what is or has been". The
Page 5
policy infrastructure conceptualizes what "could be" or "should be" within the
context of a de ned portfolio of a rm's resources and capabilities [5]. In this
paper we focus on the concepts associated with the REA accountability infras-
tructure. Concepts associated with the policy infrastructure like commitments,
agreements, and type images are not covered here but are subject to future work.
In the following section the REA ontology is applied to a simpli ed example
in order to demonstrate how the REA constructs can be used to specify an
entrepreneurial script.
2.2 REA Ontology example
The simpli ed example used to apply the REA ontology is taken from [13]. The
example is based on an actual company and is called Sy's Fish and is introduced
below:
Sy's Fish is a distributor of seafood and provides his base of restaurant cus-
tomers with over 50 types of sh which can be stored at all locations or stores.
However, each store usually specializes in local favorites. Fish are purchased from
local shers, cleaned at the store, and then sold at restaurants to customers. Cus-
tomers are allowed to buy on credit, and all pay on the last day of the month.
Most employees are generalists and can perform many duties such as purchasing,
cleaning, and delivering sh. They ll out time cards fortnightly upon which they
may note the percentage of time devoted each day to buying, cleaning and selling
sh. Non-generalist employees for the most part comprise of cashiers. Sy's also
possesses a
eet of trucks, used to bring sh from the docks and to deliver sh
to the restaurants. Both the truck and the employees involved in each purchase
and sale of sh are noted. All trucks are leased on yearly contracts, and lease
payments are made monthly. Cash receipts and disbursements are made to/from
one of the multiple checking accounts of the rm.
«Ec
ono
mic
Re
sou
rce
»
Ca
sh
«Ec
ono
mic
Eve
n...
Fis
hO
bta
ine
d
«Ec
ono
mic
Eve
nt»
Ca
shD
isb
urs
em
ent
Lea
se
«Ec
ono
mic
Eve
nt»
Lab
orA
cqu
isit
ion
«Ec
ono
mic
Eve
nt»
Lea
se
«Ec
ono
mic
Re
sou
rc..
.
Ca
sh
«Ec
ono
mic
Re
sou
rc..
.
Tru
ck
«Ec
ono
mic
Re
sou
rce
»
Fis
h
«Ec
ono
mic
Eve
nt»
Ca
shD
isb
urs
em
ent
Lab
or
«Ec
ono
mic
Re
sou
rc..
.
Ca
sh
«Ec
ono
mic
Re
sou
rc..
.
Lab
or
«Ec
ono
mic
Eve
nt»
Pay
me
ntF
orF
ish
«Ec
ono
mic
Eve
nt»
Tra
nsp
ort
In
«Ec
ono
mic
Eve
nt»
Tra
nsp
ort
Ou
t
«Ec
ono
mic
Re
sou
rce
»
Fis
h
«Ec
ono
mic
Eve
...
Cle
anI
n
«Ec
ono
mic
Eve
...
Cle
anO
ut
«Ec
ono
mic
Eve
nt»
Fis
hS
ale
«Ec
ono
mic
Eve
...
Pay
me
nt
«Ec
ono
mic
Re
sou
rc..
.
Fis
h
«Ec
ono
mic
Eve
nt»
Ca
sh
«du
alit
y»
giv
e
«st
ock
-flo
w»
«du
alit
y»
«du
alit
y»
tak
e
«st
ock
-flo
w»
«du
alit
y»
giv
e
«st
ock
-flo
w»
tak
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
«du
alit
y»
giv
e
«st
ock
-flo
w»
«du
alit
y»
giv
e
«st
ock
-flo
w»
tak
e
«st
ock
-flo
w»
tak
e
«st
ock
-flo
w»
tak
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
tak
e
«st
ock
-flo
w»
Fig. 2. REA Sy's Fish Example
The value chain of Sy's Fish (i.e. the entrepreneurial script) comprises the
processes payroll, truck acquisition and store processing. The last process is
further decomposed into the processes buying, cleaning, and selling. The process
context of a de ned portfolio of a rm's resources and capabilities [5]. In this
paper we focus on the concepts associated with the REA accountability infras-
tructure. Concepts associated with the policy infrastructure like commitments,
agreements, and type images are not covered here but are subject to future work.
In the following section the REA ontology is applied to a simpli ed example
in order to demonstrate how the REA constructs can be used to specify an
entrepreneurial script.
2.2 REA Ontology example
The simpli ed example used to apply the REA ontology is taken from [13]. The
example is based on an actual company and is called Sy's Fish and is introduced
below:
Sy's Fish is a distributor of seafood and provides his base of restaurant cus-
tomers with over 50 types of sh which can be stored at all locations or stores.
However, each store usually specializes in local favorites. Fish are purchased from
local shers, cleaned at the store, and then sold at restaurants to customers. Cus-
tomers are allowed to buy on credit, and all pay on the last day of the month.
Most employees are generalists and can perform many duties such as purchasing,
cleaning, and delivering sh. They ll out time cards fortnightly upon which they
may note the percentage of time devoted each day to buying, cleaning and selling
sh. Non-generalist employees for the most part comprise of cashiers. Sy's also
possesses a
eet of trucks, used to bring sh from the docks and to deliver sh
to the restaurants. Both the truck and the employees involved in each purchase
and sale of sh are noted. All trucks are leased on yearly contracts, and lease
payments are made monthly. Cash receipts and disbursements are made to/from
one of the multiple checking accounts of the rm.
«Ec
ono
mic
Re
sou
rce
»
Ca
sh
«Ec
ono
mic
Eve
n...
Fis
hO
bta
ine
d
«Ec
ono
mic
Eve
nt»
Ca
shD
isb
urs
em
ent
Lea
se
«Ec
ono
mic
Eve
nt»
Lab
orA
cqu
isit
ion
«Ec
ono
mic
Eve
nt»
Lea
se
«Ec
ono
mic
Re
sou
rc..
.
Ca
sh
«Ec
ono
mic
Re
sou
rc..
.
Tru
ck
«Ec
ono
mic
Re
sou
rce
»
Fis
h
«Ec
ono
mic
Eve
nt»
Ca
shD
isb
urs
em
ent
Lab
or
«Ec
ono
mic
Re
sou
rc..
.
Ca
sh
«Ec
ono
mic
Re
sou
rc..
.
Lab
or
«Ec
ono
mic
Eve
nt»
Pay
me
ntF
orF
ish
«Ec
ono
mic
Eve
nt»
Tra
nsp
ort
In
«Ec
ono
mic
Eve
nt»
Tra
nsp
ort
Ou
t
«Ec
ono
mic
Re
sou
rce
»
Fis
h
«Ec
ono
mic
Eve
...
Cle
anI
n
«Ec
ono
mic
Eve
...
Cle
anO
ut
«Ec
ono
mic
Eve
nt»
Fis
hS
ale
«Ec
ono
mic
Eve
...
Pay
me
nt
«Ec
ono
mic
Re
sou
rc..
.
Fis
h
«Ec
ono
mic
Eve
nt»
Ca
sh
«du
alit
y»
giv
e
«st
ock
-flo
w»
«du
alit
y»
«du
alit
y»
tak
e
«st
ock
-flo
w»
«du
alit
y»
giv
e
«st
ock
-flo
w»
tak
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
«du
alit
y»
giv
e
«st
ock
-flo
w»
«du
alit
y»
giv
e
«st
ock
-flo
w»
tak
e
«st
ock
-flo
w»
tak
e
«st
ock
-flo
w»
tak
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
giv
e
«st
ock
-flo
w»
tak
e
«st
ock
-flo
w»
Fig. 2. REA Sy's Fish Example
The value chain of Sy's Fish (i.e. the entrepreneurial script) comprises the
processes payroll, truck acquisition and store processing. The last process is
further decomposed into the processes buying, cleaning, and selling. The process
Page 6
level speci cation is provided in 2 as a UML class diagram. It depicts the object
constellations for the entire value chain. Due to space limitations, economic
agents have been omitted from the process level speci cation. The process level
speci cation with its sequence of duality relationships determines the rationale
of how Sy's makes money. Initial outlays (cash disbursements) are followed by
subsequent cash
ows. These cash
ows can be used to pay creditors and interests
or to nance further economic activities.
2.3 Limitations of the REA Ontology
A rst limitation of REA is given with regard to the clarity of ontological state-
ments. Although the REA model has evolved into a domain ontology, REA leaves
space for diverging interpretations of the relationships between core concepts.
In particular, the multiplicities for individual relationships are not clearly de-
ned. It remains unclear, whether a resource increment or decrement is caused
by one event or by multiple events. Moreover, it is not clear if an event can
be related to multiple stock-
ows and agents. Furthermore, there is no axiom
restricting event-resource (stock-
ow) relationships. In its original speci cation,
REA would allow for simultaneously connecting both increment and decrement
stock-
ows with a single event. However, stock-
ows have to be compatible with
their associated events. A decrement event should only be related with decrement
stock-
ows and vice versa. Connecting an increment stock-
ow with a decrement
event would result in semantic inconsistencies.
These ambiguities and potential semantic inconsistencies may be due to a
lack of a dedicated speci cation language. In the current state no means have
been proposed to specify precisely how to relate the REA concepts and how to
enforce compliance of REA models with the axioms stated above. The compli-
ance of REA models can only be checked manually. Thus, domain experts are
responsible for including the REA pattern into particular enterprise information
architectures. There is no conceptual facility that enables the development of
interoperable REA models (a modeling language would help here).
A third limitation is the complexity of REA models which increases pro-
gressively even with very few processes to be modeled. Each process generally
includes eight entities which have to be modeled (two events, two resources, two
times an inside and an outside agent) [10]. A dedicated REA modeling notation
would help to overcome the complexity. However, proposing a modeling notation
necessitates resolving the inconsistencies and incompleteness of the ontological
statements. This is subject to the following sections.
3 The REA Domain Speci c Language
Given these limitations we started to develop a graphical domain speci c lan-
guage (DSL) for REA. We based the development of the REA DSL on The
Object Management Group's(OMG) metamodeling architecture called Meta-
Object Facility (MOF) [14]. MOF comes with a meta-meta model (M3 layer)
constellations for the entire value chain. Due to space limitations, economic
agents have been omitted from the process level speci cation. The process level
speci cation with its sequence of duality relationships determines the rationale
of how Sy's makes money. Initial outlays (cash disbursements) are followed by
subsequent cash
ows. These cash
ows can be used to pay creditors and interests
or to nance further economic activities.
2.3 Limitations of the REA Ontology
A rst limitation of REA is given with regard to the clarity of ontological state-
ments. Although the REA model has evolved into a domain ontology, REA leaves
space for diverging interpretations of the relationships between core concepts.
In particular, the multiplicities for individual relationships are not clearly de-
ned. It remains unclear, whether a resource increment or decrement is caused
by one event or by multiple events. Moreover, it is not clear if an event can
be related to multiple stock-
ows and agents. Furthermore, there is no axiom
restricting event-resource (stock-
ow) relationships. In its original speci cation,
REA would allow for simultaneously connecting both increment and decrement
stock-
ows with a single event. However, stock-
ows have to be compatible with
their associated events. A decrement event should only be related with decrement
stock-
ows and vice versa. Connecting an increment stock-
ow with a decrement
event would result in semantic inconsistencies.
These ambiguities and potential semantic inconsistencies may be due to a
lack of a dedicated speci cation language. In the current state no means have
been proposed to specify precisely how to relate the REA concepts and how to
enforce compliance of REA models with the axioms stated above. The compli-
ance of REA models can only be checked manually. Thus, domain experts are
responsible for including the REA pattern into particular enterprise information
architectures. There is no conceptual facility that enables the development of
interoperable REA models (a modeling language would help here).
A third limitation is the complexity of REA models which increases pro-
gressively even with very few processes to be modeled. Each process generally
includes eight entities which have to be modeled (two events, two resources, two
times an inside and an outside agent) [10]. A dedicated REA modeling notation
would help to overcome the complexity. However, proposing a modeling notation
necessitates resolving the inconsistencies and incompleteness of the ontological
statements. This is subject to the following sections.
3 The REA Domain Speci c Language
Given these limitations we started to develop a graphical domain speci c lan-
guage (DSL) for REA. We based the development of the REA DSL on The
Object Management Group's(OMG) metamodeling architecture called Meta-
Object Facility (MOF) [14]. MOF comes with a meta-meta model (M3 layer)
Page 7
that allows us to de ne the structure, i.e. the abstract syntax, of the REA DSL
as a meta-model (M2 layer). The resulting REA DSL meta-model comprises
three interlinked views, of which we describe in detail the duality and the value
chain perspective. Due to space limitations, we do not elaborate in detail on the
third view on economic resources. However, it is important to note that economic
resources - scarce objects having utility and under control of an enterprise [15]
- may form a generalization hierarchy.
3.1 The Duality Model
The duality relationship is a core concept in REA. It links an increment in the
resource set with a corresponding decrement; where increments and decrements
must be members of two di erent event entity sets. In the REA ontology, the
duality relationship is characterized by the unary relationship assigned to the
concept of an economic event (see Figure 3(a)).
Duality
Transfer Transformation
nonConsumingTransformation ConsumingTransformation
Give Take Use Produce Consume
EntitySet
Economic Event StockFlow Participation
EconomicResource EconomicAgent
EconomicResourceSeries InsideAgent OutsideAgent
Economic EventSeries
Economic Resource
Economic ResourceSeries
nonConsuming Transformation Consuming Transformation
Stock Flow
Economic Agent
Inside Agent Outside Agent
artici ation
Transf ation
Entity Set
(a) Meta Model
E1
E2 E3 E4
R1
R2
R3
R4 R5
Inside1 Outside1
Inside2 Inside3 Outside1 Inside4 Inside5 Outside1 Inside6 Outside1
(b) Example
Fig. 3. Duality
In the REA DSL, duality becomes a core model element that serves as a
building block for further purposes. The duality meta model is depicted in Figure
3(a). Figure 3(b) presents a (rather abstract) example model - which should help
understanding the meta model concepts and to introduce the concrete syntax
and the corresponding stencils.
The duality concept applies to transfers (exchanges with external actors) and
transformations (value creation inside the enterprise). In the case of transforma-
tions we distinguish between resource-consuming and non-resource consuming
transformation. As a consequence, the meta model de nes consuming transfor-
mation and non-consuming transformation as special kinds of transformation as
well as transfer and transformation as specializations of duality.
as a meta-model (M2 layer). The resulting REA DSL meta-model comprises
three interlinked views, of which we describe in detail the duality and the value
chain perspective. Due to space limitations, we do not elaborate in detail on the
third view on economic resources. However, it is important to note that economic
resources - scarce objects having utility and under control of an enterprise [15]
- may form a generalization hierarchy.
3.1 The Duality Model
The duality relationship is a core concept in REA. It links an increment in the
resource set with a corresponding decrement; where increments and decrements
must be members of two di erent event entity sets. In the REA ontology, the
duality relationship is characterized by the unary relationship assigned to the
concept of an economic event (see Figure 3(a)).
Duality
Transfer Transformation
nonConsumingTransformation ConsumingTransformation
Give Take Use Produce Consume
EntitySet
Economic Event StockFlow Participation
EconomicResource EconomicAgent
EconomicResourceSeries InsideAgent OutsideAgent
Economic EventSeries
Economic Resource
Economic ResourceSeries
nonConsuming Transformation Consuming Transformation
Stock Flow
Economic Agent
Inside Agent Outside Agent
artici ation
Transf ation
Entity Set
(a) Meta Model
E1
E2 E3 E4
R1
R2
R3
R4 R5
Inside1 Outside1
Inside2 Inside3 Outside1 Inside4 Inside5 Outside1 Inside6 Outside1
(b) Example
Fig. 3. Duality
In the REA DSL, duality becomes a core model element that serves as a
building block for further purposes. The duality meta model is depicted in Figure
3(a). Figure 3(b) presents a (rather abstract) example model - which should help
understanding the meta model concepts and to introduce the concrete syntax
and the corresponding stencils.
The duality concept applies to transfers (exchanges with external actors) and
transformations (value creation inside the enterprise). In the case of transforma-
tions we distinguish between resource-consuming and non-resource consuming
transformation. As a consequence, the meta model de nes consuming transfor-
mation and non-consuming transformation as special kinds of transformation as
well as transfer and transformation as specializations of duality.
Page 8
No matter which kind of specialization it is, a duality always covers two dis-
tinct entity sets; one describing the increments in the resource sets and the other
one the decrements in the resource sets. In the case of a transfer the decrement is
called give and the increment is called take. The decrement of a non-consuming
transformation is denoted as use and the one of a consuming transformation
is denoted as consume. In both kinds of transformation the increment is called
produce.
Figure 3(b) shows an abstract example of a duality model to illustrate the
concepts and their stencils. An entity set is modeled as a partition. Accordingly,
a duality model includes two partitions. The duality shown in the example is a
transfer. It follows that one entity set is a give and the other one is a take. The
kind of entity set is denoted in the upper left corner of the partition.
According to the meta model, an entity set covers at least one but up to
multiple economic events. An economic event is considered as a class of phe-
nomena re
ecting changes in scare means [16]. An economic event is speci c
to the entity set it belongs to. Following the principles of duality, all economic
events in the decrement entity set (give, use, consume) are counterbalanced by
the economic events in the corresponding increment entity set (take, produce)
of the same duality model.
An economic event is usually executed at a certain point in time. However, in
certain cases the increment/decrement is realized by a series of economic events
each of the same nature. For example, consider that the payment for goods is
split into a number of partial payments. For this purpose we use the concept of an
economic event series, which is de ned as a specialization of the economic event
in the meta model. Consequently, the concept of an economic event series may
substitute an economic event. This means, instead of modeling each economic
event of each partial payment, one may model a single economic event series of
partial payments.
An economic event in a give/use/consume entity set decrements resource(s).
Similarly, an economic event in a take/produce entity set increments resource(s).
This relationship between an economic event and an economic resource is de-
scribed by the concept of a stock-
ow. A stock-
ow models an association be-
tween exactly one economic event and exactly one economic resource. An eco-
nomic event will a ect most of the time one economic resource only, but it
may a ect multiple ones. Thus, an economic event may have one up to many
stock-
ows connected. An economic resource usually is a ected by many di er-
ent economic events (in di erent entity sets of di erent duality models). At a
minimum an economic resource is a ected by one economic event - otherwise it
would not be worth considering the economic resource at all. Consequently, an
economic resource is connected to one up to many stock-
ows.
A similar concept to economic event series is de ned for economic resources:
the economic resource series. An economic resource series is used if an economic
event a ects a number of economic resources of the same kind. For example, on a
high level of abstraction one may de ne raw material as an economic resource and
an economic event may a ect a number of economic resources. It is important
tinct entity sets; one describing the increments in the resource sets and the other
one the decrements in the resource sets. In the case of a transfer the decrement is
called give and the increment is called take. The decrement of a non-consuming
transformation is denoted as use and the one of a consuming transformation
is denoted as consume. In both kinds of transformation the increment is called
produce.
Figure 3(b) shows an abstract example of a duality model to illustrate the
concepts and their stencils. An entity set is modeled as a partition. Accordingly,
a duality model includes two partitions. The duality shown in the example is a
transfer. It follows that one entity set is a give and the other one is a take. The
kind of entity set is denoted in the upper left corner of the partition.
According to the meta model, an entity set covers at least one but up to
multiple economic events. An economic event is considered as a class of phe-
nomena re
ecting changes in scare means [16]. An economic event is speci c
to the entity set it belongs to. Following the principles of duality, all economic
events in the decrement entity set (give, use, consume) are counterbalanced by
the economic events in the corresponding increment entity set (take, produce)
of the same duality model.
An economic event is usually executed at a certain point in time. However, in
certain cases the increment/decrement is realized by a series of economic events
each of the same nature. For example, consider that the payment for goods is
split into a number of partial payments. For this purpose we use the concept of an
economic event series, which is de ned as a specialization of the economic event
in the meta model. Consequently, the concept of an economic event series may
substitute an economic event. This means, instead of modeling each economic
event of each partial payment, one may model a single economic event series of
partial payments.
An economic event in a give/use/consume entity set decrements resource(s).
Similarly, an economic event in a take/produce entity set increments resource(s).
This relationship between an economic event and an economic resource is de-
scribed by the concept of a stock-
ow. A stock-
ow models an association be-
tween exactly one economic event and exactly one economic resource. An eco-
nomic event will a ect most of the time one economic resource only, but it
may a ect multiple ones. Thus, an economic event may have one up to many
stock-
ows connected. An economic resource usually is a ected by many di er-
ent economic events (in di erent entity sets of di erent duality models). At a
minimum an economic resource is a ected by one economic event - otherwise it
would not be worth considering the economic resource at all. Consequently, an
economic resource is connected to one up to many stock-
ows.
A similar concept to economic event series is de ned for economic resources:
the economic resource series. An economic resource series is used if an economic
event a ects a number of economic resources of the same kind. For example, on a
high level of abstraction one may de ne raw material as an economic resource and
an economic event may a ect a number of economic resources. It is important
Page 9
to note that an economic resource series may substitute an economic resource
in a stock-
ow, but the economic resource series must not be used in economic
resource generalization hierarchies. This is prohibited by a corresponding OCL
constraint to the meta model. An economic resource series must always be based
on exactly one existing economic resource. For an economic resource one may
de ne zero to many economic resource series (used in di erent stock-
ows).
An economic event involves economic agents. We distinguish between outside
agents, i.e. trading partners outside the company, and inside agents who are ac-
countable inside the company. The involvement of economic agents in economic
events is denoted by the concept of participation. A participation is an associ-
ation that connects exactly one economic event with one economic agent. An
economic event is associated to at least one, but up to many economic agents.
Hence, an economic event has one to many participation associations. An eco-
nomic agent participates in at least one, but up to many economic events (in the
same, but also in di erent entity sets of the same or di erent duality models).
Thus, an economic agent has one to many participations connected.
In addition, there are further constraints assigned to the meta model to
handle speci cs of transfers. In case of a transfer, each economic event must be
assigned to exactly one outside agent and, in addition, to at least one inside
agent. All economic events of the same transfer (both in the give and the take
entity set) must involve one and the same outside agent.
The relationships among economic events, economic resources, and economic
agents within an entity set is also illustrated in the abstract example of Figure
3(b). Economic events are denoted by an hexagon. In the give partition, there
is only one economic event E1. This economic event leads to a decrement of
the only associated economic resource R1, notated by a drop. Since the duality
model is a transfer, exactly one outside agent Outside 1 and an inside agent
Inside 1 is involved. The symbol for economic agents is a stickman - outside
agents have a black head, whereas inside agents have a white head.
The take partition includes three economic events E2, E3, and E4; one of
which (E3) is an economic event series. The symbol of an economic event series
is a pack of hexagons. All three economic events together compensate the single
economic event E1 of the give partition. E2 is associated with two economic
resources R2 and R3 that are incremented, whereas the series E3 increments
only the economic resource R4. E4 leads to an increase of the economic resource
series R5. An economic resource series is depicted by a pack of drops. Given the
example of a transfer, all three events in the take partition must be associated
with the same outside agent as the economic event of the give partition: Outside
1. In addition, E2 and E3 involve two inside agents and E4 only one.
3.2 The Value Chain
According to Geerts and McCarty [13], the duality relationships introduced in
the previous chapter are the glue that binds a company's economic events to-
gether into rational economic processes, while "stock-
ow" relationships weave
these processes together into an enterprise value chain. In the latter case, they
in a stock-
ow, but the economic resource series must not be used in economic
resource generalization hierarchies. This is prohibited by a corresponding OCL
constraint to the meta model. An economic resource series must always be based
on exactly one existing economic resource. For an economic resource one may
de ne zero to many economic resource series (used in di erent stock-
ows).
An economic event involves economic agents. We distinguish between outside
agents, i.e. trading partners outside the company, and inside agents who are ac-
countable inside the company. The involvement of economic agents in economic
events is denoted by the concept of participation. A participation is an associ-
ation that connects exactly one economic event with one economic agent. An
economic event is associated to at least one, but up to many economic agents.
Hence, an economic event has one to many participation associations. An eco-
nomic agent participates in at least one, but up to many economic events (in the
same, but also in di erent entity sets of the same or di erent duality models).
Thus, an economic agent has one to many participations connected.
In addition, there are further constraints assigned to the meta model to
handle speci cs of transfers. In case of a transfer, each economic event must be
assigned to exactly one outside agent and, in addition, to at least one inside
agent. All economic events of the same transfer (both in the give and the take
entity set) must involve one and the same outside agent.
The relationships among economic events, economic resources, and economic
agents within an entity set is also illustrated in the abstract example of Figure
3(b). Economic events are denoted by an hexagon. In the give partition, there
is only one economic event E1. This economic event leads to a decrement of
the only associated economic resource R1, notated by a drop. Since the duality
model is a transfer, exactly one outside agent Outside 1 and an inside agent
Inside 1 is involved. The symbol for economic agents is a stickman - outside
agents have a black head, whereas inside agents have a white head.
The take partition includes three economic events E2, E3, and E4; one of
which (E3) is an economic event series. The symbol of an economic event series
is a pack of hexagons. All three economic events together compensate the single
economic event E1 of the give partition. E2 is associated with two economic
resources R2 and R3 that are incremented, whereas the series E3 increments
only the economic resource R4. E4 leads to an increase of the economic resource
series R5. An economic resource series is depicted by a pack of drops. Given the
example of a transfer, all three events in the take partition must be associated
with the same outside agent as the economic event of the give partition: Outside
1. In addition, E2 and E3 involve two inside agents and E4 only one.
3.2 The Value Chain
According to Geerts and McCarty [13], the duality relationships introduced in
the previous chapter are the glue that binds a company's economic events to-
gether into rational economic processes, while "stock-
ow" relationships weave
these processes together into an enterprise value chain. In the latter case, they
Page 10
do not refer to the stock-
ows within a single duality model. Rather, they mean
that an increment in a resource as a result of one duality model will serve as a
chance to decrement this resource in a subsequent duality model. In other words,
a value chain is built by a well de ned series of duality models. The transitions
between the duality models re
ect the resource dependencies between the duality
models.
1..*
Economic Resource Value Chain
Business Process Transfer Business Event
Economic ResourceFlow Economic Resource
1
1
1
0..1 0..1 0..* 0..*
0..*
1..* 1..*
2..* +target +source +in +out
(a) Meta Model (b) Example
Fig. 4. Value Chain
These general ideas are re
ected in the value chain meta model of Figure 4(a).
A value chain is built by a number of business processes. A business process is
de ned as a transfer or a transformation, and in addition the tasks needed to
execute the transfer/transformation. It follows a business process points to an
underlying transfer/transformation described by a duality model. Furthermore,
a business process results in a number of business events, two of which are the
economic events of the underlying duality model.
A value chain includes one to many business processes. A business process
is used only once in one distinctive value chain. A business process points to
exactly one duality model. A duality model is usually the basis of one business
process, but may be referred to by multiple business processes. A business process
involves at least two business events, these are the two economic events (which
are considered as special kinds of business events). Usually, there will be more
business events associated to a business process as a result of the tasks needed
to execute the transfer/transformation.
Economic resources tie the business processes together. Thus, an economic
resource
ow - which points to exactly one economic resource - connects two
business processes. An economic resource
ow is a directed association that usu-
ally starts from a source business process and ends at a target business process.
However, we also allow for economic resource
ows that have either no source
business process or no target business process. This allows for a partial analysis,
when one considers a certain economic resource as given or when an economic
ows within a single duality model. Rather, they mean
that an increment in a resource as a result of one duality model will serve as a
chance to decrement this resource in a subsequent duality model. In other words,
a value chain is built by a well de ned series of duality models. The transitions
between the duality models re
ect the resource dependencies between the duality
models.
1..*
Economic Resource Value Chain
Business Process Transfer Business Event
Economic ResourceFlow Economic Resource
1
1
1
0..1 0..1 0..* 0..*
0..*
1..* 1..*
2..* +target +source +in +out
(a) Meta Model (b) Example
Fig. 4. Value Chain
These general ideas are re
ected in the value chain meta model of Figure 4(a).
A value chain is built by a number of business processes. A business process is
de ned as a transfer or a transformation, and in addition the tasks needed to
execute the transfer/transformation. It follows a business process points to an
underlying transfer/transformation described by a duality model. Furthermore,
a business process results in a number of business events, two of which are the
economic events of the underlying duality model.
A value chain includes one to many business processes. A business process
is used only once in one distinctive value chain. A business process points to
exactly one duality model. A duality model is usually the basis of one business
process, but may be referred to by multiple business processes. A business process
involves at least two business events, these are the two economic events (which
are considered as special kinds of business events). Usually, there will be more
business events associated to a business process as a result of the tasks needed
to execute the transfer/transformation.
Economic resources tie the business processes together. Thus, an economic
resource
ow - which points to exactly one economic resource - connects two
business processes. An economic resource
ow is a directed association that usu-
ally starts from a source business process and ends at a target business process.
However, we also allow for economic resource
ows that have either no source
business process or no target business process. This allows for a partial analysis,
when one considers a certain economic resource as given or when an economic
Page 11
resource is considered as nal output of the value chain. Typically, cash is often
assigned to such economic resource
ows.
A business process has at least one, but up to many outgoing economic re-
source
ows. Similarly, a business process has at least one, but up to many
ingoing economic resource
ows. In order to deliver a consistent value chain
two important constraints on the business processes have to be considered: For
each economic resource in the give / use / consume entity set of the underly-
ing duality model, at least one corresponding incoming economic resource
ow
pointing to the same economic resource must exist. In analogy, for each eco-
nomic resource in the take / produce entity set of the underlying duality model,
at least one corresponding outgoing economic resource
ow pointing to the same
economic resource must exist. However, one may consider the substitutability
concept of more general and more speci c economic resources as de ned in a
resource generalization hierarchy. In other words, a duality model expecting an
economic resource in give / use / consume entity set may also accept a more
speci c economic resource in the incoming economic resource
ow.
In Figure 4(b) we depict an abstract example of a value chain to illustrate
its concepts and their stencils. Our value chain example includes four business
processes BP1, BP2, BP3, and BP4. The symbol for a business process is a
rounded rectangle. Economic resource
ows are depicted by an arrow with the
economic resource assigned to this arrow. For example, the economic resource R2
ows from BP1 to BP3. However, for esthetic reasons we provide an alternative
(but still similar) notation for economic resource
ows that do not start from
or do not end in a business process. In this case, the arrow may start from the
economic resource or may lead to the economic resource, instead of assigning the
economic resource to the arrow. Examples are the resource
ows of R1 leading
into BP1 and of R8 starting from BP4.
BP1 points to the duality model of Figure 3(b). In this duality model the
economic resource R1 sits in the give partition; and the economic resources R2,
R3, R4, and R5 - where the latter is a economic resources series - are included in
the take partition. This is consistent with the economic resource
ows to/from
BP1 in the value chain model of Figure 4(b). BP1 receives R1 and delivers
R2, R3, R4, and R5. It should be noted that the number of ingoing/outgoing
transitions does not necessarily meet the number of economic resources in the
duality model, since a business process may provide an economic resource or -
more unlikely in practice - receive a resource from multiple business processes.
For example, BP2 provides R6 to both BP3 and BP4.
Let us assume that the duality model on which BP2 is based requires the
economic resources R3, R5, and R7. In this case one might think of an incon-
sistency since BP2 receives R3 and R5, but R4 instead of R7. However, if R4
is de ned as a specialization of R7 in a resource generalization hierarchy, the
model is consistent since in this case R4 may substitute R7 in the duality model
of BP2.
assigned to such economic resource
ows.
A business process has at least one, but up to many outgoing economic re-
source
ows. Similarly, a business process has at least one, but up to many
ingoing economic resource
ows. In order to deliver a consistent value chain
two important constraints on the business processes have to be considered: For
each economic resource in the give / use / consume entity set of the underly-
ing duality model, at least one corresponding incoming economic resource
ow
pointing to the same economic resource must exist. In analogy, for each eco-
nomic resource in the take / produce entity set of the underlying duality model,
at least one corresponding outgoing economic resource
ow pointing to the same
economic resource must exist. However, one may consider the substitutability
concept of more general and more speci c economic resources as de ned in a
resource generalization hierarchy. In other words, a duality model expecting an
economic resource in give / use / consume entity set may also accept a more
speci c economic resource in the incoming economic resource
ow.
In Figure 4(b) we depict an abstract example of a value chain to illustrate
its concepts and their stencils. Our value chain example includes four business
processes BP1, BP2, BP3, and BP4. The symbol for a business process is a
rounded rectangle. Economic resource
ows are depicted by an arrow with the
economic resource assigned to this arrow. For example, the economic resource R2
ows from BP1 to BP3. However, for esthetic reasons we provide an alternative
(but still similar) notation for economic resource
ows that do not start from
or do not end in a business process. In this case, the arrow may start from the
economic resource or may lead to the economic resource, instead of assigning the
economic resource to the arrow. Examples are the resource
ows of R1 leading
into BP1 and of R8 starting from BP4.
BP1 points to the duality model of Figure 3(b). In this duality model the
economic resource R1 sits in the give partition; and the economic resources R2,
R3, R4, and R5 - where the latter is a economic resources series - are included in
the take partition. This is consistent with the economic resource
ows to/from
BP1 in the value chain model of Figure 4(b). BP1 receives R1 and delivers
R2, R3, R4, and R5. It should be noted that the number of ingoing/outgoing
transitions does not necessarily meet the number of economic resources in the
duality model, since a business process may provide an economic resource or -
more unlikely in practice - receive a resource from multiple business processes.
For example, BP2 provides R6 to both BP3 and BP4.
Let us assume that the duality model on which BP2 is based requires the
economic resources R3, R5, and R7. In this case one might think of an incon-
sistency since BP2 receives R3 and R5, but R4 instead of R7. However, if R4
is de ned as a specialization of R7 in a resource generalization hierarchy, the
model is consistent since in this case R4 may substitute R7 in the duality model
of BP2.
Page 12
3.3 REA DSL Example
Having introduced the meta models and rather abstract examples, we now
demonstrate our REA DSL by means of a simple, but more realistic example.
For this purpose we use again the Sy's Fish example [13] of subsection 2.2. The
resulting Sy's Fish value chain and duality models are depicted in Figure 5.
PayrollProcess Buying
TruckAcquisition Transport Cleaning Selling
CashDisbursementLabor
LaborAquisition
PaymentForFish
FishObtained
PaymentForTruck
TruckLease
TransportIn
TransportOut
CleanIn
CleanOut
FishSale
Payment
Cash Cash
Cash
Cash
Labor
Labor
Labor
Labor
Labor
Fish
Fish Fish
Fish
Fish Fish Truck
Truck
Logistics Truck Rental
Cashier Truck Rental
Cashier Cashier Worker
Worker
Fisherman
Fisherman Purchaser Human Resource
Driver
Driver
Cleaner
Cleaner Cashier
Salesman Customer
Customer
Fig. 5. Sy's Fish Value Chain and Dualities
The value chain begins with the payroll process which receives some cash
to realize a transfer. By the duality relationship of the economic events cash
disbursement and labor acquisition, the economic resource cash is decre-
mented and labor is incremented. The resulting labor is then input for the busi-
ness processes buying, transport, cleaning, and selling.
Buying transfers labor and additionally cash to fish, by the dual economic
events payment for fish and fish obtained. Truck acquisition is another
transfer that turns cash into a truck that is leased by the duality of payment
for truck and truck lease.
Transport uses all acquired resource - labor, fish, and truck - to deliver the
sh to the company. The transport is done by the company itself, so it does not
involve an outside actor. Accordingly, it is a transformation process that uses,
but does not consume its economic resources in the economic event transport
in. The dual economic event transport out results again in the fish, this time
Having introduced the meta models and rather abstract examples, we now
demonstrate our REA DSL by means of a simple, but more realistic example.
For this purpose we use again the Sy's Fish example [13] of subsection 2.2. The
resulting Sy's Fish value chain and duality models are depicted in Figure 5.
PayrollProcess Buying
TruckAcquisition Transport Cleaning Selling
CashDisbursementLabor
LaborAquisition
PaymentForFish
FishObtained
PaymentForTruck
TruckLease
TransportIn
TransportOut
CleanIn
CleanOut
FishSale
Payment
Cash Cash
Cash
Cash
Labor
Labor
Labor
Labor
Labor
Fish
Fish Fish
Fish
Fish Fish Truck
Truck
Logistics Truck Rental
Cashier Truck Rental
Cashier Cashier Worker
Worker
Fisherman
Fisherman Purchaser Human Resource
Driver
Driver
Cleaner
Cleaner Cashier
Salesman Customer
Customer
Fig. 5. Sy's Fish Value Chain and Dualities
The value chain begins with the payroll process which receives some cash
to realize a transfer. By the duality relationship of the economic events cash
disbursement and labor acquisition, the economic resource cash is decre-
mented and labor is incremented. The resulting labor is then input for the busi-
ness processes buying, transport, cleaning, and selling.
Buying transfers labor and additionally cash to fish, by the dual economic
events payment for fish and fish obtained. Truck acquisition is another
transfer that turns cash into a truck that is leased by the duality of payment
for truck and truck lease.
Transport uses all acquired resource - labor, fish, and truck - to deliver the
sh to the company. The transport is done by the company itself, so it does not
involve an outside actor. Accordingly, it is a transformation process that uses,
but does not consume its economic resources in the economic event transport
in. The dual economic event transport out results again in the fish, this time
Page 13
at the right place. Cleaning is another non-consuming transformation process
that receives the fish and turns it into a cleaned fish.
The nal transfer in the value chain is selling. Labor and fish are given
in the fish sale in order to take cash as result of the payment.
4 REA DSL Tool Support
Having developed our DSL approach towards REA modeling in theory, we wanted
to evaluate our approach by practical means. Thereby, we focused on three major
steps. Firstly, we evaluated the technical feasibility of the DSL by an implemen-
tation based on Microsoft DSL tool kits. Thereby, we were able to eliminate
technical
aws in the meta model. Secondly, we wanted to test our tool if it
properly supports existing REA models. For this purpose we had 32 REA mod-
els available and successfully completed the reengineering task to represent these
models within our REA DSL. During our reengineering activities the strengths
of our rather strict meta model became evident - we were able to recognize
aws
in the existing REA models which have not been recognized before due to the
complexity in the ontology representation and missing tool support. Thirdly,
we approached the originator of REA - William McCarthy - to seek his advise
and to report inconsistencies in existing REA models. McCarthy sanctioned our
DSL approach and provided valuable comments which we integrated into our
approach. However, so far we have not yet done any usability studies, nor have
we used the tool set in real world case studies. We will conduct these kinds of
evaluations once we extend our approach to cover the REA policy infrastructure.
We implemented the graphical REA-DSL tool based on Microsoft's Visual
Studio 2010 Visualization & Modeling SDK (V&M SDK). Accordingly we used
the V&M SDK to create the meta models explained before in Section 3. Addi-
tional custom code enables to set further constraints, necessary for the validation
of the REA model. In a second step the designer - see Figure 6 - is created to
support the REA modeling.
AB C
D
E
Fig. 6. REA-designer
that receives the fish and turns it into a cleaned fish.
The nal transfer in the value chain is selling. Labor and fish are given
in the fish sale in order to take cash as result of the payment.
4 REA DSL Tool Support
Having developed our DSL approach towards REA modeling in theory, we wanted
to evaluate our approach by practical means. Thereby, we focused on three major
steps. Firstly, we evaluated the technical feasibility of the DSL by an implemen-
tation based on Microsoft DSL tool kits. Thereby, we were able to eliminate
technical
aws in the meta model. Secondly, we wanted to test our tool if it
properly supports existing REA models. For this purpose we had 32 REA mod-
els available and successfully completed the reengineering task to represent these
models within our REA DSL. During our reengineering activities the strengths
of our rather strict meta model became evident - we were able to recognize
aws
in the existing REA models which have not been recognized before due to the
complexity in the ontology representation and missing tool support. Thirdly,
we approached the originator of REA - William McCarthy - to seek his advise
and to report inconsistencies in existing REA models. McCarthy sanctioned our
DSL approach and provided valuable comments which we integrated into our
approach. However, so far we have not yet done any usability studies, nor have
we used the tool set in real world case studies. We will conduct these kinds of
evaluations once we extend our approach to cover the REA policy infrastructure.
We implemented the graphical REA-DSL tool based on Microsoft's Visual
Studio 2010 Visualization & Modeling SDK (V&M SDK). Accordingly we used
the V&M SDK to create the meta models explained before in Section 3. Addi-
tional custom code enables to set further constraints, necessary for the validation
of the REA model. In a second step the designer - see Figure 6 - is created to
support the REA modeling.
AB C
D
E
Fig. 6. REA-designer
Page 14
The designer is separated into ve major areas: The modeling canvas (A) the
toolbox (B), the solution explorer (C), the property window (D) and the valida-
tion window (E). By dragging the modeling elements from the toolbox on the
modeling canvas, a REA model can be assembled in a graphical representation.
The solution explorer provides a tree based overview of the elements of the cur-
rently displayed model as well as a le and directory structure to hold di erent
model instances. Properties of the selected model element can be changed in the
property window. The validation window informs the user of any errors/warnings.
5 Summary and Future Work
In this paper we proposed a domain speci c modeling language to support the
conceptual modeling of economic events based on the REA ontology. Conceptual
models based on our modeling language { the REA-DSL { aim at facilitating the
requirements elicitation process during the design of accounting and enterprise
information systems. The original REA ontology has a long history in accounting
and is based on well grounded accounting concepts. However, REA leaves space
for diverging interpretations of the relationships between core concepts. This has
also been criticized by Gailly and Poels who have proposed a new conceptual
representation of REA guided by proven ontology engineering principles [17].
They come up with a presentation based on UML class diagrams, which we feel
results in a complex visual representation that is hardly understood by domain
experts (cf. Figure 2). Similar to Gailly and Poels, we developed a representation
format that does not leave space for divergent interpretations. In our case, the
relationships between the core concepts are precisely de ned by using OMG's
Meta Object Facility (MOF) leading to a dedicated REA domain speci c model-
ing language. The MOF-based approach enables the development of a graphical
syntax that is dedicated to the needs of business modeling. Furthermore, our
REA-DSL comes with a graphical syntax covering a set of stencils that facilitate
the understanding of the domain expert. Beside proposing a meta model and a
graphical language, we developed a REA-DSL tool as a proof of concept.
In this work, we concentrated on the basic REA principles [4] and the value
chain perspective [13]. In future work, we will gradually extend the REA-DSL.
Currently, we are working on an integration of the REA policy infrastructure
[18] covering commitments, agreements and, furthermore, the typi cation of the
operational concepts [5]. Next, we plan to extend the REA-DSL by concepts to
derive a database design for enterprise information systems, which has been one
of the original goals of REA. For this purpose, we have already introduced the
concepts of economic event series and economic resource series in the current
DSL, because they will a ect the multiplicities in the database design. By in-
cluding the policy infrastructure and the REA-driven database design, it will
become more obvious that REA models are of structural nature rather and do
not concentrate on the behavioral aspects of process models. Another intended
REA-DSL extension addresses the perspective from which the REA models are
described. Currently, we focus on the perspective of a single enterprise which ex-
toolbox (B), the solution explorer (C), the property window (D) and the valida-
tion window (E). By dragging the modeling elements from the toolbox on the
modeling canvas, a REA model can be assembled in a graphical representation.
The solution explorer provides a tree based overview of the elements of the cur-
rently displayed model as well as a le and directory structure to hold di erent
model instances. Properties of the selected model element can be changed in the
property window. The validation window informs the user of any errors/warnings.
5 Summary and Future Work
In this paper we proposed a domain speci c modeling language to support the
conceptual modeling of economic events based on the REA ontology. Conceptual
models based on our modeling language { the REA-DSL { aim at facilitating the
requirements elicitation process during the design of accounting and enterprise
information systems. The original REA ontology has a long history in accounting
and is based on well grounded accounting concepts. However, REA leaves space
for diverging interpretations of the relationships between core concepts. This has
also been criticized by Gailly and Poels who have proposed a new conceptual
representation of REA guided by proven ontology engineering principles [17].
They come up with a presentation based on UML class diagrams, which we feel
results in a complex visual representation that is hardly understood by domain
experts (cf. Figure 2). Similar to Gailly and Poels, we developed a representation
format that does not leave space for divergent interpretations. In our case, the
relationships between the core concepts are precisely de ned by using OMG's
Meta Object Facility (MOF) leading to a dedicated REA domain speci c model-
ing language. The MOF-based approach enables the development of a graphical
syntax that is dedicated to the needs of business modeling. Furthermore, our
REA-DSL comes with a graphical syntax covering a set of stencils that facilitate
the understanding of the domain expert. Beside proposing a meta model and a
graphical language, we developed a REA-DSL tool as a proof of concept.
In this work, we concentrated on the basic REA principles [4] and the value
chain perspective [13]. In future work, we will gradually extend the REA-DSL.
Currently, we are working on an integration of the REA policy infrastructure
[18] covering commitments, agreements and, furthermore, the typi cation of the
operational concepts [5]. Next, we plan to extend the REA-DSL by concepts to
derive a database design for enterprise information systems, which has been one
of the original goals of REA. For this purpose, we have already introduced the
concepts of economic event series and economic resource series in the current
DSL, because they will a ect the multiplicities in the database design. By in-
cluding the policy infrastructure and the REA-driven database design, it will
become more obvious that REA models are of structural nature rather and do
not concentrate on the behavioral aspects of process models. Another intended
REA-DSL extension addresses the perspective from which the REA models are
described. Currently, we focus on the perspective of a single enterprise which ex-
Page 15
changes value with other enterprises. In the Open-edi reference model (ISO/IEC
15944-4) REA concepts are used to describe the exchanges of value among en-
terprises from a neutral observer's point of view. We plan to integrate the ob-
server's perspective into our REA-DSL and to support semi-automatic mappings
between the perspectives.
References
1. Andersson, B., Bergholtz, M., Edirisuriya, A., Ilayperuma, T., Johannesson, P.,
Gordijn, J., Gregoire, B., Schmitt, M., Dubois, E., Abels, S., Hahn, A., Wangler,
B., Weigand, H.: Towards a Reference Ontology for Business Models. In: Proc. of
the 25th Int. Conf. on Conceptual Modeling, Springer LNCS (2006) 482{496
2. Osterwalder, A., Pigneur, Y., Tucci, C.L.: Clarifying Business Models: Origins,
Present and Future of the Concept. Communications of the Association for Infor-
mation Science (CAIS) 15 (2005) 751{775
3. Gordijn, J., Akkermans, H.: E3-value: Designing and evaluating e-business models.
IEEE Intelligent Systems 16(4) (Jul{Aug 2001) 11{17
4. McCarthy, W.E.: The REA Accounting Model: A Generalized Framework for
Accounting Systems in a Shared Data Environment. The Accounting Review 57(3)
(1982)
5. Geerts, G.L., McCarthy, W.E.: An ontological analysis of the economic primitives
of the extended-rea enterprise information architecture. International Journal of
Accounting Information Systems 3(1) (2002) 1 { 16
6. van Deursen, A., Klint, P., Visser, J.: Domain-speci c languages: an annotated
bibliography. SIGPLAN Not. 35(6) (2000) 26{36
7. Kelly, S., Tolvanen, J.P.: Domain-Speci c Modeling. Wiley-IEEE Computer Soci-
ety Press (2008)
8. Green eld, J., Short, K., Cook, S., Kent, S.: Software Factories. John Wiley (2008)
9. Strembeck, M., Zdun, U.: An approach for the systematic development of domain-
speci c languages. Softw., Pract. Exper. 39(15) (2009) 1253{1292
10. Geerts, G.L., McCarthy, W.E.: An Accounting Object Infrastructure for
Knowledge-Based Enterprise Models. IEEE Intelligent Systems 14(4) (1999) 89{94
11. Geerts, G.L., McCarthy, W.E.: The Ontological Foundations of REA Enterprise
Systems (August 2000)
12. Motal, T., Schuster, R.: From e3-value to REA: Modeling multi-party eBusiness
Collaborations. In: Proc. of the 11th IEEE Conference on Commerce and Enter-
prise Computing, IEEE CS (2009) 202{208
13. Geerts, G.L., , McCarthy, W.E.: Modeling business enterprises as value-added
process hierarchies with resource-event-agent object templates. In: In Business
Object Design and Implementation, Springer-Verlag (1997) 94{113
14. OMG: Meta Object Facility (MOF) Core Speci cation, Version 2.0 (January 2006)
15. Ijiri, Y.: Theory of accounting measurement. American Accounting Association,
Sarasota, Fla (1975)
16. Yu, S.C.: The Structure of Accounting Theory. The University Presses of Florida
(1976)
17. Gailly, F., Poels, G.: Ontology-driven business modelling: Improving the conceptual
representation of the rea ontology. In: Conceptual Modeling - ER 2007, Springer
LNCS (2007) 407{422
18. Geerts, G.L., , McCarthy, W.E.: Policy-level speci cation in rea enterprise infor-
mation systems. Journal of Information Systems 20(2) (2006) 37{63
15944-4) REA concepts are used to describe the exchanges of value among en-
terprises from a neutral observer's point of view. We plan to integrate the ob-
server's perspective into our REA-DSL and to support semi-automatic mappings
between the perspectives.
References
1. Andersson, B., Bergholtz, M., Edirisuriya, A., Ilayperuma, T., Johannesson, P.,
Gordijn, J., Gregoire, B., Schmitt, M., Dubois, E., Abels, S., Hahn, A., Wangler,
B., Weigand, H.: Towards a Reference Ontology for Business Models. In: Proc. of
the 25th Int. Conf. on Conceptual Modeling, Springer LNCS (2006) 482{496
2. Osterwalder, A., Pigneur, Y., Tucci, C.L.: Clarifying Business Models: Origins,
Present and Future of the Concept. Communications of the Association for Infor-
mation Science (CAIS) 15 (2005) 751{775
3. Gordijn, J., Akkermans, H.: E3-value: Designing and evaluating e-business models.
IEEE Intelligent Systems 16(4) (Jul{Aug 2001) 11{17
4. McCarthy, W.E.: The REA Accounting Model: A Generalized Framework for
Accounting Systems in a Shared Data Environment. The Accounting Review 57(3)
(1982)
5. Geerts, G.L., McCarthy, W.E.: An ontological analysis of the economic primitives
of the extended-rea enterprise information architecture. International Journal of
Accounting Information Systems 3(1) (2002) 1 { 16
6. van Deursen, A., Klint, P., Visser, J.: Domain-speci c languages: an annotated
bibliography. SIGPLAN Not. 35(6) (2000) 26{36
7. Kelly, S., Tolvanen, J.P.: Domain-Speci c Modeling. Wiley-IEEE Computer Soci-
ety Press (2008)
8. Green eld, J., Short, K., Cook, S., Kent, S.: Software Factories. John Wiley (2008)
9. Strembeck, M., Zdun, U.: An approach for the systematic development of domain-
speci c languages. Softw., Pract. Exper. 39(15) (2009) 1253{1292
10. Geerts, G.L., McCarthy, W.E.: An Accounting Object Infrastructure for
Knowledge-Based Enterprise Models. IEEE Intelligent Systems 14(4) (1999) 89{94
11. Geerts, G.L., McCarthy, W.E.: The Ontological Foundations of REA Enterprise
Systems (August 2000)
12. Motal, T., Schuster, R.: From e3-value to REA: Modeling multi-party eBusiness
Collaborations. In: Proc. of the 11th IEEE Conference on Commerce and Enter-
prise Computing, IEEE CS (2009) 202{208
13. Geerts, G.L., , McCarthy, W.E.: Modeling business enterprises as value-added
process hierarchies with resource-event-agent object templates. In: In Business
Object Design and Implementation, Springer-Verlag (1997) 94{113
14. OMG: Meta Object Facility (MOF) Core Speci cation, Version 2.0 (January 2006)
15. Ijiri, Y.: Theory of accounting measurement. American Accounting Association,
Sarasota, Fla (1975)
16. Yu, S.C.: The Structure of Accounting Theory. The University Presses of Florida
(1976)
17. Gailly, F., Poels, G.: Ontology-driven business modelling: Improving the conceptual
representation of the rea ontology. In: Conceptual Modeling - ER 2007, Springer
LNCS (2007) 407{422
18. Geerts, G.L., , McCarthy, W.E.: Policy-level speci cation in rea enterprise infor-
mation systems. Journal of Information Systems 20(2) (2006) 37{63
Sign up today - FREE
Mendeley saves you time finding and organizing research. Learn more
- All your research in one place
- Add and import papers easily
- Access it anywhere, anytime
Start using Mendeley in seconds!
Readership Statistics
5 Readers on Mendeley
by Discipline
by Academic Status
40% Ph.D. Student
20% Student (Master)
20% Researcher (at an Academic Institution)
by Country
20% Italy
20% Algeria
20% Austria