=Paper=
{{Paper
|id=Vol-3637/paper41
|storemode=property
|title=Grid2Onto: An Application Ontology for Knowledge Capitalisation to Assist Power Grid
Operators
|pdfUrl=https://ceur-ws.org/Vol-3637/paper41.pdf
|volume=Vol-3637
|authors=Emna Amdouni,Mostepha Khouadjia,Maroua Meddeb,Antoine Marot,Laure Crochepierre,Walid Achour
|dblpUrl=https://dblp.org/rec/conf/jowo/AmdouniKMMCA23
}}
==Grid2Onto: An Application Ontology for Knowledge Capitalisation to Assist Power Grid
Operators==
https://ceur-ws.org/Vol-3637/paper41.pdf
Grid2Onto: An Application Ontology for Knowledge
Capitalisation to Assist Power Grid Operators
Emna Amdouni1,* , Mostepha Khouadjia1 , Maroua Meddeb1 , Antoine Marot2 ,
Laure Crochepierre2 and Walid Achour1
1
IRT SystemX, Palaiseau, France
2
RTE, R&D and AI Lab, Paris, France
Abstract
The development of real-time decision-making AI virtual assistant systems requires semantic artefacts
such as taxonomies, controlled vocabularies, and ontologies. These artefacts assist human operators in
dealing with heterogeneous information. This paper presents Grid2Onto, an application ontology that
leverages agent-oriented AI recommendations to aid power grid operators in solving future problems
based on past observations stored in a knowledge database. The main contribution is a unified semantic
model that formalises Grid2Op’s output, a realistic simulation environment for electrical supervision. The
proposed Grid2Onto ontology enhances a real-time power grid recommender system by automatically
generating a knowledge graph and reasoning capabilities. This paper highlights the added value of the
proposed ontology.
Keywords
Application ontology, ontology engineering in industrial domain, AI virtual power assistant system,
real-time simulation
1. Introduction
Power grids, such as the one managed by the French transmission system operator RTE have
always been complex artificial systems. Their complexity keeps increasing due to the current
energy transition. The integration of more intermittent renewable energies on the production
side and the emergence of prosumers on the demand side, coupled with the globalization of
energy markets, have made power grids more interconnected and challenging to operate [1].
Additionally, the grid is also ageing, but grid asset developments are facing social opposition.
As a result, operators must operate closer to the grid’s limits, deal with greater uncertainty, and
manage increasing grid automation.
In power grid control centres, AI assistance for real-time decision-making has emerged as
a promising solution to improve the efficiency, reliability, and safety of the electric grid [2].
AI assistance and real-time decision-making can improve the performance of power systems,
Ontology Showcase and Demonstrations Track, 9th Joint Ontology Workshops (JOWO 2023), co-located with FOIS 2023,
19-20 July, 2023, Sherbrooke, Québec, Canada
*
Corresponding author.
$ emna.amdouni@irt-systemx.fr (E. Amdouni)
0000-0002-2930-5938 (E. Amdouni); 0000-0003-4610-991X (M. Khouadjia); 0000-0003-1580-213X (M. Meddeb);
0000-0003-1753-5493 (A. Marot); 0000-0001-7400-765X (L. Crochepierre); 0009-0005-5902-4511 (W. Achour)
© 2023 Copyright for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
CEUR
Workshop
Proceedings
http://ceur-ws.org
ISSN 1613-0073
CEUR Workshop Proceedings (CEUR-WS.org)
CEUR
ceur-ws.org
Workshop ISSN 1613-0073
Proceedings
reduce the risk of blackouts, and improve the overall resiliency of the grid. In particular, manag-
ing congestion over power lines is critical. In the form of remedial action recommendations,
assistance is becoming utterly needed to deal with more numerous and complex events. More
precisely, the operator needs to be assisted during his operations to improve his performance
and reactivity to complex or unexpected situations. This assistance goes through the recom-
mendation of actions to be deployed and through the contextualization of the situations and
the explainability behind the recommended actions. Such assistance is explored through the
training of artificial agents [3, 4, 5], but gaps in terms of Human expert knowledge integration
and transparency remain for such solutions to be better aligned and fully adopted. By focusing
on energy management recommender systems, we can identify the lack of recommendation
systems which adopt the recent trends of explainable AI [6]. One of the main purposes of
ontologies is to capture knowledge about a domain and represent them in a machine-readable
and interpretable form. Here, the purpose is to help grid operators find the most relevant actions
to deploy regarding the contextual observations and the expected KPIs.
For that purpose, we propose a real-time ontology-based recommender system that can assist
operators in finding the relevant actions to operate to regulate and manage the power grid
with the help of the semantic Web techniques and knowledge graphs. We believe that using
ontologies facilitates the parsing, reasoning, sharing and reusing of knowledge to assist grid
operators in finding the most appropriate actions to regulate power grid-related issues. In
power grid management, ontologies and knowledge graphs are usually used to model higher
power grid concepts, their topology, physical components and their relationships, and the
logical rules governing these concepts’ functioning, constraints and requirements and their
relationships [7, 8]. It will also help improve the quality and personalisation of results via
recommendation systems and the understanding of these results.
This work describes the proposed Grid2Onto application ontology and showcases its usage
in a real-time power grid recommender system. The paper is organised as follows: Section 2
presents a brief overview of the research background and related works. Section 3 describes
the methodological and technological choices we considered in the development process of the
Grid2Onto application ontology. Section 4 details the main ontology development steps, i.e.,
requirements formalisation, conceptualisation, implementation, and evaluation, followed by the
development of the underlying ontology. Section 4 presents the results of a scripted incident
management scenario and demonstrates its usage. Finally, conclusions and directions for future
study are illustrated in the last section.
2. Background and Related Work
Recommender Systems are information filtering systems designed to assist users in finding
content, products, or services by collecting their preferences either implicitly or explicitly or by
analyzing their behaviours. In recent years, there has been a growing interest in semantic and
knowledge-based recommenders that suggest items based on specific domain knowledge about
how certain features meet users’ needs, preferences, and usefulness. Knowledge bases can be
used to build intelligent and explainable systems, and methods based on description logic can
provide new information through deductive reasoning in mathematically proven ways.
In literature, several ontologies and knowledge-based recommender systems have been de-
veloped for various domains, such as e-learning [9], entertainment [10], nutritional and health-
care [11], manufacturing [12], and driver assistance systems [13]. However, most recommender
systems are either collaborative or content-based in the power grid energy domain [14, 15]. On-
tologies have primarily been used to develop task or event ontologies for Energy Management
Systems (EMS) in the power systems domain [16, 17].
OntoPowSys [18] is an ontology-based management system within a Knowledge Management
System (KMS) for a virtual Eco-Industrial Park (EIP). It uses an ontology to demonstrate Optimal
Power Flow (OPF) in the electrical network and Cross Ontology Domain Interactions. The
knowledge base is implemented in OWL and is populated automatically from J-Park Simulator
data. Open Energy Ontology (OEO) [19] is an open OWL ontology that covers all aspects of the
energy modelling domain and has three modules covering models and data, social and economic
aspects, and the physical side of energy systems. The OEO-model module is for entities related
to data and models, OEO-social is for socio-economic entities, and OEO-physical includes all
entities related to the physical world of energy systems.
Ontologies have the capability to receive real-time data from various sources, including IoT
devices, sensors, logs, and virtual environments such as simulators, emulators, and virtual
reality. Raw data from underlying systems can be calibrated and simulated in practical industrial
scenarios with real data. For our study, we acquired data from the Grid2Op power grid simulator
and used it to develop the Grid2Onto ontology. This ontology captures the relationships
and concepts relevant to power system operations and covers electrical grid components,
observations, actions, and KPIs. By creating a knowledge graph and semantic model, Grid2Onto
serves as decision-making support for operators, allowing them to establish effective action
strategies based on current observations and expected KPIs. Gird2Onto has a well-specified
role within the Grid2Op framework, and the usage scenarios are restricted and very specific.
Considering this, we decided to implement the first version of grid2onto based only on the
experts’ and application needs, then extend it using existing standards such as the Common
Information Model 1 (CIM), mid-level ontologies such as the Sensor, Observation, Sample, and
Actuator (SOSA) Ontology [20] or initiatives such as the OEO.
The following section will describe the methodological approach used to construct the
proposed ontology. Our goal is to cover the key simulation outputs of Grid2Op, including
recommendations for actions based on observed situations and desired performance metrics.
3. Used Methodologies and Tools
In this study, we applied the realism-based approach [21, 22] to design and conceptualise
the proposed application ontology. By adopting such an approach, Gid2Onto can accurately
represent the entities, properties, and relationships observed in the electrical grid. Moreover,
we followed the Linked Open Terms (LOT) ontology engineering methodology [23] as a
basis to develop our ontology for capitalising the power grid observations. We selected this
approach because it is a specialised method for developing ontologies and vocabularies in
industry. LOT describes a main workflow composed of four major steps aligned with the
1
https://www.dmtf.org/standards/cim
software development practices, namely (i) ontology requirement specification, (ii) ontology
implementation, (iii) ontology publication, and (iv) ontology maintenance and recommends a list
of tasks and tools for each step. We are considering the FAIR (Findable, Accessible, Interoperable,
Reusable) principles [24] to facilitate the reuse and interoperability of the proposed application
ontology via the metadata properties curation.
Referring to LOT recommendations, we reused the Ontology Requirements Specification
Document (ORSD) templates 2 for collecting and formalizing the functional and non-functional
requirements, diagrams.net for the graphical conceptualization of the major classes and relations
of the proposed ontology schema, the Protégé ontology editor tool for the implementation
of an ontology in an OWL format, Pellet OWL-DL reasoner [25] for consistency checking
and validation, WebVowl (Web-based Visualisation of Ontologies) 3 for semantic network
visualization, WIDOCO (WIzard for DOCumenting Ontologies) 4 for generating HTML ontology
documentation, and Git system for ontology versioning and issue tracking.
As our goal is to automatically reason on the different Grid2Op5 simulation data, we developed
an end-to-end reproducible pipeline using the Owlready2 Python library [26] for handling data
RDFisation, knowledge graph generation and SPARQL querying in the developed power grid
operator’s recommender system. The main stakeholders for the ontology design, development
and integration within the virtual assistant system are:
• The Grid2Op and industrial power system experts contribute to defining the use case
scope and specify the specifications of the ontology.
• The ontology developers implement and test the ontology.
• The architect software engineers develop a real-time supervision framework referring to
the Grid2Onto application ontology.
Figure 1 depicts the main steps and I/O of the capitalisation pipeline within the assistant
platform and how the knowledge base is continuously enriched with new RDF triples (e.g., grid
state, context, actions, etc.) associated with an n+1 observed experience.
4. Development of the Grid2Onto Application Ontology
4.1. Ontology Requirements Specification
To elaborate on the ontology requirements, the ontology developer team interviewed Grid2Op
and industrial power system experts to specify the use case specifications, the documentation
needed, the purpose and scope of the ontology, and its requirements. The major purpose of
developing the Grid2Onto ontology is to provide a semantic description of observed actions
in a power network and a recommendation of future actions for optimizing some specific Key
Performance Indicators (KPIs). It will first cover two aspects, pointing to the relevant context
for applying an action, second, retrieving similar past situations for action. Thus, we identified
the following intended uses regarding the flow congestions in lines on a power grid: (i) Use 1:
2
https://github.com/oeg-upm/LOT-resources
3
https://vowl.visualdataweb.org/webvowl.html
4
https://github.com/dgarijo/Widoco/
5
https://github.com/rte-france/Grid2Op
Figure 1: The knowledge capitalisation process for assisting power grid operators within the recom-
mender system and its use of the Grid2Onto ontology
provide an “appropriate” recommendation using KPI: Given an identified action, add context
from observations associated with it to make a recommendation for the operator. (ii) Use 2:
Search for associated actions: Identify other related actions from a given action in this context.
(iii) Use 3: Rank a list of actions based on KPI value(s): Suggest to the operator ones with similar
or better KPI values among those actions. Ranking the actions. (iv) Use 4: Seek for similar
situations: Given an action and context, identify other situations in the past that were similar
to help the operator get “trust”.
Additionally, we collected the specification in the form of Competency Questions (CQs); in
total, we gathered a set of 34 CQs (sprint 1) and formalized them in an ORSD file. And we
defined a pre-glossary of main terms included in the CQs and calculated their frequencies.
Here-after some examples of CQs examples:
• CQ1: What is the current status of the power network grid?
• CQ2: What is the category of a specific detected issue?
• CQ3: What is the measurement unit of a specific parameter?
• CQ4: How many powerline issues are observed at a specific time interval?
• CQ5: What is the list of actions that should be conducted to optimise KPI values for a
specific context?
• CQ6: At what time was the highest value of the measured flow of each line?
• CQ7: What past situations best match the current context?
4.2. Conceptualisation
For the modelling part, a typical power system is considered as a set of power generators
and loads connected through an electrical network consisting of buses and lines. It also has
Figure 2: The meta-hierarchy of the Grid2Onto ontology (release V1.2)
devices like power converters to manage the quality and quantity of power flow and the voltage
level. In addition, inclusion axioms are defined to introduce the subsumption relations between
concepts, whereas role inclusion axioms are defined as subsumption relations between role
names. Finally, relations between individuals are established using concept and role names
(individual assertions). The results showcase the advantages of using ontologies in developing
decision support tools.
In particular, we followed a three-step process to develop the Grid2Onto ontology and
establish subsumption relationships between concepts. Firstly, we examined the technical
documentation of Grid2Op, with a focus on entities related to powerline flow congestion, such
as "rho," "powerline," "bus," and others. Secondly, we interviewed three experts in Grid2Op and
industrial power systems. Thirdly, we expanded our analysis by reviewing existing ontologies,
including the Information Entity Ontology, Agent Ontology, Quality Ontology, Units of Measure
Ontology, Event Ontology, Relation Ontology, etc., to incorporate mid-level content. Figure ??
depicts the structure of the provided in the ontology.
4.3. Ontology Coding, Evaluation and Publication
We coded the class and relation hierarchy (including the top-level relationships shown in
Figure 2, and then we formalised their definitions into the ontology using axioms (i.e., logical
assertions):
• HasPart(x,y): electrical_grid x hasPart electrical_grid_component y
• HasQuality (x,y): electrical_grid x has quality y
The current version (grid2onto-owl-v1.0) of the Grid2Onto has 121 classes, 25 object prop-
erties (with the inverse relations in count), 7 data properties, and 16 annotation metadata;
the ontology content is structured into five modules: infrastructure module (consists of 7
classes), Grid infrastructure module (consists of 23 classes), Grid operations module (consists
of 30 classes), stakeholders module (consists of 6 classes), environment module (consists of
Figure 3: A screenshot of automatic classification process of issue instances as a powerline overload
issue class using the equivalent class axiom definition (Protégé implementation)
23 classes). Following the W3C best practices, Grid2Onto adopts the HTTP namespace for
the URIs of all classes, properties, and individuals to facilitate the RDF querying. Each word
is lowercase and joined by underscores (e.g., electrical_grid). Thus the URIs are as follows:
http://www.semanticweb.org/emna.amdouni/ontologies/2023/1/Grid2Onto/electrical_grid
We formalised Grid2Onto knowledge using axioms, for example, the concept Power-
line_Overload_Issue = Issue and (is_about some powerline) and (has_measurement some (Rho and
has_value some xsd:float[>1])). Figure 3 illustrates how the resulted from ontology automatically
infers the category of a list of new issues based on their minimal asserted features.
During the evaluation step, we first checked the consistency of the ontology schema using
the Pellet reasoner as a Protégé plugin. Second, the populated ontology via the same reasoner
but the owlrady2 library. We validated its completeness/coverage regarding the defined set
of CQs associated with the current release of the power grid recommender. Grid2Onto (owl
version 1.2) is online and shared in the IndustryPortal 6 , an ontology repository for industrial
semantic artefacts. We aim to allow other users to reuse the ontology once we produce the
final release as decided internally in the ongoing French ANR CAB (Cockpit and Bidirectional
Assistant) project 7 .
5. Demonstration: Grid2Onto Application Scenario in the Power
Grid Recommender Platform
To perform real-time virtual assistance of RTE power grids based on Grid2Onto ontology, we
developed a Web application based on JavaScript, Angular, and Python technologies. Figure
6
http://industryportal.enit.fr
7
https://www.irt-systemx.fr/en/projets/cab-cockpit-and-bidirectional-assistant
Figure 4: An overview of the power grid recommender system enabling semantic reasoning and
recommendation for RTE power grid operators
4 details the architecture of the power grid recommender system, which is composed of five
major parts:
1. Assistant connectors that implement the entry points used to receive context data and
events from the Grid2Op simulator,
2. Context and notification service that displays the electrical grid with its current line
status and the reported notifications about issues such as powerline overload,
3. History service that tracks all events and contexts received from the simulator and used
to populate the ontology,
4. Knowledge acquisition and reasoning service that stores the new/updated RDF triples that
are automatically generated according to the Grid2Onto model and performs semantic
requesting using SPARQL,
5. Recommendation service that returns a list of recommendations based on the knowledge
and reasoning service in the case of an observed issue.
As illustrated in Figure 4, retrieving recommendations from the RDF triple store is part of the
data layer. The sequence diagram in Figure 5 details the whole process from the overload alarm
notification to the RDF data retrieved from collected JSON-LD information.
The selection of relevant recommendations based on the observed context relies on the
defined SPARQL queries that are handled by the knowledge acquisition and reasoning service,
an example is provided hereafter:
S e l e c t d i s t i n c t ? s i m i l a r I s s u e ? l i n e ? rhovalue ? pastAction
{
? s i m i l a r I s s u e a Grid2Onto : p o w e r l i n e _ o v e r l o a d _ i s s u e .
Figure 5: Sequence UML diagram for RDF data retrieved from collected JSON-LD information
? s i m i l a r I s s u e Grid2Onto : i s _ a s s o c i a t e d _ w i t h ? p a s t A c t i o n .
? rho a G r i d 2 O n t o : Rho .
? rho G r i d 2 O n t o : h a s _ v a l u e ? r h o v a l u e .
? rho G r i d 2 O n t o : i s _ a b o u t ? l i n e .
? s i m i l a r I s s u e G r i d 2 O n t o : h a s _ m e a s u r e m e n t ? rho .
}
Listing 1: A SPARQL query example to retrieve similar powerline issue and their associated
number of the line, rho value, and past action from the acquired knowledge in the RDF triple
store.
This section describes a demonstration of a powerline incident management scenario. We
launch the Grid2Op simulator to push context and event data to the recommender system. As
displayed in Figure 6, the assistant allows us to visualise the real-time state of the grid network.
When an overload is detected, an alarm notification is raised to inform the operator about a
powerline issue (a). The operator clicks on the notification for more details. The grid network
plots the observed overload with a red line and information related to the line identifier and the
rho value (b). The operator can ask the recommender system for help to resolve this issue. The
assistant then displays recommendations for the observed powerline issue (c).
6. Conclusion and Perspectives
This work introduces a promising application ontology called Grid2Onto, conceived to be a
semantic reference model for the grid2op RTE simulator, and aims to capitalise on the power
grid operator’s knowledge. The proposed ontology is integrated into an assistant platform
showcasing the industrial use cases identified during interviews.
Figure 6: The role of Gird2Onto ontology in the assistance platform and its usage in the real-time
recommendation process for RTE power grid operators, (a) ranked list of powerline overload issue
notifications, (b) current grid network plot highlighting the observed powerline status, (c) list of returned
AI recommended actions for resolving the selected issue, and (d) timeline of issues management
The preliminary results demonstrate some reasoning advantages of using semantic web
technologies in developing real-time decision support tools in power systems. The industrial
use case experts validated the current ontology recommendation results.
The proposed Grid2Onto is an ongoing French ANR RD project output, and the model is under
frequent updating. In the future, the ontology modules will be based on existing relevant open
ontologies, such as SOSA ontology, to cover specific domain knowledge (agent, event, quality,
units of measure, etc.). And we will follow an upper-level ontology, for example, Basic Formal
Ontology (BFO)[27] to harmonise the different representations and assess the FAIRness using
specific tools [28] to improve the quality of our proposal. Additionally, efforts to expand the
recommender system’s reasoning capabilities are needed to fully cover the Grid2Op simulation
parameters and generate more appropriate recommendations based on a given observed power
grid context.
Acknowledgments
We acknowledge the support of the French government through the CAB (Cockpit and Bidi-
rectional Assistant) project, which is being conducted as part of the SystemX Technological
Research Institute and the IA2 French national program. We are grateful to Kahina Amokrane
Ferkat and Virgil Rousseaux (SystemX) for their assistance in designing and validating the IHM
and to Alexandre Rozier (RTE) for his contributions to the specification of the industrial use
case and his participation in the ontology requirement specification workshops.
References
[1] A. Marot, A. Kelly, M. Naglic, et al., Control centers for energy transition, Journal of Modern
Power Systems and Clean Energy 10 (2022) 328–344. doi:10.35833/MPCE.2021.000673.
[2] A. Marot, A. Rozier, M. Dussartre, L. Crochepierre, B. Donnot, Towards an ai assistant
for power grid operators, in: Hybrid Human Artificial Intelligence (HHAI), 2022. doi:10.
3233/FAIA220191.
[3] A. Marot, B. Donnot, C. Romero, et al., Learning to run a power network challenge
for training topology controllers, Electric Power Systems Research 189 (2020) 106635.
doi:https://doi.org/10.1016/j.epsr.2020.106635.
[4] A. Marot, B. Donnot, K. Chaouache, et al., Learning to run a power network with
trust, Electric Power Systems Research 212 (2022) 108487. URL: https://www.sciencedirect.
com/science/article/pii/S0378779622006137. doi:https://doi.org/10.1016/j.epsr.
2022.108487.
[5] A. Marot, B. Donnot, G. Dulac-Arnold, et al., Learning to run a power network challenge:
a retrospective analysis, in: H. J. Escalante, K. Hofmann (Eds.), Proceedings of the NeurIPS
2020 Competition and Demonstration Track, volume 133 of Proceedings of Machine Learning
Research, PMLR, 2021, pp. 112–132.
[6] F. K. Došilović, M. Brčić, N. Hlupić, Explainable artificial intelligence: A survey, in:
2018 41st International Convention on Information and Communication Technology,
Electronics and Microelectronics (MIPRO), 2018, pp. 0210–0215. doi:10.23919/MIPRO.
2018.8400040.
[7] R. Liu, R. Fu, K. Xu, et al., A review of knowledge graph-based reasoning technology in
the operation of power systems, Applied Sciences 13 (2023). doi:10.3390/app13074357.
[8] S. Fan, X. J. Liu, Y. Chen, et al., How to construct a power knowledge graph with dispatching
data?, Sci. Program. 2020 (2020) 8842463:1–8842463:10.
[9] C. Obeid, I. Lahoud, H. El Khoury, P.-A. Champin, Ontology-based recommender system
in higher education, International World Wide Web Conferences Steering Committee,
Republic and Canton of Geneva, CHE, 2018.
[10] P. Sheridan, M. Onsjö, C. Becerra, et al., An ontology-based recommender system with an
application to the star trek television franchise, Future Internet 11 (2019). doi:10.3390/
fi11090182.
[11] D. Mckensy-Sambola, M. Rodríguez-García, F. García-Sánchez, R. Valencia-García,
Ontology-based nutritional recommender system, Applied Sciences 12 (2022). URL:
https://www.mdpi.com/2076-3417/12/1/143. doi:10.3390/app12010143.
[12] M. Alkahtani, A. Choudhary, A. De, J. A. Harding, A decision support system based on
ontology and data mining to improve design using warranty data, Computers Industrial
Engineering 128 (2019) 1027–1039. doi:https://doi.org/10.1016/j.cie.2018.04.
033.
[13] A. Armand, D. Filliat, J. Ibañez-Guzman, Ontology-based context awareness for driving
assistance systems, in: 2014 IEEE Intelligent Vehicles Symposium Proceedings, 2014, pp.
227–233. doi:10.1109/IVS.2014.6856509.
[14] Y. Zhang, K. Meng, W. Kong, Z. Y. Dong, Collaborative filtering-based electricity plan
recommender system, IEEE Transactions on Industrial Informatics 15 (2019) 1393–1404.
doi:10.1109/TII.2018.2856842.
[15] R. K. Patel, A. Kumari, S. Tanwar, W.-C. Hong, R. Sharma, Ai-empowered recommender
system for renewable energy harvesting in smart grid system, IEEE Access 10 (2022)
24316–24326. doi:10.1109/ACCESS.2022.3152528.
[16] S. Fan, X. Liu, Y. Chen, Z. Liao, Y. Zhao, H. Luo, H. Fan, How to construct a power
knowledge graph with dispatching data?, Scientific Programming 2020 (2020) 1–10.
[17] R. Liu, R. Fu, K. Xu, X. Shi, X. Ren, A review of knowledge graph-based reasoning
technology in the operation of power systems, Applied Sciences 13 (2023) 4357.
[18] A. Devanand, G. Karmakar, N. Krdzavac, et al., Ontopowsys: A power system ontology
for cross domain interactions in an eco industrial park, Energy and AI 1 (2020) 100008.
doi:https://doi.org/10.1016/j.egyai.2020.100008.
[19] M. Booshehri, L. Emele, S. Flügel, et al., Introducing the open energy ontology: Enhancing
data interpretation and interfacing in energy systems analysis, Energy and AI 5 (2021)
100074. doi:https://doi.org/10.1016/j.egyai.2021.100074.
[20] K. Janowicz, A. Haller, S. J. Cox, D. Le Phuoc, M. Lefrançois, Sosa: A lightweight ontology
for sensors, observations, samples, and actuators, Journal of Web Semantics 56 (2019)
1–10.
[21] B. Smith, W. Ceusters, Ontological realism: A methodology for coordinated evolution of
scientific ontologies, Appl. Ontol. 5 (2010) 139–188.
[22] B. Smith, New desiderata for biomedical terminologies, in: K. Munn, B. Smith (Eds.),
Applied Ontology: An Introduction, Ontos, 2008, pp. 83–109.
[23] M. Poveda-Villalón, A. Fernández-Izquierdo, M. Fernández-López, R. García-Castro, Lot:
An industrial oriented ontology engineering framework, Engineering Applications of
Artificial Intelligence 111 (2022) 104755. doi:https://doi.org/10.1016/j.engappai.
2022.104755.
[24] M. D. Wilkinson, M. Dumontier, I. J. Aalbersberg, et al., The fair guiding principles for
scientific data management and stewardship, Scientific Data 3 (2016) 160018–.
[25] E. Sirin, B. Parsia, B. C. Grau, et al., Pellet: A practical owl-dl reasoner, Journal of Web
Semantics 5 (2007) 51–53. doi:https://doi.org/10.1016/j.websem.2007.03.004,
software Engineering and the Semantic Web.
[26] J.-B. Lamy, Owlready: Ontology-oriented programming in python with automatic clas-
sification and high level constructs for biomedical ontologies, Artificial Intelligence in
Medicine 80 (2017) 11–28. doi:https://doi.org/10.1016/j.artmed.2017.07.002.
[27] R. Arp, B. Smith, A. D. Spear, Building Ontologies With Basic Formal Ontology, The MIT
Press, 2015. doi:10.7551/mitpress/9780262527811.001.0001.
[28] E. Amdouni, S. Bouazzouni, C. Jonquet, O’faire: Ontology fairness evaluator in the
agroportal semantic resource repository, in: The Semantic Web: ESWC 2022 Satellite
Events, Springer International Publishing, Cham, 2022, pp. 89–94.