OTRA: A Risk Management Ontology for Transparent Service Level

Agreements in Federated Cloud Environment

Giulia Biagioni

1 a

, Resul Serkan Keskin

1 b

, Lawrence Cook

2 c

, Edwin Harmsma

1 d

and

Erik Langius

1 e

Netherlands Organisation for Applied Scientiﬁc Research (TNO), The Hague, The Netherlands

BIT, Ede, The Netherlands

{giulia.biagioni, serkan.keskin, edwin.harmsma, erik.langius}@tno.nl, lawrence@bit.nl

Keywords:

Cloud Federation, Service Level Agreement, Risk Management Ontology, Bow-Tie RDF-Based

Implementation.

Abstract:

Cloud Federation, a model where multiple cloud providers collaborate to offer interconnected services, has

become increasingly renowned in cloud computing. In this environment, Service Level Agreements (SLAs)

play a critical role by formalising service expectations and fostering transparency between providers and con-

sumers. Although existing SLA ontologies provide a semantic foundation for standardised SLA management,

they lack components for detailed risk management and representation within SLA. This paper presents our

work on developing OTRA - Ontology for Transparent Agreement - speciﬁcally designed for risk manage-

ment within SLAs in federated cloud environments. By integrating structured risk identiﬁcation, prevention

and mitigation strategies, our ontology extends the Gaia-X Core Ontology. We validate our ontology through a

controlled simulation that demonstrates its effectiveness in structuring SLA risk information, offering a foun-

dation for transparent, accountable service agreements.

1 INTRODUCTION

In the evolving landscape of cloud computing, cloud

federation - a model where multiple cloud service

providers (CSPs) collaborate to deliver an intercon-

nected network of services (Kurze et al., 2011) -

is gaining increasing popularity. This environment

offers signiﬁcant beneﬁts such as data sovereignty,

which allows organisations to store and manage

their data within speciﬁc jurisdictions, ensuring com-

pliance with regional regulations (Holfelder et al.,

2022)(De Filippi and McCarthy, 2012). Addition-

ally, cloud federation helps avoid vendor lock-in, giv-

ing organisations the ﬂexibility to choose services

from different providers without being tied to a sin-

gle vendor’s ecosystem (Bermbach et al., 2013). This

ﬂexibility ensures better resource optimisation, cost

management, and control over critical infrastructure

choices (Bonﬁglio, 2021).

https://orcid.org/0000-0002-9005-7945

https://orcid.org/0000-0002-4844-7982

https://orcid.org/0009-0008-0536-0511

https://orcid.org/0009-0009-8892-4593

https://orcid.org/0009-0006-2765-6781

Within this model, Service Level Agreements

(SLAs) play a crucial role. SLAs are formal con-

tracts that deﬁne the expected level of service between

providers and consumers. They ensure transparency

and accountability, helping to maximise value for all

parties by clearly deﬁning expectations and mitigat-

ing legal risks (L

e and Nguyen, 2020).

Several ontologies have been developed to repre-

sent SLAs in the context of cloud federation, such as

CSLAOnto (Labidi et al., 2016) and the generic SLA

ontological model of SLAVIDES (Karamanlioglu and

Alpaslan, 2018). These ontologies address the need

for standardised, automated, and ﬂexible SLA man-

agement. They offer a semantic-based format that

enables automatic monitoring, understanding and en-

forcement of SLAs, ensuring that service agreements

are clearly deﬁned and adhered to. When integrated

into expert systems, they enable the detection of SLA

violations across multiple sectors, potentially lead-

ing towards improving compliance and performance

management in federated cloud ecosystems.

Despite the considerable efforts made in the land-

scape of SLA ontology development, semantic mod-

els speciﬁcally designed to delineate risks in SLAs for

cloud federation have not been developed to the same

Biagioni, G., Keskin, R. S., Cook, L., Harmsma, E. and Langius, E.

OTRA: A Risk Management Ontology for Transparent Service Level Agreements in Federated Cloud Environment.

DOI: 10.5220/0013209300003950

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 15th International Conference on Cloud Computing and Services Science (CLOSER 2025), pages 119-126

ISBN: 978-989-758-747-4; ISSN: 2184-5042

119

extent. Semantic-based formats that could be fed into

expert systems to trigger strategies aimed at prevent-

ing the occurrence of hazardous events, or mitigating

their impact are still under development. This gap

represents an opportunity for further advancements

in the integration of risk management and prevention

mechanisms into the SLA frameworks used in cloud

federation.

In this article, we introduce our work on develop-

ing OTRA, a specialised ontology for risk manage-

ment within SLAs in federated cloud environments.

This ontology is designed to ﬁll existing gaps in SLA

frameworks by explicitly modelling and representing

risks, which are essential in cloud federation scenar-

ios where multiple Cloud Service Providers (CSPs)

interact and collaborate across varying regulatory en-

vironments. The ontology aims to provide a struc-

tured and semantic foundation for risk identiﬁca-

tion, prevention, and mitigation within SLAs, thereby

enhancing transparency and accountability for both

providers and consumers.

This article is structured as follows: following the

introduction, Section 2 provides an overview of the

state of the art in ontology development for risk man-

agement and SLAs, summarising existing approaches

and identifying gaps that motivate our work. Section

3 details the ontology design of OTRA, presenting its

conceptual underpinnings through First-Order Logic

(FOL) axioms. This approach ensures that our con-

ceptualisation remains technologically agnostic, al-

lowing for ﬂexible implementation across different

languages and evolving standardised semantics.

In Section 4, we describe the evaluation process of

the ontology, including a simulation designed to val-

idate its effectiveness in a controlled operational en-

vironment, achieving a Technology Readiness Level

(TRL) of 5. Section 5 discusses the results of this

evaluation, interpreting the ontology’s performance

and insights gained from the simulation. Finally, Sec-

tion 6 outlines future work, exploring potential exten-

sions to enhance the ontology’s applicability and ex-

pressiveness in dynamic cloud federation contexts.

2 STATE OF THE ART

Ontologies, in the context of RDF, are formal repre-

sentations of knowledge that deﬁne concepts, entities,

and the relationship between them within a speciﬁc

domain. Regarding their technical application, they

can be used for enabling interoperability and connect-

ing data silos (Roussey et al., 2011)(Allemang et al.,

2020)(Blumauer et al., 2020), feeding expert systems

with structured knowledge (Prapakorn and Chittaya-

sothorn, 2009), where example of such applications

can be found in domains like health (Mustafa et al.,

2023) and energy (Pruvost et al., 2023), allowing

machines to understand the meaning of information

(Hogan, 2020), and enhancing AI trustworthiness by

providing explainable reasoning when integrated with

machine learning-driven methods (Pan et al., 2023).

In risk management, the Open Risk community

has contributed signiﬁcantly to developing ontolog-

ical models, promoting open-source risk analysis

tools for transparency within ﬁnancial sectors. Their

frameworks, such as the Description of a Model On-

tology (DOAM)

and the Risk Function Ontology

(RFO)

, offer structured means to annotate and cat-

egorise key concepts in risk modelling and manage-

ment roles.

DOAM, for example, facilitates the semantic or-

ganisation of risk model components, while RFO

structures essential skills, occupations, and functions

to support interoperability. Though valuable for anal-

ysis, both ontologies prioritise metadata-level anno-

tation, which limits their direct application in expert

systems or within service-level agreements (SLAs)

between providers and consumers in cloud federa-

tions.

Incorporating detailed risk-related information

within SLAs between providers and consumers is crit-

ical for: reducing ambiguity, ensuring awareness of

conditions, clarifying mitigation strategies, and mak-

ing implicit assumptions explicit (Uriarte et al., 2016)

(Alkhamees, 2022). If DOAM and RFO lack the ca-

pability to directly map risk frameworks onto data due

to their analytical focus and ﬁnancial sector orien-

tation, SLA-speciﬁc ontologies like SLAVIDES and

CSLAOnto similarly miss essential components for

representing risk and management strategies within

service agreements.

CSLAOnto is speciﬁcally designed for cloud com-

puting SLAs, focusing on standardising terms related

to service quality, performance, and responsibilities

(Labidi et al., 2016), thus enhancing semantic in-

teroperability but not directly addressing risk man-

agement. SLAVIDES (Karamanlioglu and Alpaslan,

2018), while formalising SLA terms, similarly lacks

built-in risk representation and management compo-

nents. Consequently, these ontologies offer valuable

structure but are not yet at a readiness level for com-

prehensive, risk-sensitive applications in operational

Regarding the Open Risk Community, cf: https://

www.openriskmanagement.com

https://www.openriskmanual.org/ns/doam/index-

en.html

https://www.openriskmanual.org/ns/rfo/index-

en.html

CLOSER 2025 - 15th International Conference on Cloud Computing and Services Science

120

environments.

Regarding ontological efforts aimed at fostering

interoperability in cloud federations at an operational

and industry-driven level, the Gaia-X initiative offers

a notable example. Gaia-X is an initiative dedicated

to developing digital governance that can be applied

to foster transparency, controllability, portability, and

interoperability in the federated cloud ecosystem and

data spaces

. Among its various activities, Gaia-X

supports the collaborative development of ontologi-

cal modules that enable seamless data and service in-

tegration across cloud providers. One of its founda-

tional ontological modules is the Gaia-X Core On-

tology, which structures key concepts such as Par-

ticipants (e.g., Provider, Consumer), Resources (e.g.,

Data, Software Asset), and Service Offerings to en-

hance compatibility across cloud systems. While

the Gaia-X ontologies effectively establish founda-

tional roles and service interactions, they currently

lack the necessary components to deﬁne a compre-

hensive risk management framework within SLAs be-

tween providers and consumers.

To address these identiﬁed gaps, we have devel-

oped a new risk management ontology speciﬁcally

designed for SLAs, aiming to comprehensively rep-

resent risk and management strategies within service

agreements. This ontology extends the Gaia-X Core

Ontology module by facilitating structured, action-

able risk information, allowing its seamless integra-

tion within SLAs. The following section presents our

framework in detail, highlighting its potential to en-

hance transparency and risk management within fed-

erated cloud environments.

3 ONTOLOGY DESIGN FOR

RISK MANAGEMENT

Our risk management ontology, OTRA, is grounded

in the bow-tie method

. The bow-tie method is a

structured risk assessment model used in industries

to visualise risks by organising both preventive mea-

sures and mitigation strategies around a central haz-

ardous event (Omidvar et al., 2024) (Kluwer, 2017)

(Kluwer, 2019). Resembling a bow tie, the model

consists of two branches extending from the ”knot”

or core event: the left side represents potential threats

and preventive actions, while the right side outlines

https://gaia-x.eu/what-is-gaia-x/about-gaia-x/

The bow-tie method reﬂects conceptual structure out-

lines in the ISO 31000:2018. This standard identiﬁes eight

core concepts critical for effective risk management: Risk,

Risk Management, Stakeholder, Risk Source, Event, Con-

sequence, Likelihood and Control.

potential consequences and mitigation measures. The

bow-tie model is displayed in Figure 1.

Hazard

Control

(Preventive)

Control

(Remedy)

Threats

Impact

Figure 1: The bow-tie diagram for risk assessment.

Industries such as oil and gas (Kbah et al., 2020),

healthcare (McLeod and Bowie, 2020), and aviation

(Aust and Pons, 2020) use the bow-tie method to de-

velop clear action plans for preventing and addressing

the outcomes of hazardous events.

Our ontology structures each Hazardous Event to

include essential elements like Preconditions, Pre-

ventive Mechanisms, Postconditions, and Remedies,

closely mirroring the bow-tie method’s approach to

capture potential triggers and responses. This align-

ment allows our ontology to systematically map risks

and mitigation strategies that apply to SLAs in feder-

ated cloud environments. Key elements include:

• Hazardous Event: Identiﬁes risks that may im-

pact SLA parties, linked to Preconditions (factors

that may trigger the event) and Postconditions (the

resulting outcomes or states after the event).

• Prevention Mechanism: Outlines preventative

actions aimed at reducing identiﬁed risks.

• Remedy: Details corrective measures to address

adverse events if they occur.

• Agreement: Serves as the core SLA ele-

ment, connecting Gaia-X Consumer and Gaia-X

Provider roles and embedding risk considerations

within the SLA framework.

In our formal delineation of the ontology, which

subsequently guided its technical implementation,

each Hazardous Event x, representing a risk within

SLA agreements, is linked to a Precondition y that

may act as a trigger. This relationship is expressed in

First-Order Logic (FOL) as:

∀x (HazardousEvent(x) →

∃y(Precondition(y) ∧ hasPrecondition(x, y)))

To reduce the likelihood of hazardous events oc-

curring, each Hazardous Event is linked to a Preven-

tion Mechanism w, which is speciﬁcally designed to

counteract its occurrence:

OTRA: A Risk Management Ontology for Transparent Service Level Agreements in Federated Cloud Environment

121

∀x (HazardousEvent(x) →

∃w(PreventionMechanism(w)

∧hasPreventionMechanism(x, w)))

Each Precondition y associated with a Hazardous

Event is also assigned a Prevention Mechanism w to

address conditions that could lead to potential risks.

We represent this relationship in First-Order Logic as

follows:

∀y(Precondition(y) →

∃w(PreventionMechanism(w)

∧hasPreventionMechanism(y, w)))

The hasPreventionMechanism relationship in-

cludes an OR operator in its domain, applying to ei-

ther a Precondition or a HazardousEvent, with a range

on PreventionMechanism.

When a HazardousEvent x occurs, it produces an

outcome represented by the Postcondition z, indicat-

ing the resulting state:

∀x (HazardousEvent(x) →

∃z(Postcondition(z) ∧ hasPostcondition(x, z)))

Lastly, each HazardousEvent is connected to a

Remedy v, which serves as a corrective action to

prevent the resulting Postcondition from manifesting.

This relationship can be formalised as:

∀x (HazardousEvent(x) →

∃v(Remedy(v) ∧ hasRemedy(x, v)))

Each Postcondition z associated with a Hazardou-

sEvent is similarly assigned a Remedy v, ensuring

proactive response measures. This relationship is rep-

resented in First-Order Logic as:

∀z(Postcondition(z) →

∃v(Remedy(v) ∧ hasRemedy(z, v)))

Ultimately, similar to the nature of the

hasPreventionMechanism relationship, the

hasRemedy relationship includes an OR opera-

tor in its domain, indicating it applies to either a

Postcondition or a HazardousEvent, with a deﬁned

range on Remedy.

The Hazardous Events, representing risks in

SLAs, are inherently linked to Agreements. The

Agreement concept formalises the binding between

speciﬁc risks and the involved parties. An agreement

a connects to a hazardous event x via the hasRisk re-

lationship, embedding risk considerations within the

contractual terms:

∀a(Agreement(a) →

∃x (HazardousEvent(x)∧ hasRisk(a, x)))

Each agreement a also incorporates both a

consumer c and provider p, represented by the

involvesParty relationship. This connects the en-

tities directly accountable for the risk terms outlined

within the SLA. Formally, we represent this as:

∀a(Agreement(a) →

∃c, p(Consumer(c) ∧ Provider(p)

∧involvesParty(a, c) ∧ involvesParty(a, p)))

The ontology, as formally described above, has

been technically implemented using OWL and RDFs

standardised semantics, capturing the relationships

between key elements such as hazardous events, pre-

conditions, prevention mechanisms, postconditions,

and remedies. This structured semantic model can

be visualised as shown in Figure 2, where rectangles

represent classes, and arrows illustrate object property

relationships.

Hazardous Event

hasPrecondition

Precondition

Prevention

Mechanism

hasPreventionMechanism

hasPostcondition

Postcondition

Remedy

hasRemedy

⨆ ⨆

hasRisk

Agreement

Gaia-X:Consumer Gaia-X:Provider

⨆

involvesParties

Figure 2: Visual representation of the ontology.

CLOSER 2025 - 15th International Conference on Cloud Computing and Services Science

122

The OWL and RDF-based ontology is accessi-

ble for reference and collaborative contributions at

the following repository: https://github.com/OTRA-

anon/OTRA.

4 EVALUATION PROCESS

OTRA ontology was validated to ensure both struc-

tural consistency and relevance in representing do-

main knowledge effectively. For consistency, we ap-

plied the Hermit 1.4.3.456 reasoner to check the on-

tology’s logic and coherence with and without in-

stance data, conﬁrming its integrity.

To further test the ontology’s effectiveness in or-

ganising SLA risk information, we developed a sim-

ulation of a digital marketplace environment during

the ”Tech-X Conference and Hackathon 2024” event.

This conceptual marketplace was designed to enable

cloud service providers (CSPs) to offer services such

as IaaS, PaaS, and SaaS, with simulated consumers

capable of selecting and combining services across

multiple providers.

For seamless service interoperability within the

simulated cloud federation, we proposed Liqo

, an

open-source Kubernetes extension, as a mechanism

to facilitate peering between providers (Iorio et al.,

2022). Although the simulation did not involve actual

users or infrastructure, our conceptual pipeline lever-

aged Liqo’s capability to create ”virtual nodes” that

represent the capacity of a remote cluster within the

local cluster’s environment. This approach enabled

simulated clusters to expand their resource pool dy-

namically, supporting resource sharing and ﬂexible

workload distribution across Kubernetes clusters in a

federated cloud context.

In this model, the consumer cluster, depicted on

the left part of Figure 3, initiates an outgoing peer-

ing relationship with a provider cluster on the right,

thereby gaining access to additional resources from

the provider.

Within this simulated federation, SLAs played a

foundational role in formalising service agreements.

We embedded our risk management ontology in the

SLAs to clearly outline potential risks, preventive

measures, and remedies, thereby creating a struc-

tured, transparent framework for risk management.

This approach aimed to foster informed decision-

making and trust within federated cloud environ-

ments.

The SLA template from BIT

was used as a case

https://docs.liqo.io/en/stable/index.html

https://www.bit.nl/en/home-en

Figure 3: Illustration of Liqo-enabled resource sharing in

a federated cloud environment. In the ﬁgure, the directed

arrow signiﬁes the peering connection, which allows the

consumer to ofﬂoad workloads to the provider, mirroring

the provider’s capacity within its own infrastructure through

a ”virtual node.” This virtual node in the consumer clus-

ter represents the resources of the provider cluster, enabling

workloads or pods to run on the remote infrastructure while

remaining under the consumer’s management framework.

study in our simulation to demonstrate how the ontol-

ogy could support digital, legally binding agreements.

This allowed all parties (providers and consumers) to

view potential risks and planned response strategies.

In the simulated SLA, an agreement labelled

Agreement34 was formed between the consumer, Al-

ice, and the provider, Company X. This agreement

contained a termination clause specifying conditions

under which Company X could end the contract. Ac-

cording to this clause, Company X reserved the right

to terminate the SLA if Alice engaged in harmful or

non-compliant behaviours, such as unauthorised ac-

cess, denial-of-service attacks, or policy violations.

This termination scenario, modelled as the Hazardous

Event risk5678, represented actions that could lead

to contract termination.

The Hazardous Event risk5678 was directly

linked to both Precondition precondition1245

and Postcondition Termination89. Here,

precondition1245 speciﬁed behaviours (e.g.,

data interception, spam, unauthorised access) that

might trigger termination risks. If Alice engaged

in these non-compliant actions, Termination89

would be activated, resulting in contract termination.

This structured linkage created a pathway from

risky behaviour (risk5678) to its consequence

(Termination89) within the SLA.

Our ontology encouraged the CSP to incorporate

preventive and corrective mechanisms, even though

these elements were absent from the original SLA

template.

To reduce the likelihood of termination, a preven-

tion mechanism labelled communication89 (“Com-

munication”) was recommended. This mecha-

nism, associated with both the hazardous event

and the precondition, encouraged proactive man-

OTRA: A Risk Management Ontology for Transparent Service Level Agreements in Federated Cloud Environment

123

agement of minor issues before escalation. In the

event of a violation, the “Compliance as Rem-

edy” (complianceAgreement123) corrective action

would allow both parties to restore SLA adherence

before triggering termination.

5 DISCUSSION

The ontology developed in this study was designed

to enhance risk representation and management in

SLAs within federated cloud environments. Through

our simulation of a digital marketplace, we validated

the ontology’s effectiveness in structuring and con-

veying complex SLA-related risk information to both

providers and consumers. By embedding our risk on-

tology within SLAs, we enabled participants to view

potential risks, preventive measures, and remediation

strategies explicitly, fostering transparency and in-

formed decision-making in cloud service interactions.

The structural integrity of the ontology was ver-

iﬁed using the Hermit 1.4.3.456 reasoner, ensuring

logical consistency both with and without instance

data. This technical validation conﬁrmed that the

ontology operates coherently within the OWL and

RDF frameworks, supporting consistent interpreta-

tion across diverse cloud environments.

A key outcome of the simulation was the ontol-

ogy’s ability to guide CSPs in incorporating preven-

tive and corrective mechanisms, even if these were

absent in the original SLA templates. By system-

atically modelling risk triggers and potential out-

comes, the ontology encouraged CSPs to adopt proac-

tive communication and compliance-based remedies.

This structured risk pathway not only deﬁnes actions

and consequences within SLAs but also enhances re-

silience by establishing actionable approaches for risk

mitigation.

We deem that the ontology has the potential to

foster trust within the federated cloud environments

by establishing a shared, transparent structure for risk

representation. Through standardised elements like

hazardous events, preconditions, preventive mecha-

nisms, postconditions, and remedies, the ontology

provides a clear structure that can reduce ambigu-

ity and improve mutual understanding. This trans-

parency is intended to give both consumers and

providers conﬁdence in the reliability of the frame-

work, as each party gains visibility into how risks are

identiﬁed, managed, and mitigated within the SLA it-

self prior to signing the contract.

Encouraging CSPs to adopt a structured, replica-

ble approach to risk management, the ontology has

the potential to create a consistent standard within

SLAs, making contractual terms and risk mitiga-

tion strategies more predictable and accessible across

cloud federations. This standardised approach could,

over time, support trust-building in federated cloud

environments by promoting transparency, account-

ability, and a shared commitment to proactive risk

management.

Furthermore, to address international variations in

legal requirements, the ontology’s explicit delineation

of risk offers an adaptable structure well-suited for

federated environments where CSPs and consumers

operate under different jurisdictions. Since SLAs and

legal frameworks vary widely between countries, the

clear representation of risks, preventive measures, and

remedies within the ontology can harmonise under-

standing across regions. By standardising risk ele-

ments within the SLA framework, the ontology fa-

cilitates consistent communication of terms, helping

to bridge differing legal expectations and regulatory

requirements.

This explicit risk delineation is particularly ad-

vantageous in federated cloud contexts where cross-

border collaborations are common. A transparent,

well-deﬁned ontology allows each party to understand

applicable risks and mitigation strategies, regardless

of jurisdictional differences. The ontology’s ﬂexibil-

ity and comprehensiveness provide a foundation for

CSPs to maintain compliance with local regulations

while fostering mutual understanding.

Ultimately, creating a comprehensive framework

for risk management in cloud federations is inher-

ently a collaborative endeavour. While this ontology

establishes a foundational structure for representing

and managing risks within SLAs, we recognise that

the dynamic nature of cloud environments demands

ongoing adaptation and enrichment. We encourage,

therefore, readers of this article to contribute to ex-

panding the ontology’s expressiveness, ensuring that

it continues to meet the evolving needs of federated

cloud ecosystems. Collaboration toward this shared

objective will be essential in building a robust, re-

liable setting for cloud federation, one where trans-

parency, accountability, and mutual trust are embed-

ded at the core of cloud service interactions.

6 FUTURE WORK

With the development of our ontology, we have es-

tablished a framework that allows for the explicit rep-

resentation of risks associated with SLAs in feder-

ated cloud environments, providing a structured view

of risks, preventive measures, and remediation strate-

gies before digitally signing agreements. This foun-

CLOSER 2025 - 15th International Conference on Cloud Computing and Services Science

124

dation facilitates transparency in risk communication

and fosters informed decision-making among cloud

service providers (CSPs) and consumers.

Moving forward, one key area of focus will be in-

tegrating this ontology into an expert system capable

of actively tracking the occurrence of preconditions

linked to potential hazards. This system could mon-

itor real-time data to detect risk factors or precondi-

tions as they emerge, triggering preventive mecha-

nisms proactively to avoid the manifestation of haz-

ardous events. By embedding such predictive func-

tionality, we aim to shift risk management from a

reactive to a more proactive stance, signiﬁcantly en-

hancing the resilience of federated cloud services.

In addition, we plan to develop a complementary

system for post-event management. Should a haz-

ardous event occur despite preventive measures, this

system would activate remediation strategies, mitigat-

ing the impact on associated postconditions. This

dual-component approach would provide a compre-

hensive risk management framework, covering both

preventive actions and responsive remedies within

SLAs.

Beyond these technical advancements, we envi-

sion continuous collaboration with industry stake-

holders and researchers to enrich the ontology’s ex-

pressiveness, ensuring it remains a robust, adaptable

resource that reﬂects emerging needs and complexi-

ties within the cloud federation domain.

ACKNOWLEDGEMENTS

This research activity was conducted from the Dutch

ECOFED project which is part of the integrated IP-

CEI CIS project. The early foundation of this research

was established during the Gaia-X Tech-X hackathon

in 2024, in which our participation was facilitated by

the Dutch Gaia-X hub part of the CoE DSC.

REFERENCES

Alkhamees, S. (2022). Sla negotiation and renegotiation

in cloud sla management: Issue and challenges. In

Hussain, W. and Jan, M. A., editors, IoT as a Ser-

vice, pages 159–170, Cham. Springer International

Publishing.

Allemang, D., Hendler, J., and Gandon, F. (2020). Semantic

Web for the Working Ontologist: Effective Modeling

for Linked Data, RDFS, and OWL, volume 33. As-

sociation for Computing Machinery, New York, NY,

USA, 3 edition.

Aust, J. and Pons, D. J. (2020). A systematic methodology

for developing bowtie in risk assessment: Application

to borescope inspection. Aerospace.

Bermbach, D., Kurze, T., and Tai, S. (2013). Cloud federa-

tion: Effects of federated compute resources on qual-

ity of service and cost. In 2013 IEEE International

Conference on Cloud Engineering (IC2E), pages 31–

37.

Blumauer, A., Nagy, H., and edition mono/monochrom

(2020). The Knowledge Graph Cookbook: Recipes

that Work. edition mono/monochrom.

Bonﬁglio, F. (2021). Gaia-x, vision & strategy. https://gaia-

x.eu/sites/default/ﬁles/2021-12/Vision & Strategy.pdf.

CEO Gaia-X, December 16, 2021.

De Filippi, P. and McCarthy, S. (2012). Cloud computing:

Centralization and data sovereignty. European Jour-

nal of Law and Technology, 3(2).

Hogan, A. (2020). Web of data. The Web of Data.

Holfelder, W., Mayer, A., and Baumgart, T. (2022).

Sovereign cloud technologies for scalable data spaces.

In Designing Data Spaces, pages 419–436. Springer.

Iorio, M., Risso, F., Palesandro, A., Camiciotti, L., and

Manzalini, A. (2022). Computing without borders:

The way towards liquid computing. IEEE Transac-

tions on Cloud Computing.

Karamanlioglu, A. and Alpaslan, F. N. (2018). An

ontology-based expert system to detect service level

agreement violations. In Shishkov, B., editor, Busi-

ness Modeling and Software Design, pages 362–371,

Cham. Springer International Publishing.

Kbah, Z., Erdil, N. O., and Aqlan, F. (2020). Risk assess-

ment in oil and gas industry using simulation and bow-

tie analysis. International Journal of Industrial Engi-

neering: Theory, Applications and Practice, 27(1).

Kluwer, W. (2017). The history of bowtie.

https://www.wolterskluwer.com/en/solutions/

enablon/bowtie/expert-insights/barrier-based-risk-

management-knowledge-base/the-historie-of-bowtie.

Kluwer, W. (2019). The bowtie method. https:

//www.wolterskluwer.com/en/solutions/enablon/

bowtie/expert-insights/barrier-based-risk-

management-knowledge-base/the-bowtie-method.

Kurze, T., Klems, M., Bermbach, D., Lenk, A., Tai, S., and

Kunze, M. (2011). Cloud federation. The Second In-

ternational Conference on Cloud Computing, GRIDs,

and Virtualization, pages 32–38.

Labidi, T., Mtibaa, A., and Brabra, H. (2016). Cslaonto:

A comprehensive ontological sla model in cloud com-

puting. Journal on Data Semantics, 5.

e, L.-S. and Nguyen, T.-V. (2020). Digitizing service level

agreements in service-oriented enterprise architecture.

SN Computer Science, 1(5):257.

McLeod, R. W. and Bowie, P. (2020). Guidance on cus-

tomising bowtie analysis for use in healthcare. In

Charles, R. and Golightly, D., editors, Contempo-

rary Ergonomics and Human Factors 2020, CIEHF.

CIEHF, CIEHF. Based on the CIEHF white paper

‘Human factors in barrier management’.

Mustafa, E. M., Saad, M. M., and Rizkallah, L. W. (2023).

Building an enhanced case-based reasoning and rule-

OTRA: A Risk Management Ontology for Transparent Service Level Agreements in Federated Cloud Environment

125

based systems for medical diagnosis. Journal of En-

gineering and Applied Science, 70(1):139.

Omidvar, M., Zarei, E., and Ramavandi, B. (2024). The

Bow-Tie Method: A Hybrid System Safety and Risk

Analysis Approach for Safety–Critical Sociotechnical

Systems, pages 123–149. Springer Nature Switzer-

land, Cham.

Pan, J. Z., Razniewski, S., Kalo, J.-C., Singhania, S., Chen,

J., Dietze, S., Jabeen, H., Omeliyanenko, J., Zhang,

W., Lissandrini, M., Biswas, R., de Melo, G., Boni-

fati, A., Vakaj, E., Dragoni, M., and Graux, D. (2023).

Large Language Models and Knowledge Graphs: Op-

portunities and Challenges. Transactions on Graph

Data and Knowledge, 1(1):2:1–2:38.

Prapakorn, N. and Chittayasothorn, S. (2009). An rdf-based

distributed expert system. WSEAS Transactions on

Computers, 8(5):788–798.

Pruvost, H., Wilde, A., and Enge-Rosenblatt, O. (2023).

Ontology-based expert system for automated monitor-

ing of building energy systems. Journal of Computing

in Civil Engineering, 37.

Roussey, C., Pinet, F., Kang, M. A., and Corcho, O.

(2011). Ontologies for Interoperability, pages 39–53.

Springer.

Uriarte, R. B., Tiezzi, F., and Nicola, R. D. (2016). Dynamic

slas for clouds. In Villari, M., Zimmermann, W., and

Jonas, K., editors, Service-Oriented and Cloud Com-

puting: 5th European Conference, ESOCC 2016, Pro-

ceedings, volume 9846 of Lecture Notes in Computer

Science, pages 34–49. Springer, Cham. First Online:

25 August 2016.

CLOSER 2025 - 15th International Conference on Cloud Computing and Services Science

126