Elementary Multiperspective Material Ontology: Leveraging

Perspectives via a Showcase of EMMO-Based Domain

and Application Ontologies

Pierluigi Del Nostro

1 a

, Jesper Friis

2 b

, Emanuele Ghedini

3 c

, Gerhard Goldbeck

1 d

Oskar Holtz

, Otello Maria Roscioni

1 e

, Francesco Antonio Zaccarini

3 f

and Daniele Toti

1,4 g

Goldbeck Consulting Limited, Cambridge, U.K.

SINTEF AS, Trondheim, Norway

Alma Mater Studiorum, University of Bologna, Bologna, Italy

Catholic University of the Sacred Heart, Brescia, Italy

{pierluigi, gerhard, otello, daniele}@goldbeck-consulting.com, oskar-holtz@hotmail.com, jesper.friis@sintef.no,

Keywords:

Upper-Level Ontology, Domain Ontology, Application Ontology, Perspectives, Semiotics, Holism,

Materials Science.

Abstract:

The effectiveness of semantic technologies in ensuring interoperability is often hindered by the preference

for internally developed knowledge bases and the presence of diverse conceptual frameworks and implemen-

tation choices. Foundational, upper-level ontologies based on FOL and OWL2-DL address interoperability

and provide a robust foundation for domain and application ontologies. They emphasize logical rigor and ex-

pressiveness, aligning with the idea of shared ontologies for knowledge diffusion and reuse. In scientiﬁc and

industrial contexts, a framework that accommodates scientiﬁc pluralism is essential. The Elementary Mul-

tiperspective Material Ontology (EMMO) meets this need, offering a rigorous yet pluralistic representation

of knowledge through the mereocausal theory, focusing on parthood (mereology) and causation. EMMO’s

adaptable architecture includes discipline-speciﬁc modules, enabling the representation of items from mul-

tiple perspectives, such as viewing an image as both an ’Object’ and ’Data’. This paper presents EMMO’s

perspectives, including the Reductionistic, Holistic, Persistence, Contrast, Structural and Semiotics perspec-

tives. It then proceeds to showcase four recently-developed ontologies based on EMMO, one at the domain

level (CHAMEO) and three at the application level (BTO, HPO and MAEO), taking advantage of EMMO’s

perspectives and therefore demonstrating its representational capabilities and versatility.

1 INTRODUCTION

The practical effectiveness of semantic technologies

in supporting interoperability is hindered by the pre-

vailing tendency to favor internally developed Knowl-

edge bases, in an effort to exert greater control over

proprietary data. Furthermore, different conceptual

frameworks and implementation choices (related to

trade-offs between expressiveness and computational

https://orcid.org/0000-0002-5174-8508

https://orcid.org/0000-0002-1560-809X

https://orcid.org/0000-0003-3805-8761

https://orcid.org/0000-0002-4181-2852

https://orcid.org/0000-0001-7815-6636

https://orcid.org/0009-0008-8009-5009

https://orcid.org/0000-0002-9668-6961

efﬁciency Levesque and Brachman (1987)), exhibit

varying degrees of suitability for speciﬁc domains

of application and use-cases. In literature, it is

a common practice to classify ontologies hierarchi-

cally based on the generality of the concepts they in-

clude, i.e. their domain of application. These “lev-

els” have vague boundaries, but the presence of bor-

derline cases does not undermine the practical utility

of this classiﬁcation criterion. By deﬁnition, a do-

main ontology focuses on concepts, properties, and

relationships relevant to a speciﬁc area of knowledge

or ﬁeld of study. A domain ontology can either be

a specialized module of an upper-level ontology or a

standalone ontology targeting a speciﬁc domain (e.g.

additive manufacturing, composite materials). On the

other hand, an application ontology is an ontology en-

Del Nostro, P., Friis, J., Ghedini, E., Goldbeck, G., Holtz, O., Roscioni, O., Zaccarini, F. and Toti, D.

Elementary Multiperspective Material Ontology: Leveraging Perspectives via a Showcase of EMMO-Based Domain and Application Ontologies.

DOI: 10.5220/0012910200003838

In Proceedings of the 16th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2024) - Volume 2: KEOD, pages 135-142

ISBN: 978-989-758-716-0; ISSN: 2184-3228

135

gineered for a speciﬁc use or application focus, and

whose scope is usually speciﬁed or driven through

speciﬁc use cases (De Baas et al., 2023). Things are

further complicated by the fact that domain and ap-

plication ontologies tend to be highly context-speciﬁc

and lean towards technologies, such as triplestores,

that prioritize handling large volumes of instance data

over supporting expressive knowledge frameworks,

increasing the risk of mistakes in conceptualization.

Upper-level ontologies are usually axiomatized

in expressive formal languages (such as FOL and

OWL2-DL (W3C, 2012)) and can be considered the

explicit expression of a precise worldview. These

ontologies are commonly employed both to provide

a foundation for domain and application ontologies

and to address interoperability issues among differ-

ent schemas Trojahn et al. (2021). They can be con-

sidered the true heirs of the original idea of employ-

ing a common ontology to enable knowledge shar-

ing and reuse across different systems Gruber (1993),

a notion more recently expressed by Guarino et al.

(2009). Indeed, Gangemi et al. (2002) explored the

use of formal ontologies in semantic web applica-

tions, highlighting the signiﬁcance of logical rigor

and expressiveness in ontology design. Remarkably,

neuro-symbolic AI may soon enable the full exploita-

tion of richly axiomatized ontologies in practical set-

tings, addressing one of the core challenges deter-

ring practitioners from employing upper-level, foun-

dational ontologies, as exempliﬁed by Lazzari et al.

(2024) and discussed in Bouraoui et al. (2019).

Addressing the issues arising from the plurality of

frameworks presents a signiﬁcant challenge, particu-

larly in scientiﬁc and industrial contexts. These con-

texts seem to demand a framework that is not only

ﬁrmly rooted in the sciences but also accommodates

scientiﬁc pluralism – recognizing that multiple inter-

pretations or standards may exist for the same phe-

nomena or processes. The Elementary Multiperspec-

tive Material Ontology (EMMO)

, speciﬁcally, is an

expression of the common tenets and general world-

view central to the applied sciences. It aims to cater

to both practical needs and theoretical desiderata of

practitioners, striving to be both a standard represen-

tational framework for science and engineering and a

versatile ontology for broader applications.

At its core, EMMO is built on a mereocausal

theory, developed to support a rigorous yet pluralis-

tic representation of scientiﬁc and industrial knowl-

edge. This theory hinges on two formal relations,

parthood (mereology) and causation, which provide

clear extensional criteria of identity (Partridge et al.,

2020) and set up the preconditions for the qualitative

Available at https://github.com/emmo-repo/EMMO/.

and quantitative representation of systems at differ-

ent granularities, up to a high level of detail. More

ﬁne-grained, yet subjective, distinctions among en-

tities are expressed by means of exploiting perspec-

tives, a special feature of EMMO’s architecture, de-

signed to include a plurality of tools to categorize en-

tities, as well as to support user-driven downward ex-

pansions. In fact, EMMO includes various discipline-

speciﬁc modules, such as metrology and units (based

on BIPM et al. (2012)), materials, and manufactur-

ing (based on ISO and DIN standards). These mod-

ules ensure that EMMO can address speciﬁc domain

needs while maintaining a high level of expressive-

ness and interoperability. For instance, it allows for

the representation of an item from different perspec-

tives, such as viewing an image both as an ‘Object’

resulting from an imaging process and as ‘Data’.

This paper presents EMMO’s perspectives and

showcases a number of recently-developed domain

and application ontologies that are based on EMMO

and take advantage of its representational capabilities

and perspectives. The paper is structured as follows:

Section 2 describes EMMO’s perspectives, includ-

ing the Reductionistic, Holistic, Persistence, Contrast,

Structural and Semiotics perspectives; Section 3 lists

a selection of 4 ontologies (1 domain ontology and 3

application ontologies) based on EMMO, underlining

their main characteristics and their usage of EMMO’s

perspectives; ﬁnally, Section 4 draws the conclusions.

2 PERSPECTIVES IN EMMO

EMMO includes and embraces a plurality of per-

spectives, representing real-world objects according

to speciﬁc representational viewpoints. These per-

spectives allow for subjective categorization of enti-

ties above elementary particles and below the item

‘Universe’, each constituting a subclass of ‘perspec-

tive’. This approach reﬂects pluralism in carving out

portions of the world, according to different principles

and in line with diverse interests. In EMMO, only

entities described by the standard model of particle

physics have univocal deﬁnitions and clear identity

criteria: ‘physical objects’ are conceptualized as ‘spa-

tiotemporal objects composed of fundamental physi-

cal constituents’, without referring to any subjective

perspective. The different perspectives are pragmati-

cally chosen to increase expressiveness and avoid the

core limitations of reductionistic approaches. Fig-

ure 1 shows EMMO’s perspectives, which are de-

scribed in greater detail in the subsections below.

KEOD 2024 - 16th International Conference on Knowledge Engineering and Ontology Development

136

Figure 1: EMMO’s architecture, with its backbone based on mereocausality and its implementation of physics, chemistry

and materials, grounded in current natural science foundations (StandardModel). This core is complemented by modules

incorporating various perspectives: Reductionistic (tiling of space and/or time), Holistic (whole/part or role), Persistence

(process/object), Contrast (based on Floridi’s work), Structural (typed mereological relations) and Semiotics (based on Peirce).

2.1 Reductionistic Perspective

The Reductionistic perspective offers a powerful

granularity description of multi-scale entities, deter-

mined by the direct parthood relation (a non-transitive

relation). Differently from the standard transitive no-

tion of parthood employed in mereology, direct part-

hood makes it possible to individuate levels for the

analysis of entities. Every macro-individual, includ-

ing the Universe, can be iteratively tessellated down to

elementary constituents, as it is most convenient for

speciﬁc use cases. Tessellation based on spatial and

temporal criteria can be especially salient in speciﬁc

scenarios (e.g. temporal parts for the individuation of

steps in a productive workﬂow).

2.2 Holistic Perspective

The Holistic perspective offers a tool to represent re-

lations between entities whose categorization under

speciﬁc types intrinsically depends on part-whole re-

lations. The holistic class introduces important con-

cepts such as ‘role’ and/or ‘participant’, used to de-

scribe the parts of a particular system or process, re-

spectively. This enables the description of parts in

terms of how they contribute to the whole, includ-

ing functional aspects. This perspective allows for an

analysis of entities in mereological relations without

imposing a strict granularity, and is thus complemen-

tary to the reductionistic perspective.

2.3 Persistence Perspective

The Persistence perspective classiﬁes entities accord-

ing to how they persist in time Sider (2001) with re-

spect to speciﬁc types under which they are catego-

rized, allowing for a distinction between objects (for

which a type is conserved through all temporal parts)

and processes (for which it is not). This categoriza-

tion method is deeply ingrained in common sense and

entrenched in natural languages. The distinction pri-

marily hinges on whether the emphasis is placed on

the preservation of characteristics (indicative of ob-

jects) or patterns in the evolution (indicative of pro-

cesses) of spatiotemporal regions, given a principle of

salience. It also provides classes aimed to represent

concepts similar to the ones of endurant/continuant

and perdurant/occurrent, facilitating connections with

other upper-level ontologies (e.g. BFO (Arp et al.,

2015)), supporting interoperability.

2.4 Contrast Perspective

The Contrast perspective draws from Luciano

Floridi’s philosophy of information, given the idea

that data can be broadly understood as distinctions

that make a difference (Floridi, 2010). It encompasses

various types of data (semantic, environmental...) and

enhances EMMO’s expressive power by allowing it to

represent different forms of data and information.

Elementary Multiperspective Material Ontology: Leveraging Perspectives via a Showcase of EMMO-Based Domain and Application

Ontologies

137

2.5 Structural Perspective

The Structural perspective introduces formal tools for

the analysis of parthonomic relations between entities

of the same or different types. In practical scenarios,

this perspective can be used, for instance, to distin-

guish between a car’s base model and its accessories.

2.6 Semiotics Perspective

The Semiotics perspective takes inspiration from

Peirce’s theory of signs (Atkin, 2023) (and their tri-

adic structure of object, sign-vehicle, and interpre-

tant) and covers expressive needs that are usually

dealt with by using abstract objects such as numbers,

sets, and universals/tropes. This perspective encom-

passes all of the semantic and symbolic ‘world’, cov-

ering labeling, property attributions and abstractions.

Compared to Peirce, in EMMO more emphasis is put

on the role of interpreters and the underlying causal

connections between the three ‘prongs’, making it

possible to rigorously analyze observation and esti-

mation, and to keep track of interpreters’ possibly in-

consistent attributions. This is especially useful in op-

erative contexts where it is important to keep track of

the gap between data and phenomena, to circumscribe

disagreement (Bogen and Woodward, 1988). Not

only does this perspective greatly enhance EMMO’s

expressive power, but it also affords mappings to on-

tologies that utilize qualities and/or properties.

3 SHOWCASE OF FOUR

EMMO-BASED ONTOLOGIES

A number of domain and application ontologies re-

volving around EMMO that are part of its ecosystem

have been developed so far and/or are currently un-

der development. Here, one domain ontology and

three application ontologies recently developed by

this study’s authors are listed and brieﬂy showcased,

underlining their alignment with EMMO and their ex-

ploitation of EMMO’s expressiveness via the use of

perspectives. A description of each ontology and spe-

ciﬁc details about their use of EMMO’s perspectives

are provided in the following subsections.

3.1 Domain Ontology: CHAMEO

The CHAracterization MEthodology Ontology

(CHAMEO) (Del Nostro et al., 2022a,b) is a domain-

level ontology designed to offer a harmonized and

standardized representation of materials character-

ization methods and processes. The development

of CHAMEO is grounded in the CHADA (CHAr-

acterization DAta) document template, subject of

a 2021 CEN Workshop Agreement (CWA 17815),

which aims to provide a standard structure for

documenting material characterization techniques.

CHADA employs four key concepts to classify the

steps of a characterization workﬂow: “User case”

(information about the sample and testing envi-

ronment), “Experiment” (characterization process

including probe, signal, detector, noise, etc.), “Raw

data” (output of the metrological process), and “Data

processing” (analysis of the data to achieve the ﬁnal

shape). These concepts correspond to the sections

of a CHADA document. While this document was

easily interpretable by humans, it lacked structured

data for retrieving information on characterization

methodologies according to their various dimensions

(e.g. material, probe, detector, properties). In order

to address this limitation, CHADA was used as the

foundation to build a more structured and shared

knowledge base, leading to the creation of the

CHAMEO ontology. CHAMEO models the common

aspects of diverse characterization techniques by

providing high-level, methodological deﬁnitions,

facilitating the development of more specialized

ontologies at a ﬁner-grained application level. This

capability is enhanced by the intrinsic modularity

of its ontological design, positioning CHAMEO at

the domain level of a larger ontological framework,

with EMMO as its upper layer. As part of EMMO’s

framework, CHAMEO leverages EMMO’s versatility

in describing processes and data from multiple

perspectives. This integration allows CHAMEO to

model various aspects of characterization techniques

comprehensively, using EMMO’s multiperspective

approach to enhance its descriptive capabilities.

Consequently, CHAMEO is well-suited for modeling

the shared aspects of characterization methodologies,

providing a robust foundation for further ontology

development and application-speciﬁc adaptations.

3.1.1 EMMO’s Perspectives in CHAMEO

In CHAMEO

, the chameo:Characterisation

Workflow class models the overall characterization

process as a subclass (further down in the hierar-

chy) of emmo:Process within the Persistence per-

spective. Characterization workﬂows consist of var-

ious methods, each contributing to the ﬁnal character-

ization property. These methods are represented by

the chameo:CharacterisationMethod class, which

CHAMEO’s names for classes and properties follow

the Cambridge English naming convention, especially no-

table by the use of s instead of z, whereas this work com-

plies with the Oxford University Press English convention.

KEOD 2024 - 16th International Conference on Knowledge Engineering and Ontology Development

138

is itself a subclass (further down in the hierarchy)

of emmo:Process as well as emmo:Observation

from the Semiotics perspective; each method is

thus a process (Persistence), but also a part of

the chameo:CharacterisationWorkflow (Holis-

tic). Each chameo:CharacterisationMethod

is divided into stages (subclass of emmo:Stage)

capturing general steps common across differ-

ent characterization techniques. For instance,

the chameo:SpecimenPreparation stage refers to

the preparation of a specimen, where the holder

used is modeled as an emmo:Object (Persistence)

and plays a role in the process (emmo:Role,

Holistic). Specializations for speciﬁc techniques

and temporal sequences are expressed through

EMMO’s hasTemporalCause property. The char-

acterization system is modeled as a subclass of

emmo:HolisticSystem (Holistic), which is the sys-

tem used for characterization, built by assembling

and adapting characterization components. In-

stances of chameo:CharacterisationInstrument,

chameo:Probe, chameo:Detector, chameo:Holder

and all of the possible devices that can be

used to build a characterization system can be

deemed a chameo:CharacterisationComponent

(Persistence). The different types of data pro-

duced during the characterization process (e.g.

chameo:CharacterisationData) are subclasses of

emmo:EncodedData (Contrast perspective). The

characterization properties are assigned meaning via

a semiotic process (Semiotics perspective).

3.2 Application Ontology: Battery

Testing Ontology (BTO)

The Battery Testing Ontology (BTO) (Nostro et al.,

2024) is a standardized and ﬂexible framework for

representing knowledge in the areas of battery test-

ing and quality control. BTO’s purpose is to

model a range of electrical battery cell tests, such

as impedance spectroscopy (monitoring current and

voltage over time), self-discharge (current over time),

and high-voltage tests (voltage over time). BTO can

specify the necessary hardware for measuring speciﬁc

battery cell properties, like the quality of the separa-

tor layer in a high-voltage test. It references electrical

measurement data (voltage, current, time) and incor-

porates details about the mechanical ﬁxturing of the

battery cell to the test hardware and the electrical cal-

ibration procedures. In high-voltage separator tests,

BTO can specify what hardware is suitable for test-

ing a particular battery separator to achieve a speciﬁc

accuracy. It also allows users to verify if their test

hardware and voltage speciﬁcations can measure cer-

tain battery samples (e.g. separator layers with differ-

ent thicknesses), considering the maximum test volt-

age and required accuracy. Therefore, BTO assists in

designing high-voltage separator test experiments by

combining separator requirements (e.g. thicker sepa-

rators need a higher maximum test voltage), hardware

speciﬁcations (e.g. the tester’s voltage speciﬁcations),

and the required measurement accuracy. BTO oper-

ates at the application level of EMMO’s ecosystem

of ontologies. BTO aligns closely with EMMO and

CHAMEO, the latter providing common domain con-

cepts for characterization methods and experiments.

Additionally, BTO is connected and aligned with

other EMMO-based domain ontologies, including the

Battery Domain Ontology (BDO)

and the Electro-

chemistry Domain Ontology (EDO)

3.2.1 EMMO’s Perspectives in BTO

In BTO, the battery testing process is modeled as

a bto:BatteryCharacterizationMethod that

is made up of a number of potential steps, or

tasks, namely bto:BatterySamplePreparation,

bto:CalibrationForBatteryCharacterization

and bto:BatteryMeasurementProcess. Each

of these four classes is a subclass of a CHAMEO

class, i.e. chameo: CharacterisationMethod,

chameo:SamplePreparation, chameo:

CalibrationProcess and chameo:

CharacterisationMeasurementProcess, re-

spectively. Therefore, given what has been discussed

in 3.1.1, bto:BatteryCharacterizationMethod

is also a subclass of emmo:Process as well

as emmo:Observation (Semiotics perspective),

whereas each task is also a subclass of emmo:Process

(Persistence), and at the same time a part of the over-

all battery characterization method (Holistic). Fur-

thermore, the ambient in which a testing takes place

is modeled as a bto:BatteryCharacterization

Environment, which is a subclass of chameo:

Characterisation Environment, which in turn

is a emmo:SemioticObject (Semiotics); each

characteristic of the ambient is modeled as a

bto:CharacterizationEnvironmentProperty,

which is an emmo:Property, which in turn is

an emmo:Sign from the Semiotics perspective;

and the hardware used for the battery testing is

modeled as a bto:BatteryCharacterization

Hardware, which is a subclass of

chameo:CharacterisationInstrument (Persis-

tence perspective). The data resulting from a battery

https://github.com/emmo-repo/domain-battery.

https://github.com/emmo-repo/

domain-electrochemistry.

Elementary Multiperspective Material Ontology: Leveraging Perspectives via a Showcase of EMMO-Based Domain and Application

Ontologies

139

testing experiment, i.e. bto:TraceData, are modeled

as a subclass of chameo:CharacterisationData,

which is a subclass of emmo:EncodedData and thus

takes advantage of EMMO’s Contrast perspective.

3.3 Application Ontology:

Hyperdimensional Polymer

Ontology (HPO)

The Hyperdimensional Polymer Ontology (HPO) is

designed to capture the staggering diversity of poly-

meric materials and their applications, with a partic-

ular focus on their manufacturing aspects. HPO has

been developed primarily to describe the manufactur-

ing process of carbon-ﬁber-reinforced polymers and

includes a rich taxonomy of materials, processes and

properties for this application. The terms used to la-

bel the main classes are taken from a glossary for

polymers and composite materials compiled from the

Compendium of Chemical Terminology (IUPAC) and

specialized journals.

The fundamental requirement of HPO is to

describe materials as different as composite ma-

terials, energy materials and biomaterials. The

EMMO backbone, based on the classiﬁcation

of real-world objects based on physical sci-

ences, is exploited here by deﬁning polymers

as subclasses of emmo:PolymericMaterial,

emmo:ManufacturedMaterial, or emmo:Natural

Material, plus additional restrictions to make

further distinctions. Indeed, polymeric materials

can be classiﬁed in many different ways, depending

on the characteristics relevant to a speciﬁc appli-

cation. For example, polymeric materials can be

distinguished based on their chemical composition

(hpo:HomoPolymer vs. hpo:CoPolymer), struc-

ture (hpo:BranchedPolymer, hpo:Crosslinked

Polymer, or hpo:LinearPolymer), or mechani-

cal behavior. Another aspect covered by HPO is

the description of the manufacturing process of

composite materials, which includes the tools used

(mixers, molds, covers), the processes (impregnation,

molding, curing), and of course, any other material

used (prepregs, thinners, ﬁllers, catalysts).

3.3.1 EMMO’s Perspectives in HPO

The classiﬁcation of polymers by categories is based

on multiple inheritance, obtained by declaring a class

as a subclass of disconnected classes. For example,

an organic semiconductor polymer such as P3HT

can be classiﬁed as a hpo:HomoPolymer based on

the chemical composition, a hpo:CouplingPolymer

based on the polymerization mechanism, as well

as a hpo:LinearPolymer based on the chemical

structure. To achieve this expressivity, HPO uses

the multiperspective character of EMMO to enrich

the generic concept of emmo:PolymericMaterial

with predicates increasing its speciﬁcity. For

instance, the Contrast perspective provides the

emmo:ChemicalComposition class that ac-

commodates the classes hpo:HomoPolymer

and hpo:CoPolymer, whereas the concept of

emmo:Behaviour in the Persistence perspective

is used to classify polymers based on their me-

chanical properties. The Semiotics perspective

has a taxonomy of measured physical properties

(under the emmo:PhysicalQuantity class) used

to classify speciﬁc physical observables char-

acterizing the manufacturing process, e.g. the

hpo:PrepregGlassTransitionTemperature and

hpo:DegassingStepDuration. The manufacturing

aspects of the fabrication of polymers and composite

materials are also covered in the Persistence per-

spective, with classes like hpo:ShapingAndCuring

being part of the emmo:Manufacturing class, and

hpo:PrePreg and hpo:ResinMixer belonging to the

emmo:ManufacturedProduct family.

3.4 Application Ontology: MarketPlace

Agent and Expert Ontology

(MAEO)

The MarketPlace Agent and Expert Ontology

(MAEO) (Del Nostro et al., 2023; Goldbeck and Toti,

2021)

is designed to model experts, expertise, and

the broader community of knowledge providers and

seekers within the domain of Materials Modeling, and

was developed as part of the “MarketPlace” European

project, meant to create an online platform as a cen-

tral hub for connecting scientiﬁc and industrial stake-

holders with their own expertise. MAEO supports

this platform by providing a structured framework to

represent and manage the necessary information as

an application ontology of EMMO’s ecosystem. It

aligns with other ontologies, including Friend-Of-A-

Friend (FOAF) (Brickley and Miller, 2014) and ﬁve

EMMO-based domain ontologies, i.e. the Open Inno-

vation Environment (OIE) ontologies

, for classifying

materials, models, manufacturing processes, charac-

terization methods and software products. Designed

to meet the needs of the MarketPlace project’s stake-

holders, MAEO leverages the expressive power and

standardization efforts of EMMO, ensuring interop-

erability and promoting consistency and integration

https://github.com/emmo-repo/MAEO-Ontology.

https://github.com/emmo-repo/OIE-Ontologies/.

KEOD 2024 - 16th International Conference on Knowledge Engineering and Ontology Development

140

Table 1: Selection of EMMO-based ontologies: CHAMEO, BTO, HPO, MAEO. For each ontology, its acronym, the URI

with which it is accessible, its GitHub repository and the perspectives used from EMMO are reported.

EMMO-based ontologies

Acronym URI GitHub repository EMMO’s perspectives FAIR

score

CHAMEO https://w3id.org/emmo/domain/

characterisation-methodology/

chameo

https://github.com/emmo-repo/

domain-characterisation-methodology

Holistic, Persistence,

Contrast, Semiotics

100%

BTO http://w3id.org/emmo-bto/bto https://github.com/emmo-repo/

battery-testing-ontology

Holistic, Persistence,

Contrast, Semiotics

100%

HPO http://w3id.org/emmo-hpo/hpo https://github.com/emmo-repo/

hyperdimensional-polymer-ontology

Persistence, Contrast,

Semiotics

100%

MAEO http://w3id.org/emmo-maeo/

maeo

https://github.com/emmo-repo/

MAEO-Ontology

Holistic, Persistence,

Semiotics

100%

across various domains within Materials Modeling.

3.4.1 EMMO’s Perspectives in MAEO

In MAEO, an agent, i.e. a person or an entity

that is able to operate on the MarketPlace plat-

form, (a concept grouping knowledge providers and

knowledge seekers, including human experts, labo-

ratories, teams and organizations) is modeled as a

experts:MarketPlaceAgent, which is a subclass of

emmo:Participant that is seen as both an object

from the Persistence perspective and a participant in

a process from the Holistic perspective. Knowledge

providers possess a range of properties, divided into

two main categories, subjective properties and objec-

tive properties. A subjective property is inherently

non-well-deﬁned and cannot be unambiguously de-

termined; it relies on an agent acting as a “black box”

for its deﬁnition. In MAEO, the expertise of knowl-

edge providers is considered a subjective property be-

cause it cannot be objectively deﬁned and requires

validation by an agent, such as a certiﬁcation author-

ity or employer. Conversely, an objective property

can be determined through a well-deﬁned procedure

and observed via a speciﬁc perception mechanism,

making it measurable. In MAEO, objective proper-

ties include information about a knowledge provider

that can be clearly and objectively identiﬁed, such

as personal details (e.g. address), professional de-

tails (current and past employments), certiﬁcations

acquired, contractual details, and afﬁliations to orga-

nizations or companies. These two concepts are mod-

eled as expert:ExpertSubjectiveProperty and

expert:ExpertObjectiveProperty, respectively,

deﬁned as subclasses of emmo:Property, which is a

subclass of emmo:Sign, i.e. something that stands for

an object through, for instance, a convention, thus em-

bracing EMMO’s Semiotics perspective.

3.5 FAIR Score and Technical Details

All of the ontologies described above have been tested

for pitfalls and issues via the OOPS tool (Poveda-

Villal

on et al., 2014)

, and no pitfalls were detected

Furthermore, the ontologies’ compliance with the

FAIR principles (Findability, Accessibility, Interoper-

ability, Reusability) has been veriﬁed via the FOOPS!

tool (Garijo et al., 2021)

, and all of them have ob-

tained a perfect score of 100%. The four ontologies’

preﬁxes are stored in the preﬁx.cc repository

and

have their corresponding entry in the Linked Open

Vocabulary (LOV)

registry; the ontologies’ ﬁles are

stored in a corresponding GitHub repository. Table 1

shows a summary of the four ontologies; for each

ontology, its acronym, the URI with which it is ac-

cessible, its GitHub repository, the perspectives from

EMMO used and its FAIR score are reported.

4 CONCLUSIONS

This work has emphasized EMMO’s role in deﬁning

a comprehensive framework to meet the needs of sci-

entiﬁc and industrial contexts, accommodating scien-

tiﬁc pluralism. EMMO’s foundation in mereocausal

theory allows for a rigorous yet pluralistic represen-

tation of knowledge, with a focus on fundamental re-

lations such as parthood and causation. EMMO’s ar-

chitecture, notable for its modularity and adaptabil-

ity, includes discipline-speciﬁc modules that ensure

high expressiveness and interoperability. These mod-

ules enable the representation of items from various

perspectives, enhancing the ontology’s versatility. By

https://oops.linkeddata.es/index.jsp.

OOPS needs the direct URL of the ontology as input.

https://foops.linkeddata.es/FAIR validator.html.

https://preﬁx.cc.

https://lov.linkeddata.es/dataset/lov.

Elementary Multiperspective Material Ontology: Leveraging Perspectives via a Showcase of EMMO-Based Domain and Application

Ontologies

141

leveraging EMMO’s representational capabilities and

perspectives, this work has shown how four recently-

developed domain and application ontologies, i.e.

CHAMEO, BTO, HPO and MAEO, tackle different

application scenarios. EMMO offers an adaptable

framework for developing highly expressive domain

and application ontologies. By deﬁning foundational

classes and properties and tracing them back to funda-

mental axioms of parthood, causation, persistence and

semiotics, EMMO ensures a comprehensive knowl-

edge representation, crucial for the advancement of

both semantic web technologies and neuro-symbolic

AI, and paves the way for improved interoperability

across diverse scientiﬁc and industrial domains.

ACKNOWLEDGEMENTS

This work was supported by the European Union’s

Horizon 2020 Research and Innovation Programme,

via NanoMECommons (G. A. n. 952869) and Open-

Model (G. A. n. 953167).

REFERENCES

Arp, R., Smith, B., and Spear, A. D. (2015). Building on-

tologies with basic formal ontology. MIT Press.

Atkin, A. (2023). Peirce’s theory of signs.

https://plato.stanford.edu/archives/spr2023/entries/

peirce-semiotics/. The Stanford Encyclopedia of

Philosophy (Spring 2023 Edition).

BIPM, IEC, IFCC, ILAC, ISO, IUPAC, IUPAP, and OIML

(2012). International vocabulary of metrology — Ba-

sic and general concepts and associated terms (VIM).

Technical report.

Bogen, J. and Woodward, J. (1988). Saving the phenomena.

The Philosophical Review, 97(3):303–352.

Bouraoui, Z., Cornu

ejols, A., Denoeux, T., Destercke, S.,

Dubois, D., Guillaume, R., Marques-Silva, J., Men-

gin, J., Prade, H., Schockaert, S., Serrurier, M., and

Vrain, C. (2019). From shallow to deep interac-

tions between knowledge representation, reasoning

and machine learning (kay r. amel group). CoRR,

abs/1912.06612.

Brickley, D. and Miller, L. (2014). Friend-Of-A-Friend.

http://xmlns.com/foaf/spec/.

De Baas, A., Nostro, P. D., Friis, J., Ghedini, E., Goldbeck,

G., Paponetti, I. M., Pozzi, A., Sarkar, A., Yang, L.,

Zaccarini, F. A., and Toti, D. (2023). Review and

Alignment of Domain-Level Ontologies for Materials

Science. IEEE Access, 11:120372–120401.

Del Nostro, P., Goldbeck, G., Pozzi, A., and Toti, D. (2023).

Modeling experts, knowledge providers and expertise

in Materials Modeling:MAEO as an application on-

tology of EMMO’s ecosystem. Applied Ontology,

18(2):99–118.

Del Nostro, P., Goldbeck, G., and Toti, D. (2022a).

CHAMEO: An ontology for the harmonisation of ma-

terials characterisation methodologies. Applied Ontol-

ogy, 17(3):401–421.

Del Nostro, P., Goldbeck, G., and Toti, D. (2022b). The

CHAMEO Ontology: Exploiting EMMO’s Multiper-

spective Versatility for Capturing Materials Character-

ization Procedures. In Proceedings of the 12th Inter-

national Workshop on Formal Ontologies meet Indus-

try (FOMI 2022), volume 3240 of CEUR Workshop

Proceedings. CEUR-WS.org.

Floridi, L. (2010). Information: A very short introduction.

OUP Oxford. p. 23.

Gangemi, A., Guarino, N., Masolo, C., Oltramari, A., and

Schneider, L. (2002). Sweetening ontologies with

DOLCE. In Knowledge Engineering and Knowl-

edge Management: Ontologies and the Semantic Web,

pages 166–181. Springer Berlin Heidelberg.

Garijo, D., Corcho, O., and Poveda-Villal

on, M. (2021).

FOOPS!: An Ontology Pitfall Scanner for the FAIR

Principles. 2980.

Goldbeck, G. and Toti, D. (2021). MAEO: An ontology

for modeling Agents, Experts and Expertise within

an open online Materials Modeling MarketPlace. In

CEUR Workshop Proceedings, volume 2969, pages

253–263.

Gruber, T. R. (1993). A translation approach to portable

ontology speciﬁcations. Knowledge Acquisition,

5(2):199–220.

Guarino, N., Oberle, D., and Staab, S. (2009). What is an

ontology? In Handbook on ontologies, pages 1–17.

Springer.

Lazzari, N., Giorgis, S. D., Gangemi, A., and Presutti, V.

(2024). Sandra - a neuro-symbolic reasoner based on

descriptions and situations. ArXiv, abs/2402.00591.

Levesque, H. J. and Brachman, R. J. (1987). Expressive-

ness and tractability in knowledge representation and

reasoning1. Computational Intelligence, 3(1):78–93.

Nostro, P. D., Goldbeck, G., Kienberger, F., Moertelmaier,

M., Pozzi, A., Al-Zubaidi-R-Smith, N., and Toti, D.

(2024). Battery Testing Ontology: an EMMO-based

Semantic Framework for Representing Knowledge in

Battery Testing and Battery Quality Control. Comput-

ers in Industry.

Partridge, C., Mitchell, A., and West, M. (2020). A sur-

vey of Top-Level Ontologies: To inform the ontologi-

cal choices for a Foundation Data Model.

Poveda-Villal

on, M., G

omez-P

erez, A., and Su

arez-

Figueroa, M. C. (2014). OOPS! (OntOlogy Pitfall

Scanner!): An On-line Tool for Ontology Evaluation.

International Journal on Semantic Web and Informa-

tion Systems (IJSWIS), 10(2):7–34.

Sider, T. (2001). Four-dimensionalism: An ontology of per-

sistence and time. Oxford University Press.

Trojahn, C., Vieira, R., Schmidt, D., Pease, A., and Guiz-

zardi, G. (2021). Foundational ontologies meet ontol-

ogy matching: A survey. Semantic Web, 13.

W3C (2012). Web Ontology Language. https://www.w3.

org/OWL/.

KEOD 2024 - 16th International Conference on Knowledge Engineering and Ontology Development

142