=Paper=
{{Paper
|id=Vol-2205/paper22_bog1
|storemode=property
|title=The Role of Foundational Ontologies for Preventing Bad Ontology Design
|pdfUrl=https://ceur-ws.org/Vol-2205/paper22_bog1.pdf
|volume=Vol-2205
|authors=Stefan Schulz
|dblpUrl=https://dblp.org/rec/conf/jowo/Schulz18
}}
==The Role of Foundational Ontologies for Preventing Bad Ontology Design==
https://ceur-ws.org/Vol-2205/paper22_bog1.pdf
The Role of Foundational Ontologies for
Preventing Bad Ontology Design
Stefan SCHULZ a,b,1
a
Institute for Medical Informatics, Statistics and Documentation,
Medical University of Graz, Austria
b
AVERBIS GmbH, Freiburg, Germany
Abstract. Ontology engineering is error-prone, and many published ontologies
suffer from quality problems. This paper initiates a discussion about how
axiomatically rich foundational ontologies can contribute to prevent and to detect
bad ontology design. Examples T-boxes are presented, and it is demonstrated how
typical design errors can be detected by upper-level axioms, in particular disjoint
class axioms, existential and value restrictions. However, debugging large domain
ontologies under an expressive top level raises scalability issues. During domain
ontology design this can be mitigated by using small random ontology modules for
debugging. For reasoning in applications, however, less expressive variants of such
foundational ontologies are necessary.
Keywords. Foundational ontologies, description logics, quality assurance
1. Introduction
Constructing domain ontologies is a demanding endeavour. The formalization of basic
regularities of a domain requires not only familiarity with the domain but also
understanding of the basic principles of logic and formal ontology. This is especially the
case if ontology engineering is understood not as building (closed-world) models limited
to the support of well-delineated reasoning use cases in restricted domains, but as
providing interoperable and re-usable (open world) representations of the domain itself.
This is a basic principle stressed not only by the defenders of so-called realist
ontologies [1,2] but also by some (moderate) critics [3], which documents an increasing
consensus on how to represent those areas of knowledge where people tend to agree on
an observer-independent reality and benefit from standardised terms, such as in natural
science and technology domains. An important tenet of these ontologies is collaborative
ontology development and interoperability. Principles for this kind of ontology
development have been formulated by the OBO Foundry consortium [4] and within the
Good Ontology Design (GoodOD) guidelines [5]. Both propagate a concise foundational
upper-level as a mainstay for interoperability, and there is also some evidence that
foundational ontologies – domain-independent top-level or domain-related upper-level
ontologies – speed up ontology development and improve quality [6].
1
Corresponding Author: Stefan Schulz, Institute for Medical Informatics, Statistics and Documentation,
Medical University of Graz, Auenbruggerplatz 2/V, 8036 Graz, Austria, E-mail: steschu@gmail.com
This paper is intended to initiate a discussion on how foundational ontologies can
help prevent typical errors in domain ontologies. This is also the reason why prototypical,
partly made-up but easily understandable examples are used.
2. Materials and Methods
The work is based on BioTopLite (BTL2), an upper-level ontology [7], linked to BFO
[8], using description logics [9] with OWL-DL expressiveness [10]. BTL2 has been
designed with the intent to provide a rich set of constraining axioms to enforce the
consistency of ontologies modelled thereunder. Although BTL2 is, in principle, domain-
independent, its content is geared to the domains of health care and biomedical research.
This explains, e.g., the provision of more fine-grained classes for chemical and biological
entities (e.g. ‘mono molecular entity’, organism, cell, population) as well as the
disjunctive class condition, created in order to deal with the ontological heterogeneity of
key medical concepts like diseases, signs, and symptoms. Fig. 1 provides Protégé
screenshots of the class and relation hierarchies, together with sample axioms.
BTL2 Classes BTL2 Relations BTL Axioms (examples)
Figure 1. Classes, relations (object properties) and selected axioms in BioTopLite2 (BTL2)[7]
In order to test and demonstrate how BTL2 axioms are useful for preventing
ontology design errors, T-boxes with typical examples of bad modelling will be
presented in order to challenge the underlying foundational ontology. Some of these
examples are formulated very abstractly due to their high level of generality; others use
terms from a specific domain but are still understandable for a broader public.
Each T-box is modelled as an extension of BTL2. The HermIT [11] reasoner was
used to detect inconsistencies. The explanation of inconsistencies follows Protégé’s
OWL entailment explanation feature [12]. The examples are provided together with their
results in the following section. The presentation of the original OWL expressions, the
entailments and their justifications is done in the following order:
SRC – Axioms from ontology source (known satisfiable).
CHA – New axioms added that challenge the satisfiability of SRC. Cases
in which an axiom is only transiently added, are marked by CHA-T.
INF – Inference, in particular the detection of classes that are unsatisfiable
w.r.t. the T-box constituted by SRC and CHA.
EXP – Explanations of INF, with axioms collected from the explanation
plug-in [12].
3. Results
Reasoning examples under BTL2 are presented, in which deliberately erroneous axioms
lead to an incoherent ontology, i.e., where one or more named classes turn out to be
unsatisfiable, i.e. necessarily empty w.r.t. the T-box. This is expected to be detected by
a DL classifier. In the following, a distinction is made between five error types, viz. (1)
simple category errors, (2) value restrictions and transitive role errors, (3) complex
domain/range constraint errors, (4) physical granularity errors, and (5) errors regarding
unrealized realisables and non-referring information entities.
3.1. Simple category errors
A category error occurs whenever a class is a taxonomic descendent of upper-level
classes that are modelled as mutually exclusive (Disjoint Classes in OWL). Such errors
typically occur when ontology mapping and alignment is guided by lexical criteria. For
instance, when mapping content of the clinical ontology SNOMED CT [13] (namespace
sct:) to the foundational ontology BTL2 (namespace btl2:) one might be tempted to
equate sct:Process with btl2:process and place ‘sct: Qualifier Value (qualifier value)’
under ‘btl2:quality’:
SRC ‘sct:Process (qualifier value)’ SubClassOf (1a)
‘sct:Qualifier Value (qualifier value)’
NEW ‘sct:Qualifier Value (qualifier value)’ SubClassOf btl2:quality (1b)
NEW ‘sct:Process (qualifier value)’ EquivalentTo btl2:process (1c)
INF ‘sct:Process (qualifier value)’ EquivalentTo owl:Nothing (1d)
EXP btl2:quality DisjointWith btl2:process (1e)
The origin of such category errors may be manifold. A common cause is misleading
class labelling. Ontology labels should be context-independent and self-explanatory, and
they should avoid ambiguous terms [14]. A more severe problem – like here – arises
where the ontologies to be aligned fundamentally differ in upper level assumptions. In
SNOMED CT, e.g., the subhierarchy ‘sct:Qualifier Value (qualifier value)’ is currently
a badly organized reservoir for the most diverse terms, and the reason why hierarchies
of pathological and physiological processes are placed therein remains unclear 2.
3.2. Value restrictions and transitive roles
Value restrictions (universal constraints expressed by “only” in OWL Manchester
syntax) restrict the range of allowed role fillers. It is tempting to use value restrictions
together with mereological statements, e.g. stating that all members of a class have only
parts of a certain kind, like in the following example:
SRC btl2:cell SubClassOf btl2:compound (2a)
SRC ‘cell culture’ SubClassOf ‘btl2:material object’ (2b)
CHA ‘cell culture’ SubClassOf and (‘btl2:has part’ some btl2:cell) (2c)
and (‘btl2:has part’ only btl2:cell)
INF ‘cell culture’ EquivalentTo owl:Nothing (2d)
EXP ‘btl2:material object’ SubClassOf ‘btl2:has part’ some (2e)
‘btl2:subatomic particle’
EXP btl2:compound DisjointWith ‘btl2:subatomic particle’ (2f)
Axiom like (2c) may fulfil their purpose in domain ontologies in which cells are the
smallest objects, but as soon as smaller objects are allowed, the expression is inadequate
– assuming 'has part' being transitive. BTL2 obviates such a granularity restriction by
stating that all material objects have subatomic particles as (transitive) parts.
3.3. Complex domain / range restrictions
Domain / range restrictions are a suitable means to avoid ontology errors. However, in
ontologies that use a small number of object properties like BTL2, this resource may be
not expressive enough. For instance, if the relations ‘btl2:is part of’ and ‘btl2:has part’
are valid for material objects, immaterial objects, as well as for processes and information
objects, formulating just domain / range restrictions at the level of these relations would
still be compatible with an (unintended) model in which an object is part of a process or
vice versa. This is why BTL2 encodes these restrictions using axioms like the one in (3f).
SRC Object_A SubClassOf ‘btl2:material object’ (3a)
SRC Process_A SubClassOf btl2:process (3b)
CHA Object_A SubClassOf ‘btl2:is part of’ some Process_A (3c)
INF Object_A EquivalentTo owl:Nothing (3d)
EXP ‘btl2:has part’ InverseOf ‘btl2:is part of’ (3e)
EXP btl2:process SubClassOf ‘btl2:has part’ only btl2:process (3f)
EXP btl2:process DisjointWith ‘btl2:material object’ (3g)
2
Processes, like all kinds of entities, may play the role of values in information models, e.g., “Infectious
process”. Their confusion with their referents, i.e. domain entities proper (an infectious process in a patient) is
a classic case of use-mention confusion, which still haunts many domain ontologies [15], especially those
deriving from frames, thesauri and other knowledge organization systems.
A similar example is typical for medical ontologies, where domain experts are
tempted to use “diagnosis” and “disease” interchangeably, and where signs and
symptoms are referred to, in colloquial discourse, as being “parts” of diagnoses:
SRC Diagnosis_A SubClassOf Diagnosis (3h)
SRC Diagnosis SubClassOf ‘btl2:information object’ (3i)
SRC Symptom_A SubClassOf Symptom (3j)
SRC Symptom SubClassOf btl2:condition (3k)
CHA Diagnosis_A SubClassOf ‘btl2:has part’ some Symptom_A (3l)
INF Diagnosis_A EquivalentTo owl:Nothing (3m)
EXP ‘btl2:information object’ SubClassOf (3n)
‘btl2:has part’ only ‘btl2:information object’
EXP btl2:condition EquivalentTo btl2:function or btl2:disposition or (3o)
‘btl2:material object’ or btl2:process
EXP DisjointClasses: ‘btl2:material object’, btl2:process, btl2:function or (3p)
btl2:disposition, ‘btl2:information object’, ‘btl2:immaterial object’,
btl2:role, btl2:quality, ‘btl2:temporal region’, ‘btl2:value region’
3.4. Physical granularity
Ontologies for natural sciences and technology deal largely with physical objects of
several degrees of granularity. Levels of material granularity obey certain mereological
constraints, e.g. that biological cells can be parts of organisms but not vice versa, or that
polymolecular entities can never be part of single molecules. BTL2 incorporates such
constraints. They help detect modelling errors like the following one.
SRC Chromosome SubClassOf ‘btl2:poly molecular composite entity’ (4a)
SRC ProteinMolecule SubClassOf ‘btl2:mono molecular entity’ (4b)
CHA Chromosome SubClassOf ‘btl2:is part of’ some ProteinMolecule (4c)
INF Chromosome EquivalentTo owl:Nothing (4d)
EXP ‘btl2:poly molecular composite entity’ and (‘btl2:is part of’ some (4e)
(btl2:atom or ‘btl2:mono molecular entity’ or
‘btl2:subatomic particle’)) SubClassOf owl:Nothing
3.5. Unrealised realisables and non-referring information entities
Realisable entities like functions and dispositions [16] depend on material entities and
are realised in processes. However, the existence of realisables does not imply their
realisation: The function of a screwdriver is to drive screws, and the disposition of a glass
is to break under certain circumstances, but as there are screwdrivers that are never used
and glasses that are never thrown to the floor. Because for all types of functions and
dispositions there are instances that have never been realised, ontologies have to deal
with unrealised realisables, which could be, e.g., expressed by (5a).
SRC Unrealized_Function EquivalentTo btl2:function and not (5a)
(‘btl2:has realization’ some owl:thing)
In order not to preclude the possibility that dispositions and functions happen to be
never realised, ontologies under BTL2 should define them using the value restriction
constructor:
SRC btl2:function SubClassOf ‘btl2:has realization’ only btl2:process (5b)
SRC Function_A EquivalentTo btl2:function and (5c)
(‘btl2:has realization’ only Process_A)
Nevertheless, BTL2 does not reject a definition using an existential quantifier like in the
following definition:
CHA Function_A EquivalentTo btl2:function and (5d)
(‘btl2:has realization’ some Process_A)
Function_A would then exclude all unrealised function instances. Such a class (which
might be considered anti-rigid [17] if assuming that realisable entities are always
unrealised when they come into existence) is most likely not intended by the modeller.
The detection of these errors requires checking consistency after transiently adding
axiom (5e).
CHA-T btl2:function EquivalentTo Unrealized_Function (5e)
INF btl2:function EquivalentTo owl:Nothing (5f)
The explanation is given by the conjunction of axioms (5a), (5d), and (5e).
The axiomatization of non-referring information entities follows the same pattern.
Information entities can be referring and non-referring [18]. A typical example is medical
diagnosis [19]. BTL2 here uses the relation represents. This relation connects
information entities with domain entity they correctly characterise. This allows to
distinguish wrong diagnoses from correct diagnoses.
SRC Diagnosis SubClassOf ‘btl2:information object (5g)
SRC False_diagnosis SubClassOf Diagnosis and (5h)
(not btl2:represents some btl2:condition)
SRC Cancer_Diagnosis EquivalentTo Diagnosis and btl2:represents (5i)
only (Cancer or not btl2:condition)
CHA Cancer_diagnosis EquivalentTo Diagnosis and btl2:represents (5j)
some Cancer
Challenged by the axiom (5k) the T-box becomes incoherent.
CHA-T Diagnosis EquivalentTo False_Diagnosis (5k)
INF Cancer_Diagnosis EquivalentTo owl:Nothing
The explanation is given by the conjunction of axioms (5h), (5i), and (5k).
4. Discussion and Further Work
It was shown how a highly axiomatised foundational ontology like BioTopLite (BTL2)
can incorporate constraints that reject bad modelling decisions that lead to unsatisfiable
classes. A distinction was made between those cases in which the upper level axioms
suffice for detecting such inconsistencies and those in which additional “challenges”, i.e.
transiently added axioms are necessary.
Most of the former cases capitalise on disjoint class axioms present in the upper
level ontology. This comes near to the so-called logical anti-patterns introduced by [20],
all of which require disjoint class axioms in order to detect inconsistencies. In contrast
to the work presented, anti-patterns are very abstract logical expressions and independent
of foundational ontologies. OntoClean [17] was presented as a methodology for detecting
improper subclass axioms based on philosophically inspired, domain-independent
properties of classes, the metaproperties unity, identity and rigidity. Although DL
reasoners do not support meta-level reasoning, it has to be investigated whether certain
elements from OntoClean could also included in DL-based foundational ontologies, e.g.
by reifying them in terms of additional top-level classes3.
Several limitations of this work have to be highlighted:
The typology presented is certainly non-exhaustive. It is primarily motivated by
the author’s experience and not yet by the relevance of those types of problems
in ontologies employed in real-world applications. It could further be related to
existing work in ontology evaluation, e.g., the OQuaRE framework [21].
The proposed approach will probably fail if application ontologies bypass the
partition of upper categories of the foundational ontology or introduce new
object properties that are not subproperties of the existing ones. The BTL2
authors claim that their inventory of object properties is close to sufficient and
recommend to introduce predicates required by the domain (e.g., in the
biomedical domain: treats, prevents, diagnoses, interacts, binds) not as object
properties but as subclasses of btl2:process.
Important causes of bad ontology design cannot be prevented or remedied by a
foundational ontology. This includes erroneous representation of individuals as
classes or vice versa (a typical error would be an A-Box OWL expression like
“SodiumAtom Type btl2:atom), bad naming and insufficient documentation, as
well as constraints on a meta-class level like in OntoClean.
Constraining axioms similar to the proposed ones can be added to domain
ontologies, e.g. to assure mereotopological non-overlapping [22].
Although BTL2 was used as a testbed, the proposed approach would lend itself
to other ontologies as well. Especially BFO would benefit from a stronger
axiomatization, as the most popular version, which is the umbrella of most OBO
ontologies lacks axioms beyond subclass and disjoint class axioms. This
deficiency has been addressed by the 2.0 version, which is, however, not fully
available in OWL due to its use of ternary relations.
The fact that BTL2 uses the whole range of constructors allowed by OWL-DL
has a negative impact on reasoning performance. This makes debugging of large
ontologies intractable. This was the case when aligning SNOMED CT with
BTL2 [23]. The solution was to use small modules created from random
3
Which would be orthogonal to the existing ones, e.g. ‘rigid entity’‚ ‘anti-rigid entity’, ‘whole’
signatures as described in [24]. As a solution a two-step approach was proposed:
(i) at design time using the rich foundational ontology for debugging random
modules of a (large) domain ontology under construction, and (ii) at runtime
placing the final domain ontology under a light version of the same foundational
ontology for enabling efficient reasoning.
References
[1] I. Johansson, Bioinformatics and biological reality. J. Biomed Inform 39 (2006), 274–287
[2] B. Smith, W. Ceusters, Ontological realism: A methodology for coordinated evolution of scientific
ontologies. Applied Ontology 5 (2010) 139–188.
[3] P. Lord and R. Stevens. Adding a little reality to building ontologies for biology. PLoS one 5.9 (2010),
e12258.
[4] M. Ashburner, C. Rosse, J. Bard, W. Bug, W. Ceusters, L.J. Goldberg, K. Eilbeck, A. Ireland, C.J.
Mungall, the OBI Consortium, N. Leontis, P. Rocca-Serra, A. Ruttenberg, S. A. Sansone, R.H.
Scheuermann, N. Shah, P.L. Whetzel, S. Lewis, The OBO Foundry: coordinated evolution of ontologies
to support biomedical data integration, Nature Biotechnology 25 (2007), 1251.
[5] S. Schulz, D. Seddig-Raufie, N. Grewe, J. Röhl, D. Schober, M. Boeker, L. Jansen, Guideline on
Developing Good Ontologies in the Biomedical Domain with Description Logics,
http://www.purl.org/goodod/guideline , 2012.
[6] C. M. Keet. The use of foundational ontologies in ontology development: an empirical assessment.
Extended Semantic Web Conference. Springer, Berlin, Heidelberg (2011) 321-335.
[7] S. Schulz. M. Boeker, C. Martinez-Costa, The BioTop family of upper level ontological resources for
biomedicine, Stud Health Technol Inform. 235 (2017), 441–445.
[8] R. Arp, B. Smith, A. Spear, Building Ontologies with Basic Formal Ontology, MIT Press, 2015.
[9] F. Baader (ed.), The Description Logic Handbook, Second Edition. Cambridge University Press, 2010.
[10] W3C, OWL2 Web Ontology Language http://www.w3.org/TR/owl2-overview/ , 2012.
[11] B. Glimm et al. HermiT: an OWL 2 reasoner. Journal of Automated Reasoning, 3 (2014), 245–269.
[12] M. Horridge, B. Parsia and U. Sattler. Explanation of OWL entailments in Protégé 4. Proceedings of the
Poster and Demonstration Session at the 7th International Semantic Web Conference. CEUR-WS. org,
http://ceur-ws.org/Vol-401/iswc2008pd_submission_47.pdf , 2008.
[13] SNOMED international. SNOMED CT , https://www.snomed.org/snomed-ct , 2018
[14] D. Schober, B. Smith, S.E. Lewis, W. Kusnierczyk, J. Lomax, C. Mungall, C.F. Taylor, P. Rocca-Serra,
S.A. Sansone, Survey-based naming conventions for use in OBO Foundry ontology development, BMC
Bioinformatics 10.1 (2009), 125.
[15] B. Smith, Against idiosyncrasy in ontology development. Frontiers in Artificial Intelligence and
Applications 150 (2006), 15.
[16] J. Röhl, L. Jansen, Why functions are not special dispositions: an improved classification of realizables
for top-level ontologies. Journal of biomedical semantics 5.1 (2014), 27.
[17] N. Guarino, C.A. Welty. An overview of OntoClean. In Handbook on ontologies, 151–171, Springer,
Berlin, Heidelberg, 2004.
[18] S. Schulz, M. Brochhausen, R. Hoehndorf, Higgs bosons, Mars missions, and unicorn delusions: How to
deal with terms of dubious reference in scientific ontologies. Proc. of ICBO 2011. CEUR-WS 833
(2011), http://ceur-ws.org/Vol-833/paper24.pdf .
[19] S. Schulz, C. Martínez-Costa, D. Karlsson, R. Cornet, M. Brochhausen, A.L. Rector, An ontological
analysis of reference in health record statements. Formal Ontology in Inf. Systems (2014), 289–302.
[20] C. Roussey, O. Corcho, L. M. Vilches-Blázquez. A catalogue of OWL ontology antipatterns,
Proceedings of the fifth international conference on Knowledge capture. ACM, 2009.
[21] A. Duque-Ramos J.T. Fernández-Breis, R. Stevens, N. Aussenac-Gilles, OQuaRE: A SQuaRE-based
approach for evaluating the quality of ontologies. Journal of Research and Practice in Information
Technology 43.2 (2011), 159.
[22] M. Boeker, J. Hastings, D. Schober, S. Schulz (2011). A T-Box generator for testing scalability of OWL
mereotopological patterns. OWLED 2011 – OWL Experiences and Directions. Eight International
Workshop, San Francisco, California, USA, June 5-6, 2011.
[23] S. Schulz, C. Martínez-Costa, Harmonizing SNOMED CT with BioTopLite: An Exercise in Principled
Ontology Alignment. Stud Health Technol Inform. 216 (2015), 832–836.
[24] P. López-García, S. Schulz, Can SNOMED CT be squeezed without losing its shape? J Biomed
Semantics 7 (2016), 56.