=Paper=
{{Paper
|id=Vol-2029/ks5
|storemode=property
|title=Challenges for Ontology Repositories and Applications to Biomedicine & Agronomy
|pdfUrl=https://ceur-ws.org/Vol-2029/ks5.pdf
|volume=Vol-2029
|authors=Clement Jonquet
|dblpUrl=https://dblp.org/rec/conf/simbig/Jonquet17
}}
==Challenges for Ontology Repositories and Applications to Biomedicine & Agronomy==
https://ceur-ws.org/Vol-2029/ks5.pdf
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
Challenges for ontology repositories and
applications to biomedicine & agronomy
Clement Jonquet
Laboratory of Informatics, Robotics, and Microelectronics of Montpellier (LIRMM),
University of Montpellier & CNRS, France &
Center for BioMedical Informatics Research (BMIR), Stanford University, USA
jonquet@lirmm.fr
(ORCID: 0000-0002-2404-1582)
Abstract 1 Introduction
The explosion of the number of ontologies The Semantic Web produces many vocabularies
and vocabularies available in the Semantic and ontologies to represent and annotate any kind
Web makes ontology libraries and reposi- of data. However, those ontologies are spread out,
tories mandatory to find and use them. in different formats, of different size, with differ-
Their functionalities span from simple on-
ent structures and from overlapping domains. The
tology listing with more or less of metada-
scientific community has always been interested
ta description to portals with advanced on-
tology-based services: browse, search, vis- in designing common platforms to list and some-
ualization, metrics, annotation, etc. Ontol- time host and serve ontologies, align them, and
ogy libraries and repositories are usually enable their (re)use (Ding and Fensel, 2001;
developed to address certain needs and Hartmann et al., 2009; D’Aquin and Noy, 2012; ,
communities. BioPortal, the ontology re- 1995). These platforms range from simple ontol-
pository built by the US National Center ogy listings or libraries with structured metadata,
for Biomedical Ontologies BioPortal relies to advanced repositories (or portals) which fea-
on a domain independent technology al- ture a variety of services for multiple types of
ready reused in several projects from bio-
semantic resources (ontologies, vocabularies,
medicine to agronomy and earth sciences.
terminologies, taxonomies, thesaurus) such as
In this position paper, we describe six high
level challenges for ontology repositories: browse/search, visualization, metrics, recommen-
metadata & selection, multilingualism, dation, or annotation. In this paper, we will focus
alignment, new generic ontology-based on ontology repositories, they allow to address
services, annotations & linked data, and important questions:
interoperability & scalability. Then, we x If you have built an ontology, how do you let
present some propositions to address these the world know and share it?
challenges and point to our previously x How do you connect your ontology to the rest
published work and results obtained within of the semantic world?
applications –reusing NCBO technology– x If you need an ontology, where do you go to
to biomedicine and agronomy in the con-
get it?
text of the NCBO, SIFR and AgroPortal
projects.
x How do you know whether an ontology is any
good?
Keywords x If you have data to index, how do you find the
most appropriate ontology for your data?
Ontologies, ontology libraries & reposito- x If you look for data, how may the semantics
ries, ontology metadata, ontology-based of ontologies help you locate them?
services, ontology selection, semantic an- More generally, ontology repositories help “ontol-
notation, BioPortal. ogy users” to deal with ontologies without manag-
ing them or engaging in the complex and long
process of developing them.
25
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
However, with big number of ontologies new 6. Scalability & interoperability. The com-
problems have raised such as describing, select- munity of ontology developers and users is
ing, evaluating, trusting, and interconnecting growing both horizontally (i.e., new do-
them. From our experience working first on the mains) and vertically (i.e., new adopters in-
US National Center for Biomedical Ontologies side a domain). Ontology repositories shall
(NBCO) BioPortal, the most widely adopted bio- therefore scale to high number of ontologies,
medical ontology repository and later on the SIFR while facilitating their alignments, and when
BioPortal, a specific sub-portal to address the multiple repositories are created, they must
French biomedical community and AgroPortal, an be interoperable.
In the following, we will detail these challenges
ontology repository for agronomy, we review and
and briefly describe/point to results obtained in
discuss six challenges in designing such plat-
the context of our multiple ontology repository
forms:
projects. In some sense, this article is an index of
1. Metadata & selection. Ultimately, ontology
repositories are made to share and reuse on- 10-years of published research in the domain of
tologies. But which ontology should I reuse? ontology repositories. We do not report hereafter
With too many different and overlapping on- all related work for each challenge neither we
tologies, properly describing them with claim to have addressed them all. However, we
metadata and facilitate their identification believe our results illustrate potential solutions to
and selection becomes and important issue. move forward in that domain of research.
We believe, as any other data, ontologies
must be FAIR. 2 Background
2. Multilingualism. We live in a multilingual
world, so are the concepts and entities from 2.1 Ontology libraries & repositories
this world. The Semantic Web offers now With the growing number of ontologies devel-
tools and standards to develop multilingual oped, ontology libraries and repositories have al-
and lexically rich ontologies. Repositories ways been of interest in the Semantic Web com-
must be able to deal with multiple languages munity. Ding and Fensel (2001) introduced the
also. notion of ontology library and presented a review
3. Ontology alignment. No conceptualization of libraries at that time:
is an island. It is now commonly agreed data
“A library system that offers various func-
interoperability cannot be achieved by means
tions for managing, adapting and standardiz-
of a single common ontology for a domain,
and interlinking ontologies is the way for- ing groups of ontologies. It should fulfill the
ward. But the more ontologies are being pro- needs for re-use of ontologies. In this sense,
duced, the more the need for ontology an ontology library system should be easily
alignment becomes important. accessible and offer efficient support for re-
4. Ontology-based services. On reason to using existing relevant ontologies and stand-
adopt Semantic Web standards and use on- ardizing them based on upper-level ontolo-
tology repositories is to benefit from multiple gies and ontology representation languages.”
services for –and based on– ontologies. No The terms “collection”, “listing” or “registries”
one likes to reimplement something already are also used to describe ontology libraries. All
existing and that can be generalized to an- correspond to systems that help reuse or find on-
other ontology just by dropping it in a reposi- tologies by simply listing them (e.g., DAML or
tory. The portfolio of services for ontologies DERI listings) or by offering structured metadata
available in repositories should then grows. to describe them (e.g., FAIRSharing, BARTOC).
5. Annotations and linked data. Ontologies But those systems do not support any services be-
and vocabularies are the backbone of seman- yond description, especially based on the content
tically rich data (Linked Open Data, of the ontologies.
knowledge bases, etc.) as they are used to Hartmann et al., (2009) introduced the concept
semantically annotate and interlink datasets. of ontology repository, with advanced features
It is also important to facilitate semantic in-
such as search, browsing, metadata management,
dexing, search and data access directly from
visualization, personalization, and mappings and
the repositories.
an application programming interface to query
their content/services:
26
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
“A structured collection of ontologies (…) applications that pull their data from the foundry,
by using an Ontology Metadata Vocabulary. such as the NCBO BioPortal (Noy et al., 2009),
References and relations between ontologies OntoBee (Ong et al., 2016), the EBI Ontology
and their modules build the semantic model of Lookup Service (Côté et al., 2006) and more re-
an ontology repository. Access to resources is cently AberOWL (Hoehndorf et al., 2015). In ad-
realized through semantically-enabled interfac- dition, there exist other ontology libraries and re-
es applicable for humans and machines. There- pository efforts unrelated to biomedicine, such as
fore, a repository provides a formal query lan- the Linked Open Vocabularies (Vandenbussche et
guage.” al., 2014), OntoHub (Till et al., 2014), and the
By the end of the 2000’s, the topic was of high in- Marine Metadata Initiative’s Ontology Registry
terest as illustrated by the 2010 ORES and Repository (Rueda et al., 2009). More recent-
workshop (d’Aquin et al., 2010) or the 2008 On- ly, the SIFR BioPortal (Jonquet et al., 2016a) pro-
tologySummit.1 The Open Ontology Repository totype was created at University of Montpellier to
Initiative (Baclawski and Schneider, 2009) was a build a French Annotator and experiment multi-
collaborative effort to develop a federated infra- lingual issues in BioPortal (Jonquet et al., 2015).
structure of ontology repositories. At that time, the The same university is also developing AgroPor-
effort already reused the NCBO tal, an ontology repository for agronomy and
technology (Whetzel and Team, 2013) that was neighboring domains such as food, plant sciences
the most advanced open source technology for and biodiversity (Jonquet et al., 2017a).
managing ontologies but not yet packaged in an D’Aquin and Noy, (2012) and Naskar and
“virtual appliance” as it is today. More recently Dutta, (2016) provided the latest reviews of ontol-
the effort also studied OntoHub (Till et al., 2014) ogy repositories. In Table 1, we provide a non-
technology for generalization but the OOR initia- exhaustive –but quite rich– list of ontology librar-
tive is now discontinued. ies, repositories and Web indexes available today.
In parallel, there have been effort do index any Ontology libraries
Semantic Web data online (including ontologies) OBO Foundry
and offer search engines such as Swoogle and WebProtégé
Watson (Ding et al., 2004; D’Aquin et al., 2007). Romulus
We cannot talk about ontology library or reposito- DAML ontology library
ries for those “Semantic Web indexes”, even if
Colore
they support some features of ontology libraries or
VEST/AgroPortal Map of standards
repositories (e.g., search).
FAIRsharing
In the biomedical or agronomic domains there
DERI Vocabularies
are several standards and/or ontology libraries
such as FAIRSharing (fairsharing.org) (McQuilton OntologyDesignPatterns
et al., 2016), the FAO’s VEST Registry SemanticWeb.org
(aims.fao.org/vest-registry), and the agINFRA W3C Good ontologies
linked data vocabularies (vocabularies.aginfra.eu). TaxoBank
They usually register ontologies and provide a few BARTOC
metadata attributes about them. However, because GFBio Terminology Service
they are registries not especially focused on vo- agINFRA Linked Data Vocabularies
cabularies and ontologies, they do not support the oeGOV
level of features that an ontology repository offers. Ontology repositories
In the biomedical domain, the OBO NCBO BioPortal*
Foundry (Smith et al., 2007) is a reference com- Ontobee
munity effort to help the biomedical and biologi- EBI Ontology Lookup Service
cal communities build their ontologies with an en- AberOWL
forcement of design and reuse principles that have CISMEF HeTOP
made the effort very successful. The OBO Found- SIFR BioPortal*
ry Web application (http://obofoundry.org) is not OKFN Linked Open Vocabularies
an ontology repository per se, but relies on other ONKI Ontology Library Service
1 http://ontolog.cim3.net/wiki/OntologySummit2008.html MMI Ontology Registry and Repository*
27
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
ESIPportal* ta, notes, and projects are stored in an RDF2 triple
AgroPortal* store) (Salvadores et al., 2013).
OntoHub An important aspect is that NCBO
Finto technology (Whetzel and Team, 2013) is domain-
EcoPortal (proposition end 2017)* independent and open source. A BioPortal virtual
Semantic Web indexes appliance3 is available as a server machine em-
Swoogle bedding the complete code and deployment envi-
Watson ronment, allowing anyone to set up a local ontolo-
Sindice gy repository and customize it. The NCBO virtual
Falcons appliance is quite regularly reused by organiza-
tions which need to use services like the NCBO
Technology
Annotator but have to process sensitive data in
NCBO Virtual Appliance (Stanford)
house e.g., hospitals. Via the virtual appliance,
OLS technology (EBI)
NCBO technology has already been adopted for
LexEVS (Mayo Clinic)
different ontology repositories in related domains
Intelligent Topic Manager (Mondeca)
and was also originally chosen as foundational
SKOSMOS (Nat. Library of Finland) software of the OOR Initiative (Baclawski and
Abandoned projects include: Cupboard, Knoodl, Schneider, 2009). The MMI Ontology Registry
Schemapedia, SchemaWeb, OntoSelect, On- and Repository (Rueda et al., 2009) used it as its
toSearch, OntoSearch2, TONES, SchemaCache, backend storage system for over 10 years, and the
Soboleo Earth Sciences Information Partnership earth and
Table 1. Non-exhaustive list of ontology librar- environmental semantic portal (Pouchard L.
ies, repositories and Web indexes available to- Huhns M., 2012) was deployed several years ago.
day. We also included some known “technology” We are also currently working on the SIFR Bi-
that can be reused to setup an ontology library. oPortal (Jonquet et al., 2016a) and
Blue cells are projects in biomedicine and health AgroPortal (Jonquet et al., 2017a) projects de-
sciences. A * identifies ontology repositories scribed hereafter.
which reuse(d) NCBO technology.
2.3 Two collaborative ontology repository
2.2 Focus on the NCBO BioPortal: a “one projects
stop shop” for biomedical ontologies
In the context of our projects, to avoid building
In the biomedical domain, BioPortal new ontology repositories from scratch, we have
(http://bioportal.bioontology.org) (Noy et al., considered which of the previous technologies are
2009), developed by the National Center for Bio- reusable. While most of them are “open source,”
medical Ontologies (NBCO) at Stanford is a well- only the NCBO BioPortal4 and OLS5 are really
known open repository for biomedical ontologies meant for reuse, both in their construction, and
originally spread out over the Web and in different with their documentation provided. Although we
formats. There are +650 public ontologies in this 2 The Resource Description Framework (RDF) is the W3C
collection as of end 2017. By using the portal’s language to described data. It is the backbone of the seman-
features, users can browse, search, visualize and tic web. SPARQL is the corresponding query language. By
comment on ontologies both interactively through adopting RDF as the underlying format, an ontology reposi-
tory based on NCBO technology can easily make its data
a Web interface, and programmatically via Web available as linked open data and queryable through a public
services. Within BioPortal, ontologies are used to SPARQL endpoint. To illustrate this, the reader may consult
develop an annotation workflow (Jonquet et al., the Link Open Data cloud diagram (http://lod-cloud.net) that
since 2017 includes ontologies imported from the NCBO
2009) that indexes several biomedical text and da-
BioPortal (most of the Life Sciences section).
ta resources using the knowledge formalized in 3
www.bioontology.org/wiki/index.php/Category:NCBO_Virtual_Appliance
ontologies to provide semantic search features that 4
The technology has always been open source, and the ap-
enhance information retrieval experience (Jonquet pliance has been made available since 2011. However, the
product became concretely and easily reusable after BioPor-
et al., 2011). The NCBO BioPortal functionalities tal v4.0 end of 2013.
have been progressively extended in the last 12 5 The technology has always been open source but some
years, and the platform has adopted Semantic Web significant changes (e.g., the parsing of OWL) facilitating
the reuse of the technology for other portals were done with
technologies (e.g., ontologies, mappings, metada-
OLS 3.0 released in December 2015.
28
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
cannot know all the applications of other technol- tory for agronomy, food, plant sciences, and bio-
ogies, the visibly frequent reuse of the NCBO diversity (http://agroportal.lirmm.fr) (Jonquet et
technology definitively confirmed it is a good al., 2016c; Jonquet et al., 2017a). AgroPortal, is an
candidate for reuse when building a new ontology advanced prototype featuring all BioPortal ser-
repository. Also, of the two candidate technolo- vices and new ones implemented to address the
gies, we believe NCBO technology implements requirements of the agronomy community. The
the highest number of required features in our pro- platform currently hosts 77 ontologies among
jects (Jonquet et al., 2017a). which 50 are not present in any comparable repos-
itory. We have identified 93 other candidate ontol-
SIFR BioPortal ogies that will be loaded in the future to comple-
In the context of the Semantic Indexing of French ment this valuable resource.
Biomedical Data Resources (SIFR) project, we
have developed the SIFR BioPortal 3 Challenges, propositions and results
(http://bioportal.lirmm.fr) (Jonquet et al., 2016a),
In the following sections, we describe some chal-
an open platform to host French biomedical on-
lenges we identified by working on ontology re-
tologies and terminologies based on the technolo-
pository and exchanging with our user communi-
gy developed by the NCBO. The portal facilitates
ties. In each case, we describe a few results ob-
use and fostering of terminologies and ontologies
tained on the relevant topic.
which were only developed in French or translated
from English resources and are not well served in 3.1 Metadata & selection
the English-focused NCBO BioPortal. As of to-
day, the portal contains 25 public ontologies and The first questions we ask ourselves when enter-
terminologies (+ 6 private ones) that cover multi- ing a bookstore are often: “Where is the book I am
ple areas of biomedicine, such as the French ver- looking for?” or “Which book will I discover and
sions of standards terminologies (e.g., MeSH, pick up today?” The same questions are true for
MedDRA, ATC, ICD-10) but also multilingual ontology libraries. To address them, we need bet-
ontologies. In this later cases, we use the NCBO ter description of the ontologies, with precise
BioPortal as a source repository –so users do not and harmonized metadata and we need also
have to upload their multilingual ontologies means to facilitate the identification and selec-
twice– and only parse and index the French con- tion of the ontologies of interest. Ontologies
tent on the SIFR BioPortal. serve to make data FAIR (Wilkinson et al., 2016),
The original motivation in building the SIFR ontology repositories shall serve to make ontolo-
BioPortal was to develop the SIFR Annotator gies FAIR.
(http://bioportal.lirmm.fr/annotator) to address the As any resources, ontologies, vocabularies and
lack of out-of-the-shelve openly and easily acces- terminologies need to be described with relevant
sible semantic annotation system for metadata to facilitate their identification and selec-
French (Jonquet et al., 2016a; Tchechmedjiev et tion. However, none of the existing metadata vo-
al., 2017a). The service is originally based on the cabularies can completely meet this need if taken
NCBO Annotator [8], a Web service allowing sci- independently. Indeed, some metadata properties
entists to utilize available biomedical ontologies are intrinsic to the ontology (name, license, de-
for annotating their datasets automatically, but scription); others, such as community feedbacks,
was significantly enhanced and customized for or relations to other ontologies are typically in-
French. The annotator service processes raw tex- formation that an ontology library shall capture,
tual descriptions, tags them with relevant biomed- populate and consolidate to facilitate the ontology
ical ontology concepts and returns the annotations landscape comprehension (e.g., selection of an on-
to the users in several formats such as JSON-LD, tology).
RDF or BRAT. In Jonquet et al., (2017b), we have reviewed the
most standard and relevant vocabularies (23 to-
AgroPortal: a vocabulary and ontology repos- tals) currently available to describe metadata for
itory for agronomy ontologies (such as Dublin Core, Ontology
We have been reusing the NCBO BioPortal tech- Metadata Vocabulary, VoID, etc.) as well as the
different metadata implementation in multiple on-
nology to design AgroPortal, an ontology reposi-
tology libraries or repositories. We have then built
29
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
a new metadata model for AgroPortal. The reposi- ommendation approach evaluates the relevance of
tory now parses 346 standard properties that could an ontology to biomedical text data according to
be used to describe different aspects of ontologies: four different criteria: (1) the extent to which the
intrinsic descriptions, people, date, relations, con- ontology covers the input data; (2) the acceptance
tent, metrics, community, administration, and ac- of the ontology in the community; (3) the level of
cess. We use them to populate a model of 127 detail of the ontology classes that cover the input
properties implemented in the portal and harmo- data; and (4) the specialization of the ontology to
nized for all the ontologies. We have spent a sig- the domain of the input data. This new version of
nificant amount of time to edit the metadata of the a service originally released in 2010 (Jonquet et
ontologies with the goal to facilitate the compre- al., 2010) combines the strengths of its predeces-
hension of the agronomical ontology landscape by sor with a range of adjustments and new features
displaying diagrams and charts about all the on- that improve its reliability and usefulness. Be-
tologies on the portal. We have now a specific cause it is integrated in the NCBO technology, the
page (http://agroportal.lirmm.fr/landscape) dedi- Recommender is already available within the
cated to visualizing the ontology landscape in Ag- SIFR BioPortal and AgroPortal. We shall note that
roPortal that facilitates analysis of the repository these services do not yet rely on the new metadata
content. The landscape page helps to figure out model previously cited.
what are some of the main domain of interests as
well as common development practices when cre- 3.2 Multilingualism
ating an ontology in agronomy. Scientific discoveries that could be made with
In Dutta et al., (2017), we have generalized our help of ontologies to annotate, integrate, mine and
work done within AgroPortal to propose a new search data, are often limited by the availability of
Metadata vocabulary for Ontology Description ontology-based tools and services only for one
and publication, called MOD natural language, usually English, for which there
(https://github.com/sifrproject/MOD-Ontology). exist the most ontologies. Recently, ontology lo-
MOD 1.2 is defined in OWL and consists of 19 calization, i.e., “the process of adapting an ontolo-
classes and 88 properties most of them to describe gy to a concrete language and culture communi-
the mod:Ontology object. MOD 1.2 may serve as ty” (Cimiano et al., 2010), has become very im-
(i) a vocabulary to be used by ontology developers portant in the ontology development lifecycle, but
to annotate and describe their ontologies, or (ii) an when efforts are made to properly represent lexi-
explicit OWL ontology to be used by ontology li- cal (e.g., using Lemon (McCrae et al., 2011)) or
braries to offer semantic descriptions of ontologies multilingual information (e.g., using
as linked data. MOD 1.2 is an initiative which at- LexOMV (Montiel-Ponsoda et al., 2007) or Le-
tempts to overcome some of the limitations of the mon translation module (Gracia et al., 2014)) are
Ontology Metadata Vocabulary (Suarez-Figueroa made, it is rarely leveraged by ontology libraries
et al., 2005) but is still a temporary proposition and repositories. In the future, we need ontology
that will be discussed in the next months within repositories to entirely support interface and
the Research Data Alliance recently re-configured content internationalization (i.e., both display-
Vocabulary & Semantic Services Interest Group.6 ing user interfaces (e.g., menu names, help, etc.)
Automatic ontology selection or recommenda- in different languages and displaying their content
tion has been a subject of interest to facilitate on- (e.g., ontology labels, mappings, etc.) in different
tology reuse (Sabou et al., 2006)(Butt et al., languages) and be multilingual by enabling a
2016). The number and variety of ontologies in complete use of their functionalities and ser-
certain domains is now so large that choosing one vices for multilingual ontologies or monolin-
for an annotation task or for designing a specific gual ontologies linked one another.
application is quite cumbersome. In Jonquet et al., (2015), we presented a
In Martinez-Romero et al., (2017), we devel- roadmap for addressing the issues of dealing with
oped the NCBO Ontology Recommender. This multilingual or monolingual ontologies in the
service suggests relevant ontologies from the re- NCBO BioPortal, which takes English as primary
pository for annotating text data. The new rec- language. We proposed a set of representations to
6 https://www.rd-alliance.org/groups/vocabulary-services-
support multilingualism in the portal and to enable
interest-group.html a complete use of the functionalities and services
30
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
for any kind of ontologies and data: In Ghazvinian et al., (2009), we have analyzed
(i) Representation of natural language property for the mappings automatically generated within Bi-
an ontology; (ii) Representation of translation re- oPortal and what they tell us about the ontologies
lations between ontologies; (iii) Representation of themselves, the structure of the ontology reposito-
the distinction between ontologies with multilin- ry, and the ways in which the mappings can help
gual content i.e., multilingual and mono lingual in the process of ontology design and evaluation.
ontologies; (iv) Representation of multilingual This study demonstrated the value of having a
mappings. Those aspects have been addressed mapping repository goes beyond ontology-to-
now within MOD and/or the new AgroPortal ontology alignment, but concretely helps analyze
metadata model previously cited. In addition, the structures, dependencies and overlap of ontol-
in Annane et al., (2016b), we reconciled more ogies in the same domain. A similar, more recent
than 228K mappings between ten English ontolo- study about ontology terms reuse have been done
gies hosted on NCBO BioPortal and their French by Kamdar et al. (Kamdar et al., 2017). In Annane
translations hosted on the SIFR BioPortal. The et al., (2016a), we have also demonstrated that ex-
next big step is now to internationalize the portal. isting mappings between ontologies can also be
used to improve ontology alignment methods
3.3 Ontology alignment based on background knowledge; in other words,
Ontologies, or other semantic resources, will inev- a centralized mapping repository will also be an
itably overlap in coverage. Therefore, the need for excellent resource to curate and generate new
ontology alignment. This need has been explicitly mappings.
expressed by almost all our partner organizations
in biomedicine, agronomy or ecology. Surprising- 3.4 Generic ontology-based services
ly, it seems there is a gap between the state-of-the- Ontology repositories offer a large span of ser-
art results obtained at each edition of the Ontology vices: file hosting, versioning, search and browse
Alignment Evaluation Initiative (OAEI - content, visualization, metrics, notes, mapping,
http://oaei.ontologymatching.org) and the day-to- etc. These services are ‘generic’ if they are domain
day reality of ontology developers. Tools are often independent i.e., not specific to a domain, group
hardly reusable, and results cannot be easily re- of ontologies, specific format or design principles.
produced outside of the benchmarking effort. An- It is important that ontology repositories contin-
other key role of ontology repositories is to store ue to enhance ontology-based services and offer
mappings (or alignments) between ontologies. new generic ones to enlarge the spectrum of
Ontology repositories shall support the extrac- possible use of ontologies. Using standard for-
tion, generation, validation, evaluation, storage mats such as OWL or SKOS has facilitated the
and retrieval of mappings between the ontolo- development of a wide range of tools and services
gies they host. Automatic mapping generation for semantic resources. The challenge is now to
within ontology repositories shall go beyond sim- package them inside ontology repositories and
ple lexical or ID-based approaches7 and state-of- keep vertical quality (i.e., one ontology) while en-
the-art tools shall be incorporated within reposito- abling quantitative horizontal use.
ries. An equivalent effort, such as the one made to One important use of ontologies is for annotat-
harvest ontologies, must be made to harvest the ing and indexing text data (Spasic et al., 2005;
mappings between these ontologies and describe Handschuh and Staab, 2003). Therefore, we often
them with metadata and provenance information see aside of ontology repositories, ontology-based
to facilitate trust and reuse. annotation services. For instances, BioPortal has
the NCBO Annotator (Jonquet et al., 2009), OLS
7
To the best of our knowledge, only the NCBO technology had Whatizit (Rebholz-Schuhmann et al., 2008)
automatically computes ontology alignments when ontolo- and now moved to ZOOMA, HeTOP had
gies are hosted within the portal. The portal automatically FMTI (Sakji et al., 2010) and UMLS has Met-
creates some mappings when two classes share the same
identifiers properties, or when they share a common normal-
aMap (Aronson, 2001). Hereafter, we focus on
ized preferred label or synonym. Although basic lexical services for text data (annotation & terminology
mapping approaches can be inaccurate and should be used extraction).
with caution (Faria et al., 2014; Pathak and Chute, 2009), In Lossio-Ventura et al., (2014), we presented
they usually work quite well to interconnect
ontologies (Ghazvinian et al., 2009). BioTex, a Web application that implements state-
31
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
of-the-art measures for automatic extraction of bi- of more than twenty heterogeneous biomedical re-
omedical terms from English and French free text. sources (later extended to 50) included within Bi-
The application includes a new methodology for oPortal. Directly when browsing the ontologies or
automatic term extraction mixing linguistic, statis- using a dedicated search engine, users can discov-
tical, graph and Web-based approaches that have er datasets of interest. The indexing relied on the
been demonstrated quite efficient (Lossio-Ventura NCBO Annotator workflow and used the seman-
et al., 2015). Among other use of BioTex, we have tics that the ontologies encode, such as synonyms,
shown it can be part of an ontology enrichment class hierarchies, and the mappings between on-
workflow that could be highly valuable for ontol- tologies, to improve the search experience. The
ogy developers (Lossio-Ventura et al., 2016). Resource Index, was a tentative developed before
However, this work has not yet been incorporated 2010 that did not rely neither on big data technol-
within an ontology repository technology. ogies and did not followed linked open data prin-
In Tchechmedjiev et al.,(2017), we present mul- ciples. Both were in their infancies at that time.
tiple enhancement to the semantic annotation More recently, in agronomy, we have followed
workflow that we have developed on top of the new efforts such as AgroLD project (Venkatesan
NCBO Annotator and when building a French et al., 2015) to build a database of resources de-
version of the service. Some of these new func- scribed in RDF, and annotated with ontologies.
tionalities are particularly relevant to process elec- We are currently working on the interoperation of
tronic health records. These new features include: AgroLD and AgroPortal.
annotation scoring (Melzi and Jonquet, 2014), ad-
ditional output formats (for evaluation and inte- 3.6 Scalability & interoperability
gration with standard clinical systems), clinical In 2007, Swoogle claimed to “Search over 10.000
context detection (negation, experiencer and tem- ontologies”. Today, a simple Google Search for
porality through the integration of the Neg- “filetype:owl” returns around 34K results. The
Ex/ConText algorithm) (Abdaoui et al., 2017), NCBO BioPortal, which is generally considered
coarse-grained entity type annotations (using has the biggest ontology repository (not library)
UMLS Semantic Groups, e.g., anatomy, disorders, contains +650 ontologies as of end of 2017. More
devices). and more vocabularies are being developed and
hosted by the LOV platform. Multiple domain
3.5 Annotations and Linked Data specific ontology repository efforts have started
Data integration and semantic interoperability en- often inspired by results in the biomedical domain
able new scientific discoveries that could be made and usually by reusing NCBO technology (e.g.,
by merging different currently available data. MMI OOR, AgroPortal, ESIPPortal). The more
These is one major reason for adopting ontologies. ontologies and ontology repositories are being
They are used to design semantic indexes of data developed, the more scalability and interoper-
and linked open datasets that could be used for ability issues become important. Some ontolo-
various type of cross datasets studies (Handschuh gies are useful to different communities and shall
and Staab, 2003; Bizer et al., 2009). Ontology re- then be hosted in multiple repositories e.g., do-
positories must facilitate indexing/annotation, main ontologies such as the Gene
search and access to semantically described, in- Ontology (Ashburner et al., 2000), or the Envi-
teroperable, actionable, open, rich linked data ronment Ontology (Buttigieg et al., 2013). Be-
directly from the within the repositories. Work- cause no repository will host them all, ontology
ing with big data represents a set of challenges for repositories have to offer a certain level of in-
ontology repositories when designing these se- teroperability to ensure their users that they will
mantic indexes: scalability, consistency, com- not have to work with multiple web applications
pleteness in a context where both ontologies and and programming interfaces if their ontologies of
data constantly evolve. In addition, cross ontolo- interest are not all hosted by the same repositories.
gies semantics and indexed data consistency shall As previously explained standard ontology
be checked by ontology repositories using OWL metadata is a crucial aspect to achieve this.
reasoning. In Jonquet et al., ; Jonquet et al., (2016b), our
In Jonquet et al., (2011), we have built the projects described Section 2.3, we have been par-
NCBO Resource Index, an ontology-based index ticularly careful in not redeveloping features and
32
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
functionalities that to our knowledge were already tures supported by the common model. We see
available. We have designed and implemented two two general scenarios of use for these repositories:
advanced prototype ontology repositories for the x The repositories provide basic ontology li-
French biomedical community and for the agron- brary services for users with a “vertical need”
omy domain. Our choice to reuse the NCBO tech- —those who want to do very precise things
nology was justified by the large spectrum of fea- (e.g., reasoning, using specific relations) us-
tures and services, but in addition our motivation ing only suitable ontologies (developed by the
was: (i) to avoid re-developing tools that have al- same communities and in the same format).
ready been designed and extensively used and Such users may just use the repositories as li-
contribute to long term support of the commonly braries to find and download ontologies, and
used technology; and (ii) to offer the same tools, work in their own environment.
services and formats to different but still intercon- x The repositories provide many ontology-
nected communities, to facilitate the interface and based services to users with “horizontal
interaction between their domains (agro, bio, needs” —those who wants to work with a
wide range of ontologies and vocabularies
health (French)). Relying on the same original
useful in their domain but developed by dif-
technology enhance both technical reuse (for ex-
ferent communities, overlapping and in dif-
ample, enabling queries to either systems with the ferent formats. Such users greatly appreciate
same code), and semantic reuse. Then, we have the unique endpoints (Web application and
developed new functionalities –as previously de- programmatic for REST and SPARQL que-
scribed– while keeping our systems backward ries) offered by the repositories under a sim-
compatible with the original technology to facili- plified common model.
tate a convergence of the efforts. We strongly be- In this position paper, we have unfortunately not
lieve that sharing the technology is the best way to covered all related work on the cited challenges
guaranty long term support and development by and we have certainly skipped other important
engaging different ontology practitioners and challenges: semantic consistency, ontology evalu-
communities all around the world with their re- ation, visualization, community feedback. But we
spective funding and supporting schemes. Also, offered a short summary of multiple various con-
sharing the technology is the best way to make on- tributions on ontology repository and ontology-
tology repositories interoperable. As explained based service research. In the future, we will con-
in Tchechmedjiev et al.,(2017), all of the new fea- tinue our efforts to address the identified challeng-
tures implemented (e.g., NCBO Annotator + or es (and others), while continue to offer to various
the new Recommender) are available across any scientific communities the means to share and
other NCBO based platform at minimum cost. leverage their ontologies or semantic resources
and enable new science in their fields.
4 Conclusions
Acknowledgments
In this paper, we have presented our vision on
challenges and issues in building ontology reposi- This work is partly achieved within the Semantic
tories. We have illustrated our thoughts with re- Indexing of French biomedical Resources (SIFR –
sults obtained over the last 10 years within our www.lirmm.fr/sifr) project that received funding
projects in biomedicine and agronomy. By adopt- from the French National Research Agency (grant
ing NCBO technology, we inherit some ad- ANR-12-JS02-01001), the European Union’s
vantages and inconvenients but we can now con- Horizon 2020 research and innovation programme
tribute to this field of research with concrete use under the Marie Sklodowska-Curie grant agree-
cases, communities and outcomes. NCBO-based ment No 701771, the NUMEV Labex (grant
ontology repositories adopted a vision where mul- ANR-10-LABX-20), the Computational Biology
tiple semantic resources are made available in a Institute of Montpellier (grant ANR-11-BINF-
common place (though not combined and con- 0002), as well as by the University of Montpellier
sistency checked), and cast to a common model. and the CNRS. I also acknowledge the National
While doing so, the repositories arguably limits Center for Biomedical Ontologies for their in-
the full power of ontologies –which has been a re- sights and thanks all my collaborators in Montpel-
current criticism– constraining their use to fea- lier or Stanford interested like me on ontology re-
positories.
33
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
References Gàmez-Pérez. 2010. A note on ontology
localization. Applied Ontology, 5(2):127–137.
Amine Abdaoui, Andon Tchechmedjiev, William
Digan, Sandra Bringay, and Clement Jonquet. Richard G Côté, Philip Jones, Rolf Apweiler, and
2017. French ConText: a Publicly Accessible Henning Hermjakob. 2006. The Ontology Lookup
System for Detecting Negation, Temporality and Service, a lightweight cross-platform tool for
Experiencer in French Clinical Notes. Biomedical controlled vocabulary queries. BMC
Informatics, UNDER REVI. Under review. bioinformatics, 7:97, January.
Amina Annane, Zohra Bellahsene, Faical Azouaou, Mathieu D’Aquin, Claudio Baldassarre, Laurian
and Clement Jonquet. 2016a. Selection and Gridinoc, Sofia Angeletou, Marta Sabou, Enrico
Combination of Heterogeneous BK to Enhance Motta, Mathieu Aquin, Claudio Baldassarre,
Biomedical Ontology Matching. In 20th Laurian Gridinoc, Sofia Angeletou, Marta Sabou,
International Conference on Knowledge Enrico Motta, Mathieu D’Aquin, Claudio
Engineering and Knowledge Management, Baldassarre, Laurian Gridinoc, Sofia Angeletou,
EKAW’16, volume 10024 LNAI, Bologna, Italy. Marta Sabou, Enrico Motta, Mathieu Aquin, et al.
2007. Watson: A Gateway for Next Generation
Amina Annane, Vincent Emonet, Clement Jonquet, Semantic Web Applications. In Poster &
Faical Azouaou, and Clement Jonquet. 2016b. Demonstration Session at the 6th International
Multilingual Mapping Reconciliation between Semantic Web Conference, ISWC’07, page 3,
English-French Biomedical Ontologies. In 6th Busan, Korea.
International Conference on Web Intelligence,
Mining and Semantics, WIMS’16, volume 13–15– 2010. 1st Workshop on Ontology Repositories and
June, page 12, Nimes, France. ACM. Editors for the Semantic Web, ORES’10. In
Mathieu d’Aquin, Alexander G Castro, Christoph
Alan R Aronson. 2001. Effective mapping of Lange, and Kim Viljanen, editors, number 596,
biomedical text to the UMLS Metathesaurus: the Hersonissos, Greece. CEUR-WS.org.
MetaMap program. In American Medical
Informatics Association Annual Symposium, Mathieu D’Aquin and Natasha F Noy. 2012. Where to
AMIA’01, pages 17–21, Washington, DC, USA. Publish and Find Ontologies? A Survey of
Ontology Libraries. Web semantics,
Michael Ashburner, Catherine A Ball, Judith A Blake, 11(August):96–111, March.
David Botstein, Heather Butler, J Michael Cherry,
Allan P Davis, Kara Dolinski, Selina S Dwight, Li Ding, Tim Finin, Anupam Joshi, Yun Peng, R Scott
Janan T Eppig, and others. 2000. Gene Ontology: Cost, Joel Sachs, Rong Pan, Pavan Reddivari, and
tool for the unification of biology. Nature genetics, Vishal Doshi. 2004. Swoogle: A Semantic Web
25(1):25–29. Search and Metadata Engine. In D A Grossman, L
Gravano, C Zhai, O Herzog, and D Evans, editors,
Kenneth Baclawski and Todd Schneider. 2009. The 13th ACM Conference on Information and
open ontology repository initiative: Requirements Knowledge Management, CIKM’04, Washington
and research challenges. In T Tudorache, G DC, USA. ACM.
Correndo, N Noy, H Alani, and M Greaves, editors,
Workshop on Collaborative Construction, Ying Ding and Dieter Fensel. 2001. Ontology Library
Management and Linking of Structured Systems: The key to successful Ontology Re-use.
Knowledge, CK’09, volume 514, page 10, In 1st Semantic Web Working Symposium,
Washington, DC., USA. CEUR-WS.org. SWWS’01, pages 93–112, Stanford, CA, USA.
CEUR-WS.org.
Christian Bizer, Tom Heath, and Tim Berners-Lee.
2009. Linked Data - The Story So Far. Semantic Biswanath Dutta, Anne Toulet, Vincent Emonet, and
Web and Information Systems, 5(3):1–22. Clement Jonquet. 2017. New Generation Metadata
vocabulary for Ontology Description and
Anila Sahar Butt, Armin Haller, and Lexing Xie. Publication. In E Garoufallou, S Virkus, and G
2016. RecOn: Ontology recommendation for Alemu, editors, 11th Metadata and Semantics
structureless queries. Applied Ontology, 11(4):301– Research Conference, MTSR’17, Tallinn, Estonia.
324.
Daniel Faria, Ernesto Jiménez-Ruiz, Catia Pesquita,
Pier Luigi Buttigieg, Norman Morrison, Barry Smith, Emanuel Santos, and Francisco M Couto. 2014.
Christopher J Mungall, and Suzanna E and Lewis. Towards Annotating Potential Incoherences in
2013. The environment ontology: contextualising BioPortal Mappings. In 13th International
biological and biomedical entities. Biomedical Semantic Web Conference, ISWC’13, volume 8797,
Semantics, 4(1):43. pages 17–32, Riva del Garda, Italy. Springer.
Philipp Cimiano, Elena Montiel-Ponsoda, Paul Amir Ghazvinian, Natasha F Noy, Clement Jonquet,
Buitelaar, Mauricio Espinoza, and Asuncion Nigam H Shah, and Mark A Musen. 2009. What
34
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
Four Million Mappings Can Tell You about Two Clement Jonquet, Mark A Musen, and Nigam H Shah.
Hundred Ontologies. In A Bernstein, D R Karger, T 2010. Building a Biomedical Ontology
Heath, L Feigenbaum, D Maynard, E Motta, and K Recommender Web Service. Biomedical
Thirunarayan, editors, 8th International Semantic Semantics, 1(S1). Selected in Pr. R. Altman’s 2011
Web Conference, ISWC’09, volume 5823, pages Year in Review at AMIA TBI.
229–242, Washington DC, USA. Springer.
Clement Jonquet, Nigam H Shah, and Mark A Musen.
Jorge Gracia, Elena Montiel-Ponsoda, Daniel Vila- 2009. The Open Biomedical Annotator. In
Suero, and Guadalupe Aguado-de Cea. 2014. American Medical Informatics Association
Enabling Language Resources to Expose Symposium on Translational BioInformatics,
Translations as Linked Data on the Web. In N AMIA-TBI’09, pages 56–60, San Francisco, CA,
Calzolari, K Choukri, T Declerck, H Loftsson, B USA.
Maegaard, J Mariani, A Moreno, J Odijk, and S
Clément Jonquet, Anne Toulet, Elizabeth Arnaud,
Piperidis, editors, 9th International Conference on
Sophie Aubin, Esther Dzalé-Yeumo, Vincent
Language Resources and Evaluation, LREC’14,
Emonet, John Graybeal, Mark A. Musen, Cyril
pages 409–4013, Reykjavik, Iceland. European
Pommier, and Pierre Larmande. 2016c. Reusing
Language Resources Association.
the NCBO BioPortal technology for agronomy to
Siegfried Handschuh and Stephen Staab, editors. build AgroPortal. In P Jaiswal and R Hoehndorf,
2003. Annotation for the Semantic Web.volume 96 editors, 7th International Conference on
of Frontiers in Artificial Intelligence and Biomedical Ontologies, ICBO’16, Demo Session,
Applications. IOS Press. volume 1747, pages 4–6, Corvallis, Oregon, USA.
Jens Hartmann, Raúl Palma, and Asunción Gómez- Clement Jonquet, Anne Toulet, Elizabeth Arnaud,
Pérez. 2009. Ontology Repositories. Handbook on Sophie Aubin, Esther Dzalé Yeumo, Vincent
Ontologies:551–571. Emonet, John Graybeal, Marie-Angélique Laporte,
Mark A Musen, Valeria Pesce, and Pierre
Robert Hoehndorf, Luke Slater, Paul N Schofield, and
Larmande. 2017a. AgroPortal: an ontology
Georgios V Gkoutos. 2015. Aber-OWL: a
repository for agronomy. Computers and
framework for ontology-based data access in
Electronics in Agriculture, IN PRESS:30. Under
biology. BMC Bioinformatics, 16(1):1–9.
review.
Clement Jonquet, Amina Annane, Khedidja Bouarech,
Clement Jonquet, Anne Toulet, Biswanath Dutta, and
Vincent Emonet, and Soumia Melzi. 2016a. SIFR
Vincent Emonet. 2017b. Harnessing the power of
BioPortal : Un portail ouvert et générique
unified metadata in an ontology repository: the
d’ontologies et de terminologies biomédicales
case of AgroPortal. Data Semantics, UNDER
françaises au service de l’annotation sémantique. In
REVI. Under review.
16th Journées Francophones d’Informatique
Médicale JFIM’16. Maulik R Kamdar, Tania Tudorache, and Mark A
Musen. 2017. A Systematic Analysis of Term
Clement Jonquet, Amina Annane, Khedidja Bouarech,
Reuse and Term Overlap across Biomedical
Vincent Emonet, and Soumia Melzi. 2016b. SIFR
Ontologies. Semantic web, 8(6):853–871.
BioPortal : Un portail ouvert et générique
d’ontologies et de terminologies biomédicales Juan-Antonio Lossio-Ventura, Clement Jonquet,
françaises au service de l’annotation sémantique. In Mathieu Roche, Maguelonne Teisseire, Juan
16th Journées Francophones d’Informatique Antonio, Lossio-ventura Clement Jonquet, Mathieu
Médicale, JFIM’16. Roche, and Maguelonne Teisseire. 2015.
Biomedical term extraction: overview and a new
Clement Jonquet, Vincent Emonet, and Mark A
methodology. Information Retrieval, Special issue
Musen. 2015. Roadmap for a multilingual
on Medical Information Retrieval, 19(1):59–99.
BioPortal. In J Gracia, J P McCrae, and G Vulcu,
editors, 4th Workshop on the Multilingual Semantic Juan Antonio Lossio-Ventura, Clement Jonquet,
Web, MSW4’15, volume 1532, pages 15–26, Mathieu Roche, and Maguelonne Teisseire. 2016.
Portoroz, Slovenia. A Way to Automatically Enrich Biomedical
Ontologies. In 19th International Conference on
Clement Jonquet, Paea LePendu, Sean Falconer,
Extending Database Technology, EDBT’16, Poster
Adrien Coulet, Natalya F Noy, Mark A Musen, and
Session, number 305, page 2, Bordeaux, France.
Nigam H Shah. 2011. NCBO Resource Index:
OpenProceedings.org.
Ontology-Based Search and Mining of Biomedical
Resources. Web Semantics, 9(3):316–324. 1st prize Juan Antonio Juan Antonio Lossio-Ventura, Clement
of Semantic Web Challenge at the 9th International Jonquet, Mathieu Roche, and Maguelonne
Semantic Web Conference, ISWC’10, Shanghai, Teisseire. 2014. BIOTEX: A system for Biomedical
China. Terminology Extraction, Ranking, and Validation.
35
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
In M Horridge, M Rospocher, and J Ossenbruggen, Edison Ong, Zuoshuang Xiang, Bin Zhao, Yue Liu,
editors, 13th International Semantic Web Yu Lin, Jie Zheng, Chris Mungall, Mélanie
Conference, Demonstration, ISWC’14, volume Courtot, Alan Ruttenberg, and Yongqun He. 2016.
1272, pages 157–160, Riva del Garda, Italy. Ontobee: A linked ontology data server to support
ontology term dereferencing, linkage, query and
Marcos Martinez-Romero, Clement Jonquet, Martin J
integration. Nucleic acids research, 45(D1):D347–
O’Connor, John Graybeal, Alejandro Pazos, and
D352, October.
Mark A Musen. 2017. NCBO Ontology
Recommender 2.0: An Enhanced Approach for Jyotishman Pathak and Christopher G Chute. 2009.
Biomedical Ontology Recommendation. Debugging Mappings between Biomedical
Biomedical Semantics, 8(21). Ontologies: Preliminary Results from the NCBO
BioPortal Mapping Repository . In B Smith, editor,
John McCrae, Dennis Spohr, and Philipp Cimiano.
International Conference on Biomedical Ontology,
2011. Linking lexical resources and ontologies on
pages 95–98, Buffalo, NY, USA.
the semantic web with lemon. In G Antoniou, M
Grobelnik, E Simperl, B Parsia, D Plexousakis, P Depriest A Pouchard L. Huhns M. 2012. Lessons
DeLeenheer, and J Z Pan, editors, 8th Extended learned in deploying a cloud-based knowledge
Semantic Web Conference, ESWC’11, number platform for the ESIP Federation. American Geo-
6643, pages 245–259, Heraklion, Crete, Greece. physical Union Fall Meeting, poster
Springer. session:22725.
Peter McQuilton, Alejandra Gonzalez-Beltran, Dietrich Rebholz-Schuhmann, Miguel Arregui,
Philippe Rocca-Serra, Milo Thurston, Allyson Sylvain Gaudan, Harald Kirsch, and Antonio
Lister, Eamonn Maguire, and Susanna-Assunta Jimeno. 2008. Text processing through Web
Sansone. 2016. BioSharing: curated and crowd- services: Calling Whatizit. Bioinformatics,
sourced metadata standards, databases and data 24(2):296–298, January.
policies in the life sciences. Database, baw075.
Carlos Rueda, Luis Bermudez, and Janet Fredericks.
Soumia Melzi and Clement Jonquet. 2014. Scoring 2009. The MMI Ontology Registry and Repository:
semantic annotations returned by the NCBO A Portal for Marine Metadata Interoperability. In
Annotator. In A Paschke, A Burger, P Romano, M S MTS/IEEE Biloxi - Marine Technology for Our
Marshall, and A Splendiani, editors, 7th Future: Global and Local Challenges,
International Semantic Web Applications and Tools OCEANS’09, page 6, Biloxi, MS, USA.
for Life Sciences, SWAT4LS’14, volume 1320, page
Marta Sabou, Vanessa Lopez, and Enrico Motta. 2006.
15, Berlin, Germany. CEUR-WS.org.
Ontology Selection on the Real Semantic Web:
Elena Montiel-Ponsoda, Guadalupe Aguado de Cea, How to Cover the Queens Birthday Dinner? In S
Mari Carmen Suarez-Figueroa, Raul Palma, Staab and V Svátek, editors, 15th International
Asuncion Gomez-Pérez, and Wim Peters. 2007. Conference on Knowledge Engineering and
LexOMV: an OMV extension to capture Knowledge Management Managing Knowledge in
multilinguality. In P Buitelaar, K Choi, A Gangemi, a World of Networks, EKAW’06, volume 4248,
C Huang, and A Oltramari, editors, pages 96–111, Podebrady, Czech Republic.
Lexicon/Ontology Interface Workshop, Springer.
OntoLex’07, page 10, Busan, South-Korea.
Saoussen Sakji, Q Gicquel, S Pereira, I Kergoulay, D
Debashis Naskar and Biswanath Dutta. 2016. Proux, Darmoni SJ, and M H Metzger. 2010.
Ontology And Ontology Libraries: A Study From Evaluation of a French Medical Multi-Terminology
An Ontofier And An Ontologist Perspective. In Indexer for the Manual Annotation of Natural
19th International Symposium on Electronic Language Medical Reports of Healthcare-
Theses and Dissertations, ETD’16, pages 1–12, Associated Infections. In C Safran et al., editor,
Lille, France. 13th World Congress on Medical Informatics,
MedInfo’10, volume 160, pages 252–256, Cape
Natalya F Noy, Nigam H Shah, Patricia L Whetzel,
Town, South Africa. IOS Press.
Benjamin Dai, Michael Dorf, Nicholas B Griffith,
Clement Jonquet, Daniel L Rubin, Margaret-Anne Manuel Salvadores, Paul R Alexander, Mark A
Storey, Christopher G Chute, Mark A Musen, Musen, and Natalya F Noy. 2013. BioPortal as a
Cherie H Youn, Adrien Coulet, Chris Callendar, dataset of linked biomedical ontologies and
and Barry Smith. 2009. BioPortal: ontologies and terminologies in RDF. Semantic Web, 4(3):277–
integrated data resources at the click of a mouse. In 284.
B Smith, editor, Nucleic Acids Research, volume
Barry Smith, Michael Ashburner, Cornelius Rosse,
37, pages 170–173, Stockholm, Sweden, July.
Jonathan Bard, William Bug, Werner Ceusters,
Louis J Goldberg, Karen Eilbeck, Amelia Ireland,
36
Position paper – Keynote SIMBig 2017 – September 2017, Lima, Peru
Christopher J Mungall, The O B I Consortium, Jildau Bouwman, Anthony J Brookes, Tim Clark,
Neocles Leontis, Philippe Rocca-Serra, Alan Mercè Crosas, Ingrid Dillo, Olivier Dumon, Scott
Ruttenberg, Susanna-Assunta Sansone, Richard H Edmunds, Chris T Evelo, Richard Finkers, et al.
Scheuermann, Nigam H Shah, Patricia L Whetzel, 2016. The FAIR Guiding Principles for scientific
and Suzanna Lewis. 2007. The OBO Foundry: data management and stewardship. Scientific Data,
coordinated evolution of ontologies to support 3.
biomedical data integration. Nature Biotechnology,
1995. A case study in ontology library construction.
25(11):1251–1255.
Artificial Intelligence in Medicine, 7(3):227–255,
Irena Spasic, Sophia Ananiadou, John McNaught, and June.
Anand Kumar. 2005. Text mining and ontologies in
biomedicine: making sense of raw text. Briefings in
Bioinformatics, 6(3):239–251.
Suarez-Figueroa, Jens Hartmann, York Sure, Peter
Haase, and M. Suarez-Figueroa. 2005. OMV–
ontology metadata vocabulary. In C. Welty, editor,
Workshop on Ontology Patterns for the Semantic
Web, WOP’05, page 9, Galway, Irland. Springer.
Andon Tchechmedjiev, Amine Abdaoui, Vincent
Emonet, and Clement Jonquet. 2017a. ICD10
Coding of Death Certificates with the NCBO and
SIFR Annotator(s) at CLEF eHealth 2017 Task 1.
In Working Notes of CLEF eHealth Evaluation
Lab, volume 1866, page 16, Dublin, Ireland.
Andon Tchechmedjiev, Amine Abdaoui, Vincent
Emonet, Soumia Melzi, Jitendra Jonnagaddala, and
Clement Jonquet. 2017b. Enhanced Functionalities
for Annotating and Indexing Clinical Text with the
NCBO Annotator+. Bioinformatics, UNDER
REVI. Under Review.
Mossakowski Till, Oliver Kutz, and Mihai Codescu.
2014. Ontohub: A semantic repository for
heterogeneous ontologies. In Theory Day in
Computer Science, DACS’14, page 2, Bucharest,
Romania.
Pierre-Yves Vandenbussche, Ghislain A Atemezing,
Maria Poveda-Villalon, and Bernard Vatant. 2014.
Linked Open Vocabularies (LOV): a gateway to
reusable semantic vocabularies on the Web.
Semantic Web, 1:1–5.
Aravind Venkatesan, Nordine El Hassouni, Florian
Phillipe, Cyril Pommier, Hadi Quesneville, Manuel
Ruiz, and Pierre Larmande. 2015. Exposing French
agronomic resources as Linked Open Data. In
James Malone, Robert Stevens, Kerstin Forsberg,
and Andrea Splendiani, editors, 8th International
Conference on Semantic Web Applications and
Tools for Life Sciences, SWAT4LS’15, pages 205–
207, Cambridge, UK. CEUR.
Patricia L Whetzel and NCBO Team. 2013. NCBO
Technology: Powering semantically aware
applications. Biomedical Semantics, 4S1(S8):49.
Mark D Wilkinson, Michel Dumontier, Ijsbrand Jan
Aalbersberg, Gabrielle Appleton, Myles Axton,
Arie Baak, Niklas Blomberg, Jan-Willem Boiten,
Luiz Bonino da Silva Santos, Philip E Bourne,
37