=Paper=
{{Paper
|id=Vol-2931/ICBO_2019_paper_14
|storemode=property
|title=Ontology as Product-Service System: Lessons Learned from GO, BFO and DOLCE
|pdfUrl=https://ceur-ws.org/Vol-2931/ICBO_2019_paper_14.pdf
|volume=Vol-2931
|authors=Barry Smith
|dblpUrl=https://dblp.org/rec/conf/icbo/Smith19
}}
==Ontology as Product-Service System: Lessons Learned from GO, BFO and DOLCE==
https://ceur-ws.org/Vol-2931/ICBO_2019_paper_14.pdf
Ontology as Product-Service System:
Lessons Learned from GO, BFO and DOLCE
Barry Smith
Department of Philosophy and National Center for Ontological Research, University at Buffalo, Buffalo, NY, USA
Abstract The computable representation must be shared. Ontology
This paper defends a view of the Gene Ontology (GO) and of development is inherently collaborative.
Basic Formal Ontology (BFO) as examples of what the manu- Ensure that there is access to help. Does a warm body
facturing industry calls product-service systems. This means answer help email within a reasonable time (say 2 working
that they are products bundled with a range of ontology ser- days)?
vices such as updates, training, help desk, and permanent iden-
Every ontology improves when it is applied to actual
tifiers. The paper argues that GO and BFO are contrasted in
instances of data.
this respect with DOLCE, which approximates more closely to
a scientific theory or a scientific publication. The paper pro- There will be fewer problems in the ontology and more
vides a detailed overview of ontology services and concludes commitment to fixing remaining problems when important
with a discussion of some implications of the product-service research data is involved that scientists depend upon.
system approach for the understanding of the nature of applied The basic thesis underlying these principles is that an ontology
ontology. Ontology developer communities are compared in becomes more valuable to the extent that it is aggressively used.
this respect with developers of scientific theories and of stand- Ashburner and his associates accordingly devised a multi-
ards (such as W3C). For each of these we can ask: what kinds pronged strategy designed to maximize GO usage, including:
of products do they develop and what kinds of services do they 1. developing a simple ontology editing tool (called OBO-
provide for the users of these products? Edit) designed to suit the needs of biologists (3)
Keywords 2. making the GO easy to find by placing it in the public
Basic Formal Ontology, Domain Ontology for Linguistic and domain
Cognitive Engineering, product-service system 3. making the results of using the GO in annotating biological
literature and data easy to find, by creating the GO
Preamble Annotation (GOA) database (4) and associated software
tools
The success of an ontology can be measured using a range of 4. providing a set of evidence codes, which allow literature
metrics, including number and variety of associated software and data curators to record the types of evidence (for
applications; quantities of data and literature annotated using instance experimental, phylogenetic, computational) on
terms from the ontology; and number, size and degree of utili- which their annotations are based (5)
zation of major databases incorporating terms from the ontol- 5. ensuring sustainability (6), for example (i) by ensuring that
ogy. By any of these metrics, the Gene Ontology (GO) (1) is the the ontology provides permanent identifiers (7), that it is
world’s most successful scientific ontology. In a tutorial pre- updated speedily in light of advances in science and in the
sented at the Intelligent Systems for Molecular Biology (ISMB) needs of users, and (ii) by providing easy and enduring
conference in 2005, Michael Ashburner and Suzanna Lewis for- online access to all successive versions of the GO
mulated a set of “Principles of Biomedical Ontology Construc-
tion” (2) extracted from their experience in developing the GO. 6. providing assistance in creation and use of persistent
These principles can, I believe, help us to account for the re- identifiers and repositories for ontology content (8)
markable success of the GO, which still, after 20 years, contin- 7. providing online user forums and help desk, and an issue
ues on its upward trajectory as concerns numbers of users and tracker that allows users to report errors or omissions in the
applications, and scientific influence and utility. ontology and to obtain rapid feedback
The most important of the Ashburner-Lewis principles for our 8. responding to needs of users by creating new ontologies
purposes here are: developed in such a way as to interoperate with the GO
Before you start building an ontology learn what is out and with each other
there. 9. making these ontologies easy for biologists to access by
creating an open portal: initially the GOBO (Global Open
Assess extant ontologies critically and realistically. Do not
Biology Ontologies) portal, launched already in 2001 (9);
reinvent. Start building – but not in isolation. Collaborate.
Copyright © 2019 for this paper by its author. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
now the OBO (Open Biological and Biomedical facturers, there is a cottage industry of smaller companies that
Ontologies) Foundry at http://obofoundry.org. offer repair services to its products. The latter may indeed bring
10. ensuring an effective modular architecture for these benefits. They may be quicker, and cheaper, than GE itself. At
ontologies, including providing each ontology with a name the same time, however, such companies contribute to a frag-
(such as “Cell Ontology”, “Protein Ontology”) that would mentation of the data landscape in ways that may bring adverse
make it easily findable by new users consequences for users in the future.
11. ensuring an effective division of labor by devising Product-Service System Business Models
procedures to support resolution of overlaps between
The business model underlying the activities of GE, as well as
ontology modules and an editorial process that allows each
– somewhat later – IBM, Microsoft, Adobe, is that each of these
OBO ontology to be managed by scientists with
organizations wants to provide services. They are, in effect, us-
corresponding subject-matter expertise
ing the products that they create as delivery mechanisms for
12. contributing to the development of ontology software (for services. This model contributes to a move away from the
example ROBOT (10) and the Ontozoo tools (11)) throw-away approach characteristic of manufacturing in the
13. maintaing a reliable license regime to provide legal second half of the 20th century, to a situation in which one prod-
certainty for users and reusers of the ontologies. uct is maintained in active use for as long as possible and is
This strategy has yielded (and continues to yield) a positive continually updated to ensure a maximally productive life.
snowball effect whereby, when one community encounters As Mont points out (13), to achieve this goal
gaps or errors when using the GO or one of its sister ontologies, requires a higher level of customer involvement and education by
these gaps and errors are rapidly fixed. This increases the value producers. For producers and service providers, product-service sys-
of the ontology, thereby making it attractive to further users, tems mean a higher degree of responsibility for the product’s full
who in turn identify further gaps and errors thereby initiating a life cycle, the early involvement of consumers in the design of the
new cycle of improvement. Such fixes occur in a way that pre- product-service system, and design of the closed-loop system.
serves the integrity of the GO as a resource on which existing A product-service system is thus not simply the result of asso-
and future users can rely as it develops over time and addresses ciating products with services. Rather, it constitutes an archi-
new sets of user needs. tecture where artifacts and processes are deliberately designed
in such a way that they fit together in a single system, which
Product-Service Systems (PSS) may involve multiple enterprises and multiple user communi-
ties joined together in complex networks.
While the role played by ontology services is familiar to all on-
tology practitioners, such services have not thus far been an ex- The GO Product-Service System
plicit topic of theoretical reflection. In the advanced manufac-
turing industries, in contrast, a set of parallel developments has In an interesting parallel to such developments in the advanced
engendered a new organizational paradigm with considerable manufacturing and commercial software industries, we can now
impact both practically and theoretically. see that the GO developer community has been providing the
Gene Ontology to its users as a (free) product bundled with a
The term ‘product-service system’ was introduced in the early range of user services. This reflects the fact that the incentives
2000s, when producers of, for instance, aircraft engines or pho- of the GO developers were from the very beginning aligned
tocopying machines had been experimenting with new business with the incentives of its users. Indeed, many of these users
models focused on the marketing of bundled services. formed part of the very same GO Consortium that was respon-
On one model, GE does not sell engines. Rather it charges for sible for developing and maintaining the ontology itself.
the use of its engines per hour of flying time. As pointed out in From the very beginning the GO pursued a strategy of assisting
(12), this fosters better alignment of incentives between GE and its users in solving the problems that arise, for example, when
its customers, both of whom want to minimize the amount of a new release of the ontology involves changes that might dis-
downtime for unscheduled maintenance. This creates a second- rupt existing workflows, or when new scientific results or new
order incentive on the part of GE to learn as much as possible sorts of data arise which need to be accommodated within the
about the reasons for engine failure and to ensure that lessons GO and GOA frameworks. Both GO developers and GO users
learned in servicing are not only quickly disseminated across its want to minimize the amount of downtime of the ontology of
staff of service technicians but also communicated to the de- the sort that would arise, for instance, if the GO failed to correct
signers of the next generation of GE engineers. errors or to provide terms relating to newly discovered biologi-
GE is thereby stimulated cal phenomena in a reliable manner and with a rapid turnaround
to develop sophisticated algorithms for predicting likely sources of time. The GO has helped its users also by developing software,
future engine failure and the optimal time to service the engine to such as the AmiGO browser (14), which enhances the value of
prevent such failures. The more data and experience GE accumu- the data entered by curators into the GOA database by making
lates, the better these algorithms become, and the more effective GE these data more easily accessible.
will become in delivering these services. (12)
The creation of the UniProt and other databases (15) which use
Manufacturing and servicing become tied together in ways that GO terms in annotations of their data provides the GO itself
allow advances on either side to be mutually reinforcing. This with an important informational advantage over potential com-
then gives rise to an additional dimension that will be of signi- petitors. This has served in the biological domain to slow the
ficance to us here, in that, for GE, as for other advanced manu-
growth of the sort of cottage industry of small (‘lite’), local on- ontology coverage (and thus division of labor) in overlap-
tologies that has, unfortunately, been a feature of ontology work ping areas
in many other domains, a phenomenon which has repeatedly - helping to align neighboring ontologies in logically fruit-
given rise to the sort of fragmentation of the data landscape ful ways
which ontology development was precisely designed to avoid.
updating ontology content wherever needed, while
The BFO Product-Service System - informing users in advance of proposed changes
- providing software support for making needed updates to
BFO was created to serve as the top-level ontology of the OBO
domain ontologies using BFO (25)
Foundry, and its three principal categories, of (i) independent
and (ii) dependent continuants and (iii) occurrents, correspond promoting sustainability in order to provide users with the
to the three Gene Ontology modules for Cellular Component, confidence that effort invested in using an ontology today
Molecular Function, and Biological Process, respectively (16). will not be wasted because the ontology ceases to be
The BFO developer community, too, offers not merely the BFO maintained at some time in the future.
product but also a range of services analogous to those provided There are a number of items which could be added to this list as
by the GO. Because BFO was established as a top-level ontol- desiderata, including targeted software tools for checking the
ogy (TLO) designed to support the coordinated development of BFO compliance of a domain ontology and also tools to assist
interoperable domain ontologies, it provides both (i) services to in the creation of BFO-conformant ontologies following the
those who are using BFO as a starting point for building domain proposals sketched in (18), (26), and (27). The OBO Foundry
ontologies, and (ii) services to the users of these domain ontol- provides useful first steps towards the provision of such ser-
ogies themselves. vices, but there is still no single submission point where some
The BFO developer community provides these services as a re- level of validation for at least some aspects of conformity could
flection of its conviction that an ontology benefits when it has be achieved for BFO-based ontologies.
a user community that is both large and diverse. Many of the BFO Version Tracking
services are provided through the OBO Foundry (of which BFO
forms a part). However, there are now significant numbers of One service that is indispensable to those who need to use an
BFO-compliant ontologies outside the domain of the life sci- ontology over long periods of time is a traceable revision his-
ences (17), and services must be provided to the developers and tory. This enables annotations of data to be kept up to date as
users of these ontologies also. the ontology used in these annotations changes. For this pur-
The BFO product consists of the ontology itself, in both formal- pose it must be possible to establish the present meanings of
ized and natural language versions, which is presented to its us- annotations created at an earlier date and using an earlier ver-
ers as a domain-neutral starting point for ontology creation in a sion of the ontology. Like the GO, BFO has a traceable history
way that brings the advantage of having been employed as TLO of this sort, and it has provided guidance to users of successive
in many peer ontology initiatives with a correspondingly broad versions both to support consistency of use from one version to
cohort of experts in BFO-based ontology development. the next and also to provide the rationale for specific changes
(28).
Services provided by the BFO developer community and its col-
laborators include: Given the large number of ontology development groups using
BFO as their common top level, it is important that updates in-
helping such users volve minimal disruption and that they are carefully managed
- to formulate definitions (18) in such a way that they do not disrupt existing workflows. Of-
- to re-engineer legacy domain ontology artifacts in such a ten, issues can be resolved without any necessary change in the
way as to achieve BFO conformance (19, 20, 21) ontology itself, for example through additional commentary on
the release document, or through associated changes in an ex-
providing
tension ontology such as the IAO (29).
- public dissemination and developer portals (for example
The consequences of proposed changes in BFO are thoroughly
ontobee (22))
evaluated before these changes are incorporated in the next pub-
- a tracker that enables users to post questions and report lic release. If such changes affect, for example, the ways Eng-
issues (23) lish-language definitions are formulated, the BFO developers
- manuals and ‘how to’ documents providing guidance on are careful to ensure that terms and relational expressions refer
developing and using BFO-conformant ontologies (24) persistently from one version to the next and that examples of
- review services for developers and users of BFO- usage provided in earlier versions continue to be applicable in
conformant ontologies later versions.
- training videos, including site visits, tutorials, workshops Ontologies Reusing BFO
and conferences
Some 300 domain ontologies have been built using BFO as top
serving as liaison between different ontology communities level (30). In some cases BFO has been used as a TLO for suites
working with BFO, for example: of mutually interoperable domain ontologies that have been
- helping to organize collaborative ontology building efforts developed with the goal of providing benefits analogous to
- helping to negotiate agreements concerning division of those brought in biological and biomedical domains by the
OBO Foundry. Examples include the Planteome Consortium
(31), the Network of Epidemiology Related Ontologies (32) and Reflecting the number of Department of Defense (DoD) and In-
the Penn TURBO (Transforming and Unifying Research with telligence Community (IC) ontology applications developed on
Biomedical Ontologies) suite (33). the basis of BFO as top level, the Joint Technical Committee on
Most recently, the Allotrope Foundation (34) has adopted BFO Information Technology (JTC 1) of ISO and the International
as its top-level ontology (35). The Foundation is funded by a Electrotechnical Commission (IEC) initiated in 2016 a process
consortium of the world’s major pharmaceutical companies to to consider BFO for adoption as an international standard. This
improve the way its members acquire, share and gain insights proposal led to the development of ISO/IEC 21838, Part 1 of
from scientific data through standardization and linked data. which sets forth the requirements for being a top-level ontology
(TLO). Part 2 then documents that BFO satisfies these require-
BFO’s ability to promote interoperability across ontology
ments, inter alia by providing formalizations of BFO in both
frameworks has made BFO attractive not only in the life
OWL and Common Logic. The requirements specify further (a)
sciences but also in other areas such as industrial engineering
that the CL formalization be proved consistent, and (b) that the
(36), manufacturing process modeling (37), and military
OWL formalization be proved to be logically derivable there-
intelligence (38). Not least importantly for our purposes here
from.
are those users of BFO who are developing ontologies of
product-service systems as set forth in (39) or (40). Examples Compliance to the ISO rules for the formulation of definitions
of institutions using BFO in the engineering domain include: required also a number of improvements in the treatment of nat-
ural language definitions of terms and relational expressions in
NSF Center for e-Design and the Realization of Engineered BFO 2 (50). Accordingly, a new version of BFO, to be called
Products and Systems (21, 41) BFO-ISO – roughly equivalent to BFO 2.1 in the conventional
Engineering Informatics Research Group (42) enumeration – will be released simultaneously with the publi-
Systems Engineering Research Center (SERC) (43) cation of Parts 1 and 2 of ISO/IEC 21838. (The backbone is-a
hierarchy of this new release is illustrated in Figure 1 below.)
In March 2019, BFO was selected, after an extensive review
process managed by the National Institute of Standards and The BFO-ISO framework offers a range of new opportunities
Technology (NIST), to serve as top-level ontology of the Indus- for ontology developers, since CL allows greater expressivity
trial Ontologies Foundry (IOF) suite (44). than OWL in the formulation of axioms and of definitions of
ontology terms. The FOL language from which CL is derived
Finally, in the military and intelligence domains, BFO is being
also brings the benefit that it is a more intuitive language when
used above all through the Common Core Ontology (CCO)
it comes to presenting formal content to human users.
suite (45), which forms a set of mid-level BFO-based reference
ontologies covering domains such as physical artifacts, geospa- What is a Top-Level Ontology?
tial entities, units of measure, time and events, together with a
ISO/IEC:21838-1 defines an ontology as:
large set of domain ontologies extending from the CCO cover-
ing, inter alia, land, sea, air, planning, operations, and sensor a collection of terms, relational expressions and associated natural-
data. language definitions together with one or more formal theories de-
signed to capture the intended interpretations of these definitions.
BFO and ISO/IEC 21838 Part of the goal of this definition is to take account of the fact
that an ontology – for example the Gene Ontology, or BFO –
A further type of service that can be of value in almost any area can exist in multiple successive versions and yet remain one and
of organized human activity is the establishment of standards. the same ontology. One solution to this problem views the on-
The International Standards Organization was founded on the tology as a document, analogous for example to a textbook, that
idea of answering the question: what’s the best way of doing exists in several successive editions (51). The ISO definition
this? (46). It began with units of measure and now embraces, above relies instead on a common reading of ‘collection’ (as in
for example, network security standards. ‘museum collection’) as representing something that – like or-
Standard terminologies can promote more effective communi- ganisms and organizations – can gain and lose included items
cation; electrotechnical standards can promote interoperability (parts) over time.
of hardware and software. They can promote improved under- Top-level ontologies deal with categories, which are general
standability of third‐party content formulated in accordance classes or types represented by domain-neutral terms such as
with the standard, allow new sorts of quality measures to be de- ‘object’ or ‘process’.
veloped and applied, and promote transportability of expertise.
On this basis ISO/IEC 21838-1 defines a top-level ontology as
Existing ISO standards relevant to ontology include: an ontology that deals with categories shared across a maximally
- Common Logic (CL, ISO/IEC 24707:2018) (47), a family of broad range of domains.
languages extending classical First-Order Logic (FOL) with The standard must then provide a procedure for determining
features designed to optimize computational use whether a candidate TLO satisfies this definition. To meet this
- Industrial Automation Systems and Integration (ISO 15926), need, Part 1 of the standard provides a list of types of entities
especially part 14: Data Model Adopted for OWL 2 Direct ranging from entities relating to time, space and spacetime, to
Semantics (draft dated 2019) change and process; qualities and quantities; material and infor-
- Process Specification Language (PSL, ISO 18629) (48) mational artefacts, and so forth. That an ontology O succeeds in
covering a ‘maximally broad range’ is shown by providing doc-
- Standard for the Exchange of Product Model Data (STEP,
umentation demonstrating that, given a type or class of entity of
ISO 10303) (49)
one or other of the listed sorts, either (i) there exists a corre- plagued ontology efforts in the past. In many cases, most nota-
sponding parent (or ancestor) term for this type of entity in O bly in the case of the Gene Ontology, we can use the results of
itself, or (ii) a definition of this class of entity can be created empirical science as a means of gaining access to those portions
through logical combination (for example disjunction) of terms of reality that form the needed benchmark. In other cases we
from O representing types or classes satisfying (i). may use for this purpose authoritative sources such as industrial
ISO/IEC 21838-1 therefore does not require that the TLO con- or military standards, or codes of law.
tains detailed treatments of entities under all the mentioned The reason for using such benchmarks as the basis for creating
headings. Thus, for example it is sufficient if it is possible to ontologies is (simplifying considerably) to arrive at a situation
point to some extension ontology which serves this purpose, in which there will be just one authoritative ontology for each
and which establishes the needed chain of subtype relations to domain of reality. This goal can be achieved only if we can per-
some term or terms in the TLO. Thus, for example there are no suade ontology developers to accept certain shared constraints
terms for information artifacts in BFO – thus no terms for data, on how they build ontologies and for this we need to employ a
signs, symbols, software, and so on. However, many terms in strategy that does not endanger, for example, the flexibility that
this family are provided in the Common Core Ontology suite is needed to keep pace with scientific advance. The OBO
(CCO) and in the Information Artifact Ontology (IAO), which Foundry has been at least partially successful in meeting these
provide definitions of the mentioned terms as representing sub- conditions, and in ways that essentially involve the use of BFO.
types or subclasses of BFO:generically dependent continuant.
BFO on the Nature of Ontology
BFO and DOLCE In 2008, I defended the proposition that ontologies like the GO,
which are created to support the retrieval, integration and anal-
As Guarino himself outlines in (52), BFO and DOLCE have
ysis of scientific data, are a part of science (58). They form what
much in common. They have a common origin: indeed a valu-
we can think of (again simplifying somewhat) as the termino-
able early presentation of BFO was published in the very same
logical scaffolding of a scientific theory. In the case of GO the
document (53) in which the formalization of DOLCE first ap-
relevant theory would be molecular biology. The ontologies in
pears. Both BFO and DOLCE rest on the same trinity of funda-
question are therefore subject to the same empirically-based
mental dichotomies: between (i) universals and instances, (ii)
methods of evaluation as are those theories themselves. An on-
dependent and independent entities, and (iii) continuants and
tology like the GO must, of course, be associated with imple-
occurrents (the latter referred to by DOLCE as, respectively,
mentations satisfying the requirements of software engineering.
endurants and perdurants). On the other hand, the two ontolo-
But the ontology is not the sort of thing that can exist only as
gies differ in a number of ways from the point of view of ontol-
embedded in some specific software framework.
ogy content – above all in the fact that DOLCE, but not BFO,
adopts a multiplicative view of continuants. This means that A counterpart view, as applied to a top-level ontology such as
DOLCE, but not BFO, distinguishes physical objects from the BFO, would see the latter as providing terminological resources
portions of matter which they contain at any given time. at a more general level, which is to say, above the level of the
specific sciences. This is to enable domain ontologies at lower
DOLCE was launched in 2002 in WonderWeb Deliverable D18
levels to be linked together. What subject-matter experts in, for
(53), and the latter contains what is still today the definitive for-
example, molecular biology or clinical medicine will see are
malization of DOLCE. This document forms one major mile-
terms in GO, or in some clinical ontology such as SNOMED
stone in a stream of important contributions from the DOLCE
CT. BFO will remain invisible.
development team, including the OntoClean methodology (54),
an ontological restructuring of WordNet (55), and a series of Certain requirements must be satisfied, however, If an ontology
contributions to domain ontology in areas such as law, engi- is to serve as “terminological scaffolding” of a scientific theory
neering, hydrology, and – interestingly for our purposes here – or of some counterpart thereof for example in the field of man-
of services science (56). It is worth remarking here also that the ufacturing industry (44) or military doctrine (59). For again, this
DOLCE community contributed to a remarkable degree in the will require services to be provided from the side of the devel-
provision of ontology-related services, above all in establishing opers of the ontology.
the FOIS ontology series, the Applied Ontology journal, and the Above all, it will require provision of services of the sort that
International Association for Ontology and Its Applications. will give users confidence that the ontology will be reliably dis-
While these are services to the broader ontology community, seminated and maintained, and will continue to offer needed
rather than services to the users of specific ontologies of the sort serviced, in the future. Only thus can we overcome a range of
that concern us here, this does not take anything away from their familiar obstacles standing in the way of adoption of an ontol-
intrinsic value and importance. ogy by new users, captured in complaints for example to the
effect that this or that ontology “was not being kept up to date”;
BFO and Ontological Realism
that it “was not clear to me how the ontology would fit my par-
BFO and DOLCE differ also with respect to the issue of onto- ticular data”; that it “would not be able to incorporate the terms
logical realism (57). For the BFO developer community, ontol- I needed in time for my funding deadline”; or that “I did not
ogy is an effort to foster consistency in the ways data are de- have the confidence that the ontology will still be supported
scribed by following a specific methodology which uses reality when I need it in the future”. Too often, potential new users of
as benchmark. The goal is to counteract the many tendencies an existing ontology are motivated to build ontologies of their
leading to ad hoc and non-interoperable coding of data, and own, resulting almost always in ad hoc contrivances with a very
thereby to the formation of data silos, of the sort that have short half-life.
DOLCE and Ontological Realism contrast, has been without update for some 17 years, which
means that it has been unaffected by the changes in the field of
The views of the developers of DOLCE on the topic of onto-
ontology around it. Thus while DOLCE provided the ontologi-
logical realism are formulated as follows:
cal foundations for important work in a series of multi-partner
the aim of DOLCE is to capture the intuitive and cognitive bias un- projects (64) involving the creation of DOLCE-based domain
derlying common-sense … DOLCE does not commit to a strong ref-
ontologies, none of these collaborations brought about changes
erentialist metaphysics (it does not make claims on the intrinsic na-
ture of the world) and does not take a scientific perspective (it is not
in DOLCE itself as a result of discoveries made in its actual use.
an ontology of, say, physics or of social sciences). Rather, it looks DOLCE (FOL) has thus not benefited from the sort of virtuous
at reality from the mesoscopic and conceptual level aiming at a for- cycle of continuous development through a mutually beneficial
mal description of a particular, yet fairly natural, conceptualization interaction with the users of domain ontologies constructed in
of the world. (60) its terms that has characterized the evolution of BFO.
DOLCE, that is to say, aims to capture the ontological catego- A possible exception in this regard is the e-Science Knowledge
ries lying behind natural language and human common sense, Infrastructure (65), which has taken important steps towards
so that its categories are to be regarded “conceptual containers” creation of a suite of ontologies in the domain of hydrology.
and thus as “cognitive artifacts ultimately depending on human (66, 67). Otherwise, while DOLCE has certainly inspired the
perception, cultural imprints and social conventions.” (53, 55) creation of appreciable bodies of domain-ontology content, it
One problem with views of this sort, however, is that they can has not been able to serve in any of the domains where it has
be detrimental to the goal of using a top-level ontology as a been applied as an easily findable, easily learnable, easily teach-
means of promoting interoperability of domain ontologies de- able hub for the development of mutually consistent extension
fined in its terms. Indeed, since ‘cultural imprints and social ontologies in sustainable unitary suites analogous to the OBO
conventions’ vary so widely, allowing these to play a role in Foundry or the Common Core (45).
determining top-level ontology content raises the problem of The domain ontology contributions created on the basis of
non-interoperability at this very top level itself. DOLCE thus survive largely as fragments, documented in sci-
entific papers, rather than as ontological going concerns.
DOLCE on the Nature of Ontology
In sum, the various domain ontologies built around the official
A more charitable, and I believe more adequate, view of DOLCE have not been provided with the services needed to
DOLCE, however, sees it as a scientific ontology of human make the domain ontologies defined in its terms work well to-
common sense, with needed benchmarks provided, for exam- gether in a sustainable, publicly accessible way. Matters are
ple, by linguistics, perceptual psychology and action theory. somewhat different in the case of DOLCE in its OWL formali-
DOLCE can in this way be viewed as a contribution to science, zations, which have a richer history of usage than the FOL ver-
that has served as inspiration for the development of many new sion of DOLCE. This is in part because most users of ontologies
ontologies, both domain ontologies based on DOLCE in its work with OWL rather than FOL. But it is in part also because
FOL version, as well as spin-offs from DOLCE, such as the of differences in management policy, reflected in the multiple
Unified Foundational Ontology (61). sorts of user assistance documented for example at (63).
In support of a view of this sort is the remarkable degree to The FOL version of DOLCE was, clearly an unusually impres-
which DOLCE (FOL) has remained stable across its entire his- sive piece of ontology design from the very start, and thus there
tory, reflecting the degree to which human common sense, too, are good reasons why it has survived so long without updates.
has also – for reasons relating to the evolutionary survival of Its record in this respect is not perfect, however, given that one
the species – manifested little change over long periods. (62) of the most important spin-offs from DOLCE was the DOLCE-
The remarkable stability of DOLCE in its FOL version has CORE proposal presented in 2009 (68). For the latter describes
however been to some degree overshadowed by the many – in a number of improvements over the official DOLCE, and rep-
some ontology user circles more conspicuous – artifacts created resents what it calls a “first step, after the release of the DOLCE
using DOLCE (FOL) as inspiration, but formalized using ontology in 2002, toward a new version of this ontological sys-
OWL, for instance as listed in (63). The proliferation of ver- tem.” DOLCE-Core has however not resulted in a new version
sions of an ontology is clearly not an unalloyed good in the on- of DOLCE even though members of the Guarino lab have been
tology context, since it will tend to diminish the degree to which using it to build domain ontologies since 2009, as for example
the ontology will be trusted by potential users as a resource that described in (69) and (70).
can be relied upon to promote interoperability in a sustainable
fashion. Guarino has accordingly (in personal communication) Conclusion
referred to the DOLCE OWL artifacts as mere “variants” over
and against the one “version” of DOLCE described in Sections 1. Ontology as Interdiscipline There are scientific disciplines;
and, it is sometimes said, there are scientific interdisciplines
3 and 4 of the WonderWeb deliverable D18 (2003).
(71). Interdisciplines work like disciplines. They involve peo-
Lessons Learned from BFO and DOLCE ple making scientific contributions – theories, experiments,
data, perhaps also ontologies – but in such a way that these con-
In the matter of update history, now, BFO lies somewhere in-
tributions cross established disciplinary boundaries.
termediate between DOLCE (FOL) on the one hand and
DOLCE (OWL) on the other. For BFO has been maintained as There is no question that the discipline of ontology as a whole
one thing through a series of updates over time motivated by is properly conceived as an interdiscipline, spanning (at least)
the experiences of its users. The one “version” of DOLCE, in philosophy, linguistics, engineering, and various branches of
computer and information science. We believe, however, that are of two sorts: material products such as laptops and servers;
what has been said above suggests a new approach to the ques- and digital products such as Adobe Acrobat. Some enterprises
tion of the interdisciplinary nature not only of ontology, but also – such as business consultants – produce no products at all but
of a range of related activities such as scientific theorizing and only services. Some enterprises product products – such as
standards development. IBM’s Watson – but they give them away free, and provide ser-
Let us therefore assume that ontology itself is an interdiscipline. vices to support their use.
What is to be said now of each single ontology? Is there some I conclude merely by noting that this perspective can be applied
single type or class whose instances are the GO, and BFO, and not only to teams of ontology developers. Standards organiza-
IAO, and UFO, and DOLCE, and DOLCE+DnS Ultralite v. tions, too, can be viewed under this heading, given that NIST,
3.31? Or do we rather need something like an ontology of on- CEN, ASME, ISO, IEC, W3C and even HL7 are both produc-
tologies that would have no single root? tion systems, producing standards, and service systems, provid-
2. Ontology as business enterprise: Much of the foregoing has ing support for the users of these standards.
rested on what I think is a novel view of ontology developer And communities of scientists, too, can be viewed as providing
teams as analogous to business enterprises. product-service systems, containing both production elements
The latter, as we have seen, have as their outputs both products – producing scientific results, publications – and service ele-
and services, more or less closely bundled together. Products ments, for example training each new cohort of scientists.
Figure 1: Is_a Hierarchy of BFO-ISO
References
1. Ashburner M, Ball CA, Blake JA, et al. Gene Ontology: tool 7. http://www.obofoundry.org/docs/Policy_for_OBO _name
for the unification of biology. Nature Genetics. 25(1), 2000. p. space_and_associated_PURL_requests
25-29. 8. http://www.obofoundry.org/id-policy.html
2. https://buffalo.app.box.com/v/Building-Ontologies. All URIs 9. Hill DP, Blake JA, Richardson JE, Ringwald M. Extension
last accessed April 8, 2019. and integration of the Gene Ontology (GO): Combining GO
3. Day-Richter J, Harris MA, Haendel M. et al. OBO-Edit – Vocabularies with external vocabularies. Genome Res. 2002;
An ontology editor for biologists. Bioinformatics, 23:16; 2198– 12(12), p. 1982-1991.
2200, 2007. 10. http://robot.obolibrary.org
4. Huntley RP, Sawford T, Mutowo-Meullenet P, Shypitsyna 11. http://www.hegroup.org/ontozoo
A, Bonilla C, Martin MJ, O’Donovan C. The GOA database: 12. Chesbrough H. Open services innovation, https: //www.in-
Gene Ontology annotation updates for 2015. Nucleic Acids Re- novationexcellence.com/blog/2012/11/04/open-services-inno-
search, 43, D1, 2015, p. D1057–D1063. vation/.
5. http://wiki.geneontology.org/index.php/Guide_to_ GO_ Ev- 13. Mont O. Clarifying the concept of product-service system.
idence_Codes Journal of Cleaner Production, 10 (2002), p. 237-245.
6. “Services to support the OBO Foundry standards”, https:// 14. http://amigo.geneontology.org/amigo
app.dimensions.ai/details/grant/grant.7071924
15. The UniProt Consortium. The Universal Protein Resource Basic Formal Ontology, 2017 International Conference on En-
(UniProt). Nucleic Acids Res. 2008 Jan; 36(Database issue). gineering, Technology and Innovation (ICE/ITMC).
16. Munn, K. & Smith, B. (eds.) Applied Ontology: An Intro- 40. DIVERSITY/FALCON White Paper: Product Service Sys-
duction, Frankfurt/Lancaster: ontos/Walter de Gruyter, 2008. tem (PSS) Ontology for Discrete Manufacturing. Draft 2019.
17. http://basic-formal-ontology.org/users.html Available from ICT for Sustainable Manufacturing, EPFL CH-
18. Seppälä S, Ruttenberg A, Smith B. Guidelines for writing 1015 Lausanne.
definitions in ontologies. Ciência da informação 46 (1), 73-88; 41. http://www.centerforedesign.org/
2017. 42. http://infoneer.wp.txstate.edu/
19. Hagedorn TJ, Krishnamurty S, Grosse IR, A knowledge- 43. Transforming Systems Engineering through Model-Cen-
based method for innovative design for additive manufacturing tric Engineering A013 Interim Technical Report SERC-2017-
supported by modular ontologies, Journal of Computing and In- TR-111. https://apps.dtic.mil/dtic/tr/fulltext/u2/1058354.pdf
formation Science in Engineering, 18(2), 021009, 2018. 44. Wallace E, Kiritsis D, Smith B, Will C. The Industrial On-
20. Arena D, Ameri F, Kiritsis D, Skill modeling for digital fac- tologies Foundry proof-of-concept project, APMS 2018: Ad-
tories, IFIP International Conference on Advances in Produc- vances in production management systems. Smart manufactur-
tion Management Systems. APMS 2018: Advances in Produc- ing for industry 4.0 (IFIP International conference on advances
tion Management Systems. Smart Manufacturing for Industry in production management systems, 536). Springer, 2018; p.
4.0, 2018, 318-326. 402-409.
21. Hagedorn TJ, Smith B, Krishnamurty S, Grosse I. Interop- 45. https://www.cubrc.org/index.php/data-science-and-infor-
erability of disparate engineering domain ontologies using mation-fusion/ontology
Basic Formal Ontology, Journal of Engineering Design, in pro- 46. https://www.iso.org/benefits-of-standards.html
duction, 2019. 47. Gruninger M, Katsumi M, Mossakowski T. Revision of
22. http://www.ontobee.org/ ISO 24707 (Common Logic), 2013. http://metadata-stand-
23. https://groups.google.com/d/forum/bfo-discuss ards.org.
24. Arp R, Smith B, Spear A. Building ontologies with Basic 48 Grüninger, M. Ontology of the Process Specifica-tion Lan-
Formal Ontology, MIT Press; 2015. guage. http://bbs.w3china.org/dragonstar/-
25. http://ontobull.hegroup.org/ papers/psl-chapter.pdf
26. Seyed P. A method for evaluating ontologies – Introducing 49. Barbau R, Krima S, Rachuri S, et al. OntoSTEP: Enriching
the BFO-Rigidity Decision Tree Wizard, Formal Ontology in product model data using ontologies. Computer-Aided Design.
Information Systems (FOIS 2012), p. 191-204. 44(6), 2012, 575-590.
27. Seppälä S. An ontological framework for modeling the con- 50. https://github.com/BFO-ontology/BFO/wiki
tents of definitions. Terminology: International Journal of The- 51. Neuhaus F. What is an Ontology? 2018. arxiv 1810.09171.
oretical and Applied Issues in Specialized Communication, 52. Guarino N. BFO and DOLCE: So far, so close …, Cosmos
2015, 21(1), + Taxis 4 (4):10-18. 2017.
23-50 53. Masolo C, Borgo S, Gangemi A, et al. Deliverable D18:
28. Seppälä S, Smith B, Ceusters W. Applying the realism- Ontology Library, IST Project 2001-33052 WonderWeb: On-
based ontology-versioning method for tracking changes in the tology infrastructure for the Semantic Web. 2003.
Basic Formal Ontology, Formal Ontology in Information Sys- 54. Guarino N, Welty C. An Overview of OntoClean. S Staab
tems (FOIS 2014), 2014, p. 227-240. and R Studer (eds.) Handbook on Ontologies, 2009, 201-220,
29. http://www.ontobee.org/ontology/IAO 55. Oltramari A, Gangemi A, Guarino N, Masolo C. Restruc-
30. http://basic-formal-ontology.org/users.html turing WordNet’s Top-Level: The OntoClean approach, Pro-
31. http://planteome.org/ ceedings of Workshop on Ontologies and Lexical Knowledge
32. Pesquita C, Ferreira JD, Couto FM, Silva MJ. The epide- Bases (OntoLex’02), 2002, p. 17-26.
miology ontology: an ontology for the semantic annotation of 56. Ferrario R, Guarino N, Janiesch C, Kiemes T, Oberle D,
epidemiological resources. J Biomed Semantics. 2014; 5: 4. Probst F. Towards an ontological foundation of services sci-
33. http://upibi.org/turbo/ ence: The General Service Model. Internationale Tagung
34. https://www.allotrope.org/ Wirtschaftsinformatik, Zürich, 2011.
35. http://www.prweb.com/releases/2018/03/prweb15337219. 57. Smith B, Ceusters W. Ontological realism: A methodology
htm for coordinated evolution of scientific ontologies. Appl Ontol.
36. http://ncorwiki.buffalo.edu/index.php/BFO-Engineering 2010, 5(3-4), p. 139-188.
37. Arena DN. Towards semantics-driven modelling and simu- 58. Smith B. Ontology (science). Formal Ontology in Infor-
lation of context-aware manufacturing systems, Dissertation, mation Systems (FOIS 2008). Amsterdam: IOS Press. p. 21-35.
EPFL – École polytechnique fédérale de Lausanne, School of 59. Morosoff P, Rudnicki R, Bryant J, Farrell R, Smith B. Joint
Engineering, June 2019. Doctrine Ontology: A benchmark for military information sys-
38. Smith B, Malyuta T, Rudnicki R, et al. IAO-Intel: An on- tems interoperability. Semantic Technology for Intelligence,
tology of information artifacts in the intelligence domain. Se- Defense and Security (STIDS 2015). (CEUR 1325), p. 2-9.
mantic Technologies for Intelligence, Defense, and Security, 60. Borgo S, Masolo C. Foundational choices in DOLCE, in R
(STIDS 2013), CEUR, vol. 1097, p. 33-40. Poli et al. (eds.), Theory and applications of ontology: Com-
39. Correia A, Stokic D, Siafaka R, Scholze S, Ontology for puter applications, Springer, 2010, p. 279-295.
collaborative development of product service systems based on 61. https://ontology.com.br/
62. Smith B. Formal ontology, common sense, and cognitive 68. Borgo S, Masolo C. Foundational choices in DOLCE, in:
science. International Journal of Human-Computer Studies, 43, Handbook on ontologies, Springer; 2009. p. 361-381.
1995. p. 641-667. 69. Masolo C. Founding properties on measurement. Formal
63. http://www.ontologydesignpatterns.org/ont/dul/ ontology in information systems (FOIS 2010); p. 89-102.
64. http://www.loa.istc.cnr.it/old/Projects.html 70. Ferrario R, Porello D. Towards a conceptualization of so-
65. Brodaric B, Reitsma F, Qiang Y. SKIing with DOLCE: to- ciomaterial entanglement, CONTEXT: Modeling and using
ward an e-Science Knowledge Infrastructure. Formal Ontology context, 2015. p. 32-46.
in Information Systems, Amsterdam: IOS Press, 2008. p. 208- 71. Newell A. Reflections on the structure of an interdiscipline,
219. in F Machlup and U Mansfield (Eds.), The study of infor-
66. Brodaric B, Hahmann T, Gruninger M. Water features and mation: Interdisciplinary messages. New York, 1983, p. 99-
their parts. Applied Ontology, 2018, 14(1) 1-42. 109.
67. https://www.researchgate.net/project/A-hydro-foundation-
al-ontology-HyFO