Repurposing Zachman Framework Principles for
"Enterprise Model"-Driven Engineering
Alisa Harkai, Mihai Cinpoeru and Robert Andrei Buchmann
Business Informatics Research Center, Babeş-Bolyai University, str. T. Mihali 58-60, Cluj Napoca, Romania
Keywords: Enterprise Modelling, Enterprise Knowledge Engineering, Zachman Framework, Resource Description
Framework.
Abstract: The paper proposes an agile modelling tool which implements a domain-specific modelling method. As a
motivational starting point for the development of this modelling tool, we employ the Zachman Framework -
an ontology which conceptualises an enterprise across a variety of abstractions and facets. We conducted our
work with respect to the Zachman Framework in order to cover several of these facets and to suggest the
possibility of further employing Agile Modelling Method Engineering to extend this coverage, with the tool
providing the ability to create hyperlinks between models expressing different enterprise views. The agile
modelling tool developed as a proof-of-concept is further coupled with semantic technology to make models
available to semantic queries and machine reasoning in the context of model-driven software engineering.
1 INTRODUCTION
Motivated by project-based requirements (the
EnterKnow project (Enter Know, 2017)), we
analysed the Zachman Framework (ZF) as a
motivational starting point for implementing an
enterprise modelling method. Although ZF is
commonly presented as an ontology, it is not one in
the formal sense employed in ontology engineering
(as promoted in the works of (Smith, 2004) or
(Guarino, 1995)). Also, it is not truly a methodology
(Sessions, 2007), so it cannot be considered a fully-
fledged modelling method in the sense defined by
(Karagiannis and Kühn, 2002). It is a bi-dimensional
schema that can guide enterprise architects and
enterprise architecture managers in making decisions
based on definition, design and analysis of
architectural information for enterprises and their
information systems (Zachman, 1982).
Consequently, we aimed to investigate how ZF's
principles can be grounded in more formal and
actionable knowledge structures, with respect to
software engineering needs. More precisely, we aim
to derive machine-readable knowledge from a ZF-
guided enterprise architecture design, in a way that
can further support knowledge-driven software
development processes for Enterprise Information
Systems. Our investigation identified several
opportunities and paradigms developed during recent
years, whose interplay supports our goal: (i) the
Multi-perspective Enterprise Modelling paradigm
which advocates the description of an enterprise
through multiple models reflecting different facets of
the same enterprise (Kingston and Macintosh, 2000)
(Frank, 2000); (ii) the Agile Modelling Method
Engineering (AMME) framework (Karagiannis,
2015) which enables the full customization and quick
prototyping of multi-view modelling tools according
to situation-specific requirements – see an example in
(Bork, 2015); (iii) the Linked Open Models proposal
(Karagiannis and Buchmann, 2016) which advocates
the serialization of diagrammatic models in RDF
knowledge graphs (W3C, 2017a) in order to make
diagrammatic content available to running
knowledge-based systems outside a modelling
environment.
Based on this investigation, we advocate the
possibility of coupling the design thinking suggested
by ZF's conceptual frame with semantic technology
that can expose it to semantic queries and machine
reasoning. In support of this proposal, we developed
a proof-of-concept agile enterprise modelling tool
that reflects the multi-faceted nature of ZF while
establishing machine-readable semantic links
governed by an overarching meta-model to expose
the heterogeneous customised models as machine-
readable knowledge graphs. One current limitation
682
Harkai, A., Cinpoeru, M. and Buchmann, R.
Repurposing Zachman Framework Principles for "Enterprise Model"-Driven Engineering.
DOI: 10.5220/0006710706820689
In Proceedings of the 20th International Conference on Enterprise Information Systems (ICEIS 2018), pages 682-689
ISBN: 978-989-758-298-1
Copyright
c
2019 by SCITEPRESS Science and Technology Publications, Lda. All rights reserved
due to project requirements and limited scope is that
not all of ZF's perspectives have been deployed in the
proof-of-concept, only a selected subset correlated
with the requirements for a model-driven software
artefact. The software engineering method employed
to develop model-driven software artefacts based on
our tool is discussed in a different paper (Buchmann
et al., 2018).
This paper is structured as follows: Section 2
provides background information about ZF, the
AMME methodology and the RDF technological
space. Section 3 comments on related works. Section
4 discusses design decisions. Section 5 presents the
proof-of-concept agile enterprise modelling tool and
the final Section 6 provides conclusions based on a
SWOT analysis.
2 BACKGROUND,
METHODOLOGY AND
ENABLERS
2.1 Motivation: The Zachman
Framework
ZF is an enterprise ontology prescribing a
fundamental structure for Enterprise Architecture –
i.e., a formal and structured way of viewing and
defining an enterprise (Zachman, 2008). The
representation of this ontology is often in the form of
a 6x6 matrix, a two-dimensional schema used in order
to represent and organise the architectural layers and
artefacts comprising an enterprise system.
Figure 1 shows a representational form of this
framework in which the columns represent the ZF
facets and the rows represent the types of
stakeholders: (i) execution perspective (contextual
level); (ii) business management perspective
(conceptual level); (iii) architect perspective (logical
level); (iv) engineer perspective (physical level); (v)
technician perspective (as built); (vi) enterprise
perspective (functioning).
The core idea is that an asset can be described in
different ways for different purposes using those
levels of abstraction and views. This has also been
recognised in multi-view enterprise modelling (Bork,
2015) – the principle can be transferred to the practice
of Agile Modelling Method Engineering with the goal
of producing modelling tools that are customised to
capture, in a semantically integrated way, the facets
of ZF or enterprise assets pertaining to those facets.
Figure 1: Zachman’s Framework abstraction overview.
2.2 Methodology: The Agile Modelling
Method Engineering
Agile Modelling Method Engineering (Karagiannis,
2015) can be considered a Design Science (Peffers et
al., 2007) approach that is specialised for the
realisation of artefacts such as modelling methods and
modelling tools, regardless of application domain and
abstraction level. It advocates the ability of agilely
evolving the modelling semantics, syntax or
functionality with the help of meta-modelling
environments such as ADOxx (BOC Group, 2017). In
AMME, agility applies on two levels: (i) artefact
agility – which means that a modelling method is
decomposed in manageable building blocks of
various granularities (i.e., modelling language,
mechanisms and algorithms, modelling procedure)
and (ii) methodological agility which refers to the
actual engineering process - an incremental and
iterative development cycle in 5 steps: Create,
Design, Formalise, Develop and Deploy – see Figure
2, adapted from (Karagiannis, 2015) and subsequent
presentations (OMiLAB, 2017).
Figure 2: AMME lifecycle.
Nowadays, the notions of Agile Enterprise
(Goldman et al. 1994) and Agile Knowledge
Management (Levy and Hazzan, 2008) are commonly
adopted in management practices and Enterprise
Repurposing Zachman Framework Principles for "Enterprise Model"-Driven Engineering
683
Architecture Management could embrace a certain
level of agility in its model-based knowledge
repositories, one that can be synchronised with
software engineering processes. With the AMME
framework, agile modelling methods reflect the fact
that different purposes for different scopes must be
addressed due to the diversity of domains and
requirements. To this, the Linked Open Models vision
proposed in (Karagiannis and Buchmann, 2016) adds
the opportunity of exposing agile models to
knowledge-driven information systems, with the help
of the Resource Description Framework.
2.3 Technological Enablers: The
Resource Description Framework
The Resource Description Framework (RDF) (W3C,
2017a) is a World Wide Web Consortium standard
originally designed as a metadata model and evolved
as a technological foundation for the Semantic Web.
It can be used as a data model for conceptual
descriptions of Web resources in Linked Data
environments and coupled with ontologies.
The core idea of this data model is to capture
knowledge as statements about Web resources, in the
form of triples (subject-predicate-object). The subject
represents the resource and the predicate asserts a
relationship between the subject and the object or an
attribute whose value is the object. A collection of
RDF statements forms a directed multi-graph that can
be managed with the help of graph databases – e.g.
(Ontotext, 2017) - and queried through SPARQL
queries (W3C, 2017b). The Linked Open Models
vision (Karagiannis and Buchmann, 2016) introduced
several patterns and a plug-in for converting
diagrammatic models of arbitrary types to RDF - this
is employed in the work at hand for the ZF-driven
modelling tool hereby discussed.
3 RELATED WORKS
ZF is a well-established guide for structuring
enterprise knowledge. It was originally used for the
purpose of business system planning and it is
discussed under multiple interpretations by managers,
designers, programmers and other types of
stakeholders. This framework, designed as a schema,
prescribes facets and perspectives on enterprise
descriptions – however it does not specify the means
of bridging them in machine-oriented ways, and how
to further expose those bridges to external processing.
During the ’90s several frameworks extended ZF
- e.g., Evernden, The Integrated Architecture
Framework (Schekkerman, 2003). ZF is also used to
map various processes relevant to enterprise
architectures - e.g., analysis of the Rational United
Process (DJ de Villiers, 2001), Model-driven
architecture (Frankel et. al., 2003), TOGAF (The
Open Group, 2017a).
This paper's proposal tries to unify and combine
ZF, AMME and Linked Open Models towards the
goal of increasing the value and reusability of
enterprise architecture knowledge, opening the
possibility of interoperability with model-driven
information systems. AMME is used to partition with
appropriate granularity a custom-made modelling
language following the guidance of ZF's conceptual
frame and to agilely implement this in a modelling
tool. Then the Linked Open Models approach is
employed to make the resulting structure (and any
models built on it) available to semantic information
systems. The properties of enterprise artefacts
conceptualised in models can be navigated through a
dereferencing mechanism (originally presented in
(Cinpoeru, 2017) with application to BPMN) or
through SPARQL queries.
Enterprise knowledge often relies on RDF multi-
graphs coupled with ontologies (Teyeb et. al., 2015).
Modelling languages and conceptual patterns driven
by domain-specific requirements are also becoming
more prominent (Brambilla and Umuhoza, 2017)
(Galkin et. al., 2017). However, there is little work in
bridging all these paradigms towards a richer model-
based semantic support for enterprise information
systems.
4 DESIGN DECISIONS
The modelling tool to be described in this paper was
developed to combine the benefits of ZF, AMME and
RDF. By combining these, we want to show the
possibility of making a multi-perspective conceptual
frame available to semantic queries. Cross-
perspective links are governed by a meta-model
which enables the representation of enterprise
knowledge as machine-readable knowledge graphs
covering the facets of ZF that are deemed relevant for
a particular purpose.
Figure 3 shows examples of mappings between
ZF's facets and enterprise modelling symbols from
two well-known languages – Archimate (Open
Group, 2017b) and EEML (Carlsen, 1998).
The work at hand takes a similar mapping as a
starting point and specifies granular semantic links
across the ZF facets that can be exported as bridges
between RDF multigraphs. To develop the modelling
ICEIS 2018 - 20th International Conference on Enterprise Information Systems
684
Figure 3: Examples of diagrammatic symbols in modelling
languages reflecting the ZF facets.
tool, we started from creating a class diagram which
represents the meta-model governing the modelling
language, which can be evolved by applying the
AMME methodology as new requirements are
adopted (see Figure 4 for a snapshot of the meta-
model).
The meta-model was implemented using the
ADOxx platform (BOC Group, 2017), which
provides some abstract classes (e.g., __D-
construct__, __D-resource__, __D-container__ etc.)
as a starting point for the multi-perspective modelling
language. The class __D-construct__ represents the
root class for every meta-model and all the other
abstract or non-abstract classes inherit from it. As
predefined abstract classes, we also used the classes
__D-resource__, __D-container__ and __D-
aggregation__. The meta-model contains so far three
types of diagrams: (i) Model of courier (make-to-
order) processes, (ii) Model of locations and (iii)
Model of participants.
These three types of models/diagrams are linked
to each other by hyperlinks, the tasks require parkings
in cities or regions and also can have assigned roles
or employees.With these three types of models we
tried to cover the following ZF facets: (i) When/How,
(ii) Where and (iii) Who, with symbols indicated in
Figure 5.
Figure 4: Meta-Model (as Class Diagram).
Because of the project’s limited scope, not all of
ZF's facets have yet been implemented, but this target
could be reached with the help of AMME and ADOxx
platform, which allows the addition of new types of
models in order to extend the actual meta-model to
cover all ZF facets. Across these facets, hyperlinks
such as those highlighted in Figure 6 can be
established, distinguished by their semantics
according to the metamodel. Different kinds of links
are visible: The task Deliver has as assigned
employee the instance Jim or, for convenience, the
reverse link: Jim is responsible for doing the task
Deliver; the employee Jim can fulfil the role Big Car
Driver (roles can also be assigned to tasks, thus
delegating the instance assignment to some external
workflow
management system); the task Deliver
Repurposing Zachman Framework Principles for "Enterprise Model"-Driven Engineering
685
Figure 5: Examples of symbols and corresponding ZF
facets.
Figure 6: Linked model fragments.
requires parking in one of the cities, in this case City
A which contains two parkings areas of different
types.
All the models developed in the implemented
modelling tool can be exported and interpreted as
machine-readable content (RDF knowledge graphs).
RDF multi-graphs are used to represent the
knowledge, in this case the modelled enterprise
knowledge, and can be written in different
serialization syntaxes, e.g.: TriG (W3C, 2017c);
Turtle (W3C, 2017d); RDF/XML (W3C, 2017e). A
graphical representation of the model fragments and
links in Figure 6 is shown in Figure 7, including the
types derived from the metamodel (Figure 4).
Figure 7: Machine-readable knowledge graph.
5 PROOF-OF-CONCEPT
For developing the agile modelling tool we adopted
the case of a transport company that needs its courier
processes mapped on human resources and
geographical coverage, including instance processes
whose tasks can be assigned in the modelling
environment to instance responsibles and locations.
The human resources are described both in terms of
roles and instance performers, grouped by
departments or organisational units depicted as visual
containers, e.g.: Production; Research/Development;
Marketing; Finance; Resources. Figure 8 depicts
examples of models developed so far in the
implemented modelling tool: Model of make-to-order
process (M1); Model of participants (M2); Model of
locations (M3).
The first model contains a business process
comprising three types of tasks and also decisions.
The second type of diagram contains some
departments of the transport enterprise (in which we
have the employees) and also some business partners.
Finally, the third model contains cities or regions with
two types of parkings: small parking area and big
parking area. All these types of models can be linked
across
the ZF facets they represent: (i) the tasks are
ICEIS 2018 - 20th International Conference on Enterprise Information Systems
686
Figure 8: Samples of models in the multi-view modelling tool.
Figure 9: RDF model serialisation sample.
performed by employees; (ii) the tasks require
parking areas in cities.
In Figure 9 some fragments in RDF are presented,
describing the link between the Deliver task and the
Jim instance (including some positioning properties
relevant in case serialised models are imported in
another tool).
Based on these statements, we can run SPARQL
queries from any programming environment, in order
to access information from client applications (e.g., a
workflow management system). Some examples of
SPARQL queries are:
SELECT ?x WHERE
{?x a :Employee.}
SELECT ?role (Count(?task) as ?count)
WHERE {
?task a :DeliveryTask.
?task :AssignedRole ?role.}
The first query returns all the employees that work in
the given enterprise and the second query counts the
tasks assigned to a role. Based on a similar example,
paper (Buchmann et al., 2018) discussed the
Repurposing Zachman Framework Principles for "Enterprise Model"-Driven Engineering
687
overarching software engineering method enabled by
our proposal, whereas (Karagiannis and Buchmann,
2018) discusses the OWL reasoning opportunities
that are open by this proposal when using a graph
database management system such as GraphDB
(Ontotext, 2017).
6 CONCLUDING SWOT
ANALYSIS
The paper repurposes ZF principles to create a new
kind of enabler for enterprise model-driven
engineering – agile modelling tools that expose ZF
abstractions and their semantic links to knowledge-
driven software development processes. We
encapsulate the concluding discussion in a SWOT
analysis:
Strengths: The proposal provides a use case for
both (i) graph databases and (ii) agile modelling
methods which can deviate from standards in order to
fulfil enterprise-specific requirements. The ZF
abstractions are connected in ways that can be
exposed as RDF graphs and can be further
complemented by ontologies with the goal of deriving
hybrid knowledge bases – an aspect detailed in
(Karagiannis and Buchmann, 2018).
Weaknesses: Evaluations are necessary for
industry-oriented or standard languages; a client-side
proof-of-concept of model-aware application
benefitting from the proposed knowledge acquisition
enabler is not discussed in this paper – details on the
software engineering method that is enabled by the
proposal are available in (Buchmann et al., 2018).
Opportunities: (i) The GeoSPARQL ontology
(OGC, 2017) creates opportunities for geospatial
inferences, which are considered for the future
evolution of the tool; (ii) An enterprise should be
modelled in a multi-perspective manner and the
relations between those perspectives can be managed
by combining ZF's prescriptive schema, AMME and
RDF; (iii) Graph-like structures are more intuitive for
humans as means of knowledge representation and
the proposed tool could be evolved towards a
graphical RDF editor.
Threats: So far the adoption of RDF and
Semantic Web principles is still slow in enterprise
applications.
ACKNOWLEDGEMENT
This work is supported by the Romanian National
Research Authority through UEFISCDI, under grant
agreement PN-III-P2-2.1-PED-2016-1140.
REFERENCES
BOC-Group GmbH, 2017. ADOxx platform page – official
website, http://www.adoxx.org/live, last accessed: 30
th
October 2017.
Bork, D., 2015. Using conceptual modelling for designing
multi-view modelling tools In: Proceedings of the 21st
Americas Conference on Information Systems,
Association for Information Systems.
Brambilla, M., Umuhoza, E., 2017. Model-driven
Development of User Interfaces for IoT Systems via
Domain-specific Components and Patterns. In
Proceedings of the 19th International Conference on
Enterprise Information Systems - Volume 2, pp. 246-
253, SCITEPRESS.
Buchmann, R.A., Cinpoeru, M., Harkai, A., Karagiannis,
D., 2018. Model-aware software engineering: a
knowledge-based approach to model-driven software
engineering. In Proceedings of the 13
th
Int. Conf. on
Evaluation of Novel Approaches to Software
Engineering, SCITEPRESS (in press).
Carlsen, S., 1998. Action port model: A mixed paradigm
conceptual workflow modeling language. In:
Proceedings of Third IFCIS Conference on
Cooperative Information Systems, ACM.
Cinpoeru, M., 2017. Dereferencing service for navigating
enterprise knowledge structures from diagrammatic
representations. In Proceedings of BIS 2017
Workshops, LNBIP 303, pp.85-96, Springer.
Enter Know, 2017. Project page http://enterknow.granturi.
ubbcluj.ro, last accessed: 30
th
October 2017.
Frank, U., 2000. Multi-perspective enterprise models as a
conceptual foundation for knowledge management In:
Proceedings of the 33rd Hawaii Conf. on System
Sciences, IEEE.
Frankel, D. S., Harmon, P., Mukerji, J., Odell, J., Owen, M.,
Rivitt, P., Rosen, M., Soley, 2003. The Zachman
Framework and the OMG's Model Driven Architecture.
White paper. Business Process Trends.
Galkin, M., Auer, S., Vidal, M., Scerri S., 2017. Enterprise
Knowledge Graphs: A Semantic Approach for
Knowledge Management in the Next Generation of
Enterprise Information Systems. In Proceedings of the
19th International Conference on Enterprise
Information Systems - Volume 2, pp. 88-98,
SCITEPRESS.
Goldman, S., Naegel, R., Preiss, K. 1994. Agile
Competitors and Virtual Organizations: Strategies for
Enriching the Customer. Wiley
ICEIS 2018 - 20th International Conference on Enterprise Information Systems
688
Guarino, N., 1995. Formal ontology, conceptual analysis
and knowledge representation. In International Journal
of Human-Computer Studies, 43(5-6), pp. 625-640.
Karagiannis, D., 2015. Agile modeling method engineering.
In Proceedings of the 19
th
Panhellenic Conf. on
Informatics, pp. 5-10, ACM.
Karagiannis, D., Buchmann, R. A., 2018. A proposal for
deploying hybrid knowledge bases: the ADOxx-to-
GraphDB interoperability case. In Proceedings of the
51
st
Hawaii Int. Conf. on System Sciences, pp. 4055-
4064, University of Hawaii.
Karagiannis, D., Buchmann, R., 2016. Linked Open
Models: Extending Linked Open Data with conceptual
model information. In Journal Information System, 56,
pp. 174-197.
Karagiannis, D., Kühn, H., 2002. Metamodelling platforms.
In Proceedings of the Third International Conference
EC-Web 2002 – DEXA 2002. LNCS 2455, 182,
Springer.
Kingston, J., Macintosh, A., 2000. Knowledge management
through multi-perspective modelling: representing and
distributing organizational memory. In Journal of
Knowledge-based Systems, 13(2-3), pp. 121-131.
Levy, M., Hazzan, O., 2008. Agile Knowledge
Management. In: Encyclopedia of Information Science
and Technology, pp. 112-117, IGI Global
OGC, 2017. GeoSPARQL – A Geographic Query Language
for RDF Data, http://www.opengeospatial.org/
standards/geosparql, last accessed: 30
th
October 2017.
OMiLAB, 2017. NEMO Summer School series lectures,
http://nemo.omilab.org, last accessed: 30
th
October
2017
Ontotext, 2017. GraphDB Downloads and Resources,
http://graphdb.ontotext.com/, last accessed: 30
th
October 2017.
Peffers, K., Tuunanen, T., Rothenberger, M. A., and
Chatterjee, S., 2007. A Design Science Research
Methodology for Information Systems Research. In
Journal of Management Information Systems, 24(3),
pp. 45-77.
Schekkerman, J., 2003. How to Survive in the Jungle of
Enterprise Architecture Frameworks. pp. 139-144,
Trafford.
Sessions, R., 2007. A Comparison of the Top Four
Enterprise Architecture Methodologies. Microsoft
Developer Network Architecture Center.
Smith, B., 2004. Beyond Concepts: Ontology as Reality
Representation. In Proceedings of the Third
International Conference on Formal Ontology in
Information Systems, pp. 73-84, IOS Press.
Teyeb, J. R., Torjmen, K., M., Hernandez N., Haemmerle
O., Ben J. M., 2015. Semantic Annotation of Images
Extracted from the Web using RDF Patterns and a
Domain Ontology. In Proceedings of the 17th
International Conference on Enterprise Information
Systems - Volume 2, pp. 137-144, SCITEPRESS.
The Open Group, 2017a. ADM and the Zachman
Framework,http://pubs.opengroup.org/architecture/tog
af8-doc/arch/chap39.html, last accessed: 30
th
October,
2017.
The Open Group, 2017b. The Archimate Enterprise
Architecture Modelling Language. http://www.
opengroup.org/subjectareas/enterprise/archimate-
overview, last accessed: 30
th
October, 2017.
Villiers, D. J. de, 2001. Using the Zachman Framework to
Assess the Rational Unified Process. In The Rational
Edge, Rational Software.
W3C, 2017a. RDF - Semantic Web Standards,
https://www.w3.org/RDF/, last accessed: 30
th
October
2017.
W3C, 2017b. SPARQL 1.1 Query Language,
http://www.w3.org/TR/sparql11-query, last accessed:
30
th
October 2017.
W3C, 2017c. RDF 1.1 TriG, http://www.w3.org/TR/trig,
last accessed: 30
th
October
W3C, 2017d. RDF 1.1 Turtle, http://www.w3.org/
TR/turtle, last accessed: 30th October
W3C, 2017e, RDF 1.1 XML Syntax, https://www.w3.org/
TR/rdf-syntax-grammar/, last accessed: 30
th
October
Zachman, J. A., 1982. Business Systems Planning and
Business Information Control Study: A comparison. In
IBM Systems Journal, 21, pp. 31-53.
Zachman, J. A., 2008. John Zachman’s Concise Definition
of the Zachman Framework.. https://www.zachman.
com/16-zachman/the-zachman-framework/35-the-
concise-definition, last accessed: 30
th
October 2017.
Repurposing Zachman Framework Principles for "Enterprise Model"-Driven Engineering
689