Enterprise Ontology

The Global University Alliance (GUA) is an open group of academics with the ambition to provide both business and academia with state-of-the-art insights. Through its ties with industry standard bodies like LEADing Practice, ISO, OMG, OASIS, IEEE, government standard organizations like NATO and UN community, industry organizations and governments, the GUA is able to evaluate and valorize its scientific output. Since 2004, the members of the GUA strive for a continuous improvement of their expertise through the research, comparison, analysis and development of Best and LEADing Practices in Business. Throughout this process, the GUA built its own implicit ontology that revolves around its expertise of Best Practices, Industry Practices and LEADing practices.

Academia Industry Design: A Collaborative Process Between Research And Industry

Arising from 5 years of previous work, the GUA was founded in 2004 as a non-profit organization and today (Nov, 2015) they are an international consortium consisting of over 450+ professors, lecturers and researchers. Their aim it is to provide a collaborative platform for academic research, analysis and development. As illustrated in Figure 1, they achieve this through defining clear research themes, with detailed research questions, where they analyze and study patterns, describe concepts with their findings. This again can lead to additional research questions/themes as well as the development of artifacts, which can then be used as reference content by practitioners and industry as a whole.

The GUA collaborates with standards bodies such as:

• ISO: ‘The International Organization for Standardization (French: Organisation Internationale de standardization)
• CEN: The European Committee for Standardization (CEN, French: Comité Européen de Normalisation).
• IEEE: Institute of Electrical and Electronics Engineers is the largest association of technical professionals with more than 400,000 members
• OMG: Object Management Group: Develops the software standards.
• NATO: North Atlantic Treaty Organizations (NATO’s) with the 28 member states across North America and Europe and the additional 37 countries participate in NATO’s Partner- ship for Peace and dialogue programs, NATO represents the biggest non-standard body that standardizes concepts across 65 countries.
• ISF: The Information Security Forum, Investigates and defined information security
• W3C: World Wide Web Consortium-The W3C purpose is to lead the World Wide Web to its full potential by developing protocols and guidelines that ensure the long-term growth of the Web/Internet.

When academics, based on their research, build concepts and artefacts for practitioners, these concepts/artefacts need to be constructed rigorously to meet academic standards and be relevant for practitioners. Construction rigor is typically considered to be the domain of academia, although practitioners are also acknowledged to create knowledge and artefacts relevant to themselves and others (Nonaka, 1996). Academic artefact design methodologies have considered academia as a source of rigorously designed knowledge and artefacts, of which the relevance can be tested in practice.

However, we observe an ever-growing involvement of practitioners in academic design science (DS). March & Smith (1996) hint towards an evaluation of academic artefacts in a real world setting, as in natural science. Hevner, March, Park and Ram (2004) consider the organisational context in which academic artefacts need to serve as a major influence on relevance. They explicitly discuss the evaluation of academic artefacts in the real world, through case studies and field studies. Peffers, Tuunanen, Rothenberger and Chatterjee (2008) identify practitioner feedback as an essential aspect of artefact evaluation in a real world setting. Finally, Sein, Henfridsson, Purao, Rossi and Lindgren (2011) model the design of rigorous and relevant artefacts as a collaborative process between academics and practitioners.

In a first phase of their action-design-research (ADR) methodology, academics give the initial version of the artefacts to a small group of practitioners (e.g., a community/panel of experts). These practitioners provide valuable feedback that helps to mature the artefact. In a second phase, this improved artefact is applied by a larger group of practitioners, whose feedback will allow the academic to improve his artefact further. If this feedback requires no further modifications of the artefact, a final version of the artefact is published.

Although ADR is a very mature methodology in academia-driven artefact design, it could be made more generic (generally applicable) by alleviating (eliminating) two implicit constraints (biases) present in all DS and ADR publications:

1. Academia is the single source of rigorously constructed knowledge
2. User requirements are invariable and provide a continuous improvement feedback loop to academia.

As regards the first point, although practitioners typically create knowledge (artifacts) that is (are) relevant for them in a specific organizational context, this does not necessarily imply that this knowledge cannot be generalized and applied in other organizational contexts. This generalization (and evaluation) would typically be the role of academia in this kind of knowledge creation scenario. There are multiple instances of such artifacts existing today. For example, the Boston Consulting Group (BCG) created a matrix in 1970 to help analyze organizations’ product lines. This has enabled organizations to allocate resources as well as use it as an analytical tool in brand marketing, product management, strategic management, and portfolio analysis. While widely used, several academic evaluations have given feedback on its usage as a growth–share matrix. (Armstrong; Scott; Brodie; Roderick, 1994) A detailed academic study from Slater and Zwirlein (1992), analyzed 129 organizations.

The conclusion of the study was that those who follow the BCG matrix as a portfolio-planning model for growth success had lower shareholder returns. The study concluded that the BCG matrix is a relevant and useful artifact, but it was applied incorrectly and should be applied in other general contexts. Such an evaluation would typically be the contribution of academia in this kind of knowledge creation scenario.

As regards the second point, in ADR, an academic artifact is handed over to practitioners as soon as they accept it. This approach does not account for new feedback when user requirements have changed and the artifact is no longer relevant in its current form. From requirement engineering, requirement modeling and requirement architecture it is known that user requirements continuously change. (Gotel and Finkelstein, 1994, Ralph and Wand, 2009) Therefore, what is needed in reality is an approach that allows for continuous artifact improvement/modification through continuous user feedback, and values user knowledge as valid (relevant) input (which could thus be made more rigorous).

A major difference between academia and practice is the way knowledge is acquired. Practitioners typically rely on Experience and Induction, while Academia use research, analysis, deduction and the scientific method. From the above discussion points, we could argue that the academia and practitioners are complementary in the following ways:

• Rigor vs. Relevance: we can determine that Academia does Rigor best, while Practitioners do Relevance best.
• Abstraction Level: Academia typically designs solutions at the type level (concepts and solution for a type of problem) while Practitioners typically design solutions at instance level (solution for a particular problem).
• Knowledge creation processes in terms of developing artifacts should interlink between rigor and relevance, of which the rigor aspect can be analyzed in theory best and the relevance can be tested in practice best. Therefore:
  • Combining explicit knowledge to develop new explicit knowledge. Academia typically combines explicit knowledge at type or instance level to create new knowledge concepts at type level. Whereas the practitioners typically combine explicit knowledge at type or instance level to create new knowledge at instance level. The latter being described by Nonaka (1996).
  • Internalization: Converting explicit knowledge (e.g. books, standards) to tacit knowl- edge (e.g. personal knowledge). Academia typically teaches explicit knowledge to be transformed into tacit knowledge of students (e.g. practitioners). Whereas practitioners typically study academic concepts and non-academic solutions to develop competencies (tacit knowledge), which was also described by Nonaka (1996).
  • Socialization: Sharing tacit knowledge through interaction. Academia research share tacit knowledge in doing research and publications together. Whereas practitioners share tacit knowledge by doing things together (and learning from each other while doing). The knowledge creation mode involving only practitioners was also identified by Nonaka (1996).
  • Externalization: The need to convert tacit knowledge into explicit knowledge. Academia studies in this context, what practitioners do (at instance level) to create new knowledge at type level. Whereas practitioners sometimes document what they do, and sometimes share this content (e.g. industry standards, best practices).
  • Feedback Loop: There should be a loop of feedback and enhancement between aca- demia and practitioners. Figure 1 visualizes the knowledge creating processes in academia and practice and how they interact. Academics develop research questions, founded on the research themes they identified. They analyze real-world situations to answer their research question through the identification of patterns (e.g. laws of physics). These patterns are documented and combined with other knowledge (patterns and concepts) to build theories that might require additional concepts, which may lead to additional research themes. Industry practitioners will use these concepts and patterns to develop artifacts that will help them structure their knowledge about the business reality they experience. These artifacts will be published to peers (e.g., as standards), used and improved by them. These improvements, which may point towards user requirements that were not identified by academics, should feedback to academia. Since practitioners will mostly use the concepts embedded in the artifacts to document their knowledge, the expected impact of practitioner feedback on the elementary concepts of business is expected to be relatively low (i.e., New business concepts are not discovered that often). However, it is very likely that aca- demics will observe new innovative ways of working with their artifacts in real-life situations, when observing the practitioners in the industry. Industry practitioners can also develop their own artifacts, which may contribute directly to the academic literature. The likelihood of this scenario is expected to be between that of identifying completely new concepts and discovering new application scenarios.

The next section discusses the value of ontology, this provides the theoretical foundations for the benefits of the ontology structure and skeleton that is presented afterwards.

The Value Of Ontology

An ontology is an artifact, more precisely an intentional semantic structure that encodes the set of objects and terms that are presumed to exist in some area of interest (i.e. the universe of discourse or semantic domain), the relationships that hold among them and the implicit rules constraining the structure of this (piece of) reality.(Genesereth & Nilsson, 1987; Nicola Guarino & Giaretta, 1995) In this definition, intentional refers to a structure describing various possible states of affairs, as opposed to extensional, which would refer to a structure describing a particular state of affairs. The word semantic indicates that the structure has meaning, which is defined as the relationship between (a structure of) symbols and a mental model of the intentional structure in the mind of the observer. This mental model is often called a conceptualization (Gruber, 1993). Semantics are an aspect of semiotics, like syntax, which distinguishes valid from invalid symbol structures, and like pragmatics, which relates symbols to their meaning within a context (e.g., the community in which they are shared). (Cordeiro & Filipe, 2004).

Ontologies can be categorized and classified according to several criteria (e.g., context, maturity) (von Rosing, Laurier, & Polovina, 2015b). When ontologies are classified accord- ing to their universe of discourse, we distinguish foundational, domain, task and application ontologies. (N. Guarino, 1997) Top-level or foundational ontologies cover a very broad area of interest as they describe very general concepts as space, time and matter that are needed in any field or domain. Task and domain ontologies all relate to a specific semantic domain (e.g., business process, infrastructure, data) or task (e.g., analysis, design). Domain and task ontology terms reuse or specialize top-level ontology terms. Finally, application ontologies relate to a very specific universe of discourse (e.g. business process design in a particular company, data analysis in a specific department).

Their vocabulary can be built from scratch or defined as specializations of both domain and task ontology terms (Nicola Guarino, 1998). For example, if ‘event’ were an ontology construct for the business process domain and ‘forecasting’ would be a task ontology construct for analysis, ‘event forecasting’ could be an application ontology construct for business process analysis. This combination approach is expected to promote reuse; standardization and mutual understanding between applications, as the same domain and task construct definitions are reapplied across applications. Business ontology is an intentional semantic structure that has business as its universe of discourse.

Business ontology research has long been focusing on two distinct axes. The first axis concentrated on the development methods for ontology engineering (from scratch) by practitioners (e.g., METHONTOLOGY, On-To-Knowledge, DOGMA, SENSUS), which enabled them to build their own corporate or enterprise ontologies. (Cardoso, 2007; Corcho, Fernández-López, & Gómez-Pérez, 2003; Lima, Amaral, & Molinaro, 2010). The second axis was dominated by the development of domain ontologies (e.g., REA, e3value, BMO, TOVE) by academics. (Fox, 1992; Geerts & McCarthy, 2002; Gordijn & Akkermans, 2001; Osterwalder, 2004) Standards bodies, which are mainly practitioner organizations, have recently started to build their own domain ontologies (e.g. FIBO). (Council, 2014).

The Global University Alliance Ontology Structure And Classifications

As ontology formally represents knowledge as a set of concepts within a domain, and the relationships between those concepts, it can be used to model a domain and support reasoning about concepts. The Global University Alliance has used the concept of ontology as their basis for categorizing and classifying all their concepts. It thereby provides the basis for both a shared vocabulary and the very definition of its objects and concepts. We realize that there are various categorization as well as classifications of ontologies both in academic literature (Gomez-Perez et al. 2004; von Rosing & Laurier, 2016; Borgo 2007, Lassila and McGuinness 2001; etc) as well as in practice OWL, OMG MOF, etc. Each of them have a specific purpose, therefore the categorization and the classification is focused on the expressivity and formality of the specific languages used/proposed: natural language, formal language, etc.

The other more general applicable categorization as well as classifications of the ontologies, is centred around the scope of the objects described by the ontology. (Roussey, C., Pinet, F., Ah Kang, M., and Corcho, O. 2011). Since the enterprise ontology of the Global University Alliance, is and should be generally applicable within any organization. The more general applicable categorization and classifications of the ontologies, was chosen. Thereby the Ontology classification is centered around the sphere, filed and level and the categories is grounded on the scope of the objects described by the ontology.

The GUA, has found that there is a benefit of categorizing and classifying the ontologies around the scope of the objects described. For instance, the scope of an application ontology is narrower than the scope of a domain ontology; domain ontologies have more specific concepts than core reference ontologies, which contains the fundamental concept of a domain. Foundational ontologies can be viewed as meta ontologies that describe the upper level concepts or primitives used to define the other ontologies. (Roussey, C., Pinet, F., Ah Kang, M., and Corcho, O. 2011, von Rosing, Zachman 2017).

The Global University Alliance Ontologies

The Global University Alliance uses MOF-Meta Object Facility (OMG), Basic Formal Ontology (BFO) and Zachman Enterprise Ontology as their Top-Level Ontologies. The Top-Level Ontology describes primitives that allow for defining very general concepts like space, time, matter, object, event, action, etc. (Adapted from N. Guarino, 1997) Provides the foundation for the formal system that allows for developing meta-meta-models, of which the completeness and clarity needs to be guaranteed trough a mapping between a top-level ontology and the formal system’s primitives (MOF). (von Rosing, Zachman 2017).

The Enterprise/Business Ontology is the Foundational Ontology. It is a generic ontologies applicable to various domains. It defines basic notions like objects, relations, structure, arrangements and so on. All consistent ontology should have a foundational ontology. (Roussey et al, 2011) Foundational ontology can be compared to the meta model of a conceptual schema (Fonseca et al. 2003). It is a system of meta-level categories that commits to a specific initial-view. We use the foundational ontology, to provide real-word semantics for general conceptual modelling languages, and to constrain the possible interpretations of their modelling primitives. As such, we map our meta-meta-model (M3) to our foundational ontology. Both to certify its comprehensiveness and clarity. It also ensures that all can and will relate through our Enterprise/Business Ontology.

The Business Layer Ontology, Information Layer Ontology and the Technology Layer Ontology are our Core Reference Ontologies.  They are the standard used by all our different groups of users. These type of ontology are linked to a specific topic/domain but it integrates different levels and tiers related to specific group of users. We know from theory that core reference ontologies as well as domain ontologies based on the same foundational ontology can be more easily integrated. (Roussey et al, 2011).

Our layered enterprise ontologies are the result of the integration of the sublayer domain ontologies. However, they are a formal (i.e., domain independent) system of categories and their ties that can be used to construct models of various domains, and not one of a specific domain. Our core reference ontologies are built to catch the central concepts and relations of the specific layers. They provide the foundations for a (generic) modelling language trough a mapping between the core reference ontology and the modelling language’s meta-model (M2).

The Domain Ontologies of Value, Competency, Service, Process, Application, Data, Platform and Infrastructure, describe, the context and vocabulary related to their specific domain by specializing the concepts introduced in the core-reference ontology. In the Enterprise/Business Ontology, the domain ontologies are linked to a specific core reference ontology layer. In terms of the MOF tiers, they provide the foundations for a domain-specific modelling languages (M2) trough a mapping between the domain ontology and the modelling language’s meta-model. Each specific domain ontology is only valid to a layer with their specific view point, however the layers relate through the semantic relations, captured in the foundational ontlogy. Therefore, the individual viewpoints, ensures the ability to engineer, architect or model across multiple sublayers. That is to say that the viewpoints defines how a group of users conceptualize and visualize some specific phenomenon of the sublayers. The domain ontologies could be linked to a specific application. (Roussey et al, 2011) They provide the foundations for a domain-specific modelling languages (M2) trough a mapping between the domain ontology and the modelling language’s meta-model. (G. Guizzardi, 2005).

The Tiering Ontology, Categorization Ontology, Classification Ontology, LiveCycle Ontology, Maturity Ontology, Governance Ontology, Blueprinting Ontology, Enterprise Requirement Ontology as well as Layered Enterprise Architecture Ontology are all a part of the Task Ontologies. They provide the basis to the generic tasks relevant to both the domain ontologies and application ontologies. They do this by specializing the terms introduced in the core-reference ontology, therefore ensuring full interoperability across the various task ontologies and the core reference, domain and the application ontologies. The task ontology contains objects and descriptions of how to achieve a specific task, on the other hand the domain ontology portrays and defines the objects where the task is applied. In terms of the MOF tiers, they provide the foundation for a task-specific modelling language (M2) trough a mapping between the task ontology and the modelling language’s meta-model.

The Application Ontologies describe concepts of the domain and task ontologies. Often the Application Ontologies are specializations of both the related ontologies in order to fulfil the specific purpose of a specific use, function, purpose and thereby application. In terms of the MOF tiers, they provide the foundation for a model (M1) trough a mapping between the application ontology and the model. The Global University Alliance has the following Application Ontologies:

• Force & Trend Ontology
• Strategy Ontology
• Planning Ontology
• Quality Ontology
• Risk Ontology
• Security Ontology
• Measurement Ontology
• Monitoring Ontology
• Reporting Ontology
• Capability Ontology
• Role Ontology
• Enterprise Rule Ontology
• Compliance Ontology
• Business Workflow Ontology
• Cloud Ontology
• Business Process Ontology
• Information Ontology
• Infrastructure Ontology
• Platform Ontology
• Enterprise Culture Ontology

The recent Enterprise Ontology workgroup within the Global University Alliance has adopted the concept of holistic Enterprise Ontology Frameworks, as it identified the necessity of introducing such a framework into today’s enterprises through the LEADing Practice community. The Enterprise Ontology concepts in the framework should provide sound semantic foundations for best and LEADing practices in different domains (e.g. process, service, value, information). The concepts and practices will be shared and published as an open standard in the LEADing Practice community. Thereby enabling all academics and practitioners in the community to build on common leading principles to identify, create and realize value, competitive advantage and agility to deal with future challenges.

To realize this vision, the GUA alliance reaches out to all business and enterprise ontology researchers to contribute to the consolidation of academic findings in a research-based Enterprise Ontology Framework.

Research Team

The Enterprise Ontology research team and contacts are:

Ontology Research Coordinator:
Professor Wim Laurier, Saint-Louis University Brussels & Ghent University, Belgium

Enterprise Semantic Research Coordinator:
Professor Simon Polovina, Sheffield Hallam University, UK

Global University Alliance Coordinator:
Professor Mark von Rosing, Head of Global University Alliance, France

Enterprise Ontology Roles Coordinator:
Maxim Arzumanyan, St. Petersburg, Russia

Public Sector Coordinator:
Miss Sumaya Kagoya, Makerere Business School, Uganda, Africa

Processes Research Coordinator:
Prof. Oneil Joesphs, University of Technology, Jamaica

Technology Innovation Research Coordinator:
Dr Anas Najdawi, Amity University, Dubai

Security Research Coordinator:
Gregory Robinson, University of Collorado, US

Cyber Security Research Coordinator:
Mr Gregg Ibbotson, Academic Researcher, UK

Architecture Research Coordinator:
Dr. Bonnie Urquhart, UNBC, Canada

Enterprise Engineering Research Coordinator
Prof. Ardavan Arminim Birmingham City University, UK

Data Research Coordinator:
Jimmy Doan, ESIEA Graduate Engineering, France

Enterprise Ontology Modelling Coordinator:
Prof. Hans Scheruhn, Harz University, Germany

Enterprise Ontology Strategies:
Jamie Caine, SHU, UK

Enterprise Ontology Testing Coordinator:
Ulrik Foldager (industry researcher), Denmark

As far as partners are involved, these are the collaboration partner contacts:

Enterprise Standard Body:
Georg Etzel
LEADing Practice, Co-CEO

Enterprise Architecture Framework:
John A. Zachman
Inventor and Father of Enterprise Architecture
Zachman International

International Organization for Standardization
Johan H Bendz
ISO SC7
WG 42 Convener

IEEE Coordinator:
Rich Hilliard
Institute of Electrical and Electronics Engineers
Editor of IEEE Std 1471:2000, Project editor, ISO/IEC/IEEE 42010

Software Standards Body:
Henk De Man
OMG VDML Chairman

NATO Coordinators:
Johan Goossens
NATO Allied Command Transformation, Branch Head, Technology & Human Factors

UNESCO Coordinator:
Dr. Selin N. Şenocak
UNESCO Chair Holder
Cultural Diplomacy, Governance and Education
Director, Occidental Studies Applied Research Center
Political Sciences and International Relations Faculty Member

CSIR Coordinator:
Rentia Barnard, Research Institute CSIR
Enterprise Architect Research Group Leader

Information Security Standards Body:
Steve Durbin, CEO of Information Security Forum

OMG, Software Standards Body:
Fred Cummins, Business Modeling & Integration Task Force, Chairman
OMG Business Architecture Special Interest Group, Co-Chair

 

GUA Enterprise Ontology Foundational Introduction

GUA Enterprise Ontology Advanced Introduction