=Paper=
{{Paper
|id=None
|storemode=property
|title=Teaching Good Biomedical Ontology Design
|pdfUrl=https://ceur-ws.org/Vol-897/sessionJ-paper25.pdf
|volume=Vol-897
|dblpUrl=https://dblp.org/rec/conf/icbo/BoekerSRGRJS12
}}
==Teaching Good Biomedical Ontology Design==
https://ceur-ws.org/Vol-897/sessionJ-paper25.pdf
Teaching Good Biomedical Ontology Design
Martin Boeker 1∗, Daniel Schober 1 , Djamila Raufie 1 , Niels Grewe 2 , Johannes Röhl 2 ,
Ludger Jansen 2 , and Stefan Schulz 1,3
1
Institute of Medical Biometry and Medical Informatics, University Medical Center Freiburg, Germany
2
Institute of Philosophy, University of Rostock, Germany
3
Institute for Medical Informatics, Statistics and Documentation, Medical University of Graz, Austria
ABSTRACT 1 INTRODUCTION
Background: In order to improve ontology quality, tool- and language- Ontology engineering continues to be an area of major interest
related tutorials are not sufficient. Care must be taken to provide within the life sciences, as computer-interpretable domain repre-
optimized curricula for teaching the representational language in the sentations are the only viable option for an efficient and intelligent
context of a semantically rich upper level ontology. The constraints exploitation of the vast amounts of high-throughput-generated data.
provided by rigid top and upper level models assure that the onto- Although numerous ontologies are publically available through
logies built are not only logically consistent but also adequately ontology libraries and access portals, they remain of heterogene-
represent the domain of discourse and align to explicitly outlined onto- ous quality and ontological rigor (Smith et al., 2004; Schulz et al.,
logical principles. Finally such a curriculum must take into account the 2009; Rector et al., 2011; Boeker et al., 2011). Coordination efforts
pre-existing skills and knowledge of the target audience. and best practice providers, such as the OBO Foundry (Smith et al.,
Objective: To develop a well-structured curriculum aligned to the par- 2007) have recently emerged from the need to assure at least some
ticular requirements of life science professionals, in order to enable basic quality with respect to ontological correctness and usability.
them to create logically sound, domain adequate and predicable While these efforts mainly target already experienced ontologists,
ontologies using the Web Ontology Language (OWL) in Protégé. only few efforts have recently focused on teaching basic notions of
Methods: Content selection for the curriculum was based on the ontology engineering to novices. This imbalance results in a grow-
literature, pre-existing tutorials, and a guideline for good onto- ing number of practitioners being forced to apply particular design
logy development (i.e ontology design enhancing domain adequacy, requirements and patterns as requested by the aforementioned policy
sustainability and interoperability) that drew on the authors previ- providers, but without even knowing the most basic foundations in
ous experiences with large ontology development projects. Learning the semantics of the representation language used.
objectives were formulated according to a needs assessment of the Another major obstacle in ontology design is that various scien-
targeted learners, who were students trained in life sciences with tific and engineering communities sustain different and sometimes
basic knowledge and practical skills in computer science. As instruc- even contradictory modeling objectives and paradigms. Although
tional format we choose an approach with a high amount of practical skilled logicians or computer scientists should normally understand
exercises. The curriculum was first implemented with 24 Students and the formal semantics of description logics, their approaches to
7 lecturers/ tutors over 5 full days. The curriculum was evaluated by certain modeling tasks are often based on other paradigms, like data-
gathering the participants feedback via a questionnaire. base technologies or object-oriented programming. Object-oriented
Results: Curricular development produced 16 modules of approxima- programming uses, for example, inheritance principles that dif-
tely 2 hours each, which covered basic principles of Applied Ontology, fer from the inheritance principle of description logics. While the
description logic syntax and semantics, as well as best design practi- inheritance principles of modern object oriented programming are
ces outlined in ontology design patterns and variants of the BioTop based on the substitution principle (Liskov, 1987), inheritance in
upper ontology. An opinion survey based on questionnaires indica- description logics is based on set theory (Baader et al., 2007).
ted that the participants took advantage from the teaching strategies Thus, programmers operating with classes in object-oriented pro-
applied, as they indicated good knowledge gain and acknowledged gramming as abstract types (signatures or interfaces) which can be
the relevance of the modules. The difficulty was rated slightly lower. instantiated and subtyped might have problems to apply description
Conclusion: The development of teaching material for principled onto- logics based classes as mere sets in a bottom-up approach. The
logy design and best practices is of crucial importance in order to inheritance/ subtyping concepts of modern programming languages
enhance the quality of biomedical ontologies. Here, we present a cur- allow the engineer for very powerful top-down driven approaches in
riculum for a week long workshop, leveraging on current educational the modeling of complex software architectures (with method over-
principles, focusing on interactive hands-on exercises, group inter- riding, method overloading and polymorphic types) that have no
actions, and problem-oriented learning. Whereas evaluation clearly correspondences in DL. While a top-down modeling approach in
showed the success of this approach, in particular regarding student’s class hierarchies is best-practice in object-oriented software para-
satisfaction, the objective measurement of traceable effects on the digms, it will cause trouble when transferred to the design of an
quality of the generated ontology, although of much higher interest, ontology that is based on description logics with a deviant formal
has just started. semantics. This does, of course, not diminish the usefulness of the
respective techniques and skills in their own field of application.
∗ To whom correspondence should be addressed: martin.boeker@uniklinik-
We only want to point out that they do not comply with the rigid
modeling requirements of description logics.
freiburg.de
1
Boeker et al
In the light of the rising demand for of formal representation of, 4. computer scientists who provide the computational framework
e.g., biomedical knowledge (Ashburner et al., 2000), such limited or for development and deployment (e.g. knowledge representa-
even mislead ontology engineering skills can lead to wrong design tion, natural language processing, software engineering).
decisions in biomedical ontology. Finally the philosophical founda- Several challenges have to be met in the ontology engineering
tions of ontology engineering, as concretized in the recent discipline process, starting with the usage of editing and reasoning artifacts
of Applied Ontology (Munn and Smith, 2008) is still largely ignored with all their idiosyncrasies and computational requirements, follo-
by ontology engineers with a computer science background (Mizo- wed by putting into practice the collaborative editing of “single-file”
guchi and Kozaki, 2009). Explicitly, we do not claim to know and or modularly structured ontologies.
teach the one and only best ontology, but we provide a robust onto- The scientific community has made considerable progress in the
logical backbone on the basis of explicitly outlined design principles understanding and development of large ontologies. Besides new
rooted in traceable/ decidable logics. efforts in continuing the generation of knowledge, we must under-
In order to arrive at a sustainable development of sound ontology stand how to educate students to become ontology engineering
artifacts, it is necessary that there are ontology developers with (1) experts. It is not acceptable that just a small community is able to
a technical background in computer science that will (2) closely co- understand a topic of high interest, which, in contrast, is often con-
operate with experts for the domain to be represented, and have (3) sidered rather esoteric by those who are supposed to use the artifacts
a foundational grounding in the principles of Applied Ontology and created. In our view, well-instructed domain experts are needed to
the logical formalisms underlying ontology description languages. build good ontologies, because they are the only ones who know
The basic idea of our curriculum is to teach such principles as a what has to be represented and how, and who will later engage
series of robust and well-calibrated ontological building blocks, tar- in the dissemination and use of ontology-enabled tools. Domain
geting students in life sciences disciplines with some background in experts with some computer science background should be edu-
computer sciences. It should teach the essential skills to novices to cated appropriately to understand the basics of philosophical and
develop sound OWL-DL ontologies based on approved methods in logical foundations needed for ontology development and be trained
formal ontology. in real-life ontology engineering.
The need for a curriculum on biomedical ontology 2 METHODS
Some educational material on formal ontology engineering for The development of a specific curriculum for building OWL-DL
non-computer scientists is publicly available, and widely used ontologies is part of a larger project in which the key elements for
for self-training or in tutorials, esp. the Pizza Ontology tutorial the design of qualitatively good ontologies are defined (GoodOD:
(Protégé-Tutorial)1 based on the Pizza ontology2 and “Ontology Good Ontology Design). In this regard, the curriculum objectives
Development 101”3 or publicly available course material4 . These and contents mirror major parts of the goals of the GoodOD project.
tutorials are well structured and found wider acceptance, but suffer
from some limitations. They are often written as quick introducti- Terminological decisions
ons and user guides to the Protégé OWL editor and only partially Initiation to ontology design faces the problem that every commu-
address more complex practically occurring modeling problems, but nity involved has its own terminology. There are the vocabularies
rather focus on easy-to-understand examples. They do not teach how of philosophical ontology, set theory, DL, Protégé, knowledge
the ontologies created should be structured, and which tools lend management, and so on. This can easily lead to ambiguities and
themselves for sustainable design decisions. accordingly to confusion on the side of the learner. We decided to
As the recent debate on the scientific value of formal ontology in choose a rigid front end orientation and to use Protégé vocabulary
the life sciences has shown, the development of real life ontologies wherever possible, as this is what modelers see when working with
applying decidable logics is an interdisciplinary and highly com- this editor.
plex process, whose inter-dependencies are not yet fully understood
(Brochhausen et al., 2011). Typically, four scientific communi- Curricular development
ties with overlapping concerns and tasks will be involved in the The curricular development followed a widely adopted method from
development of an Ontology: medical education (Kern et al., 1998), which uses a six-step itera-
1. Experts from one or more domains who provide the insights in tive development procedure: the problem identification and general
what should be represented (e.g. medicine or biology), needs assessment, the needs assessment of the targeted learners, the
2. philosophers as experts of the principles of ontology (as a definition of goals and specific measurable objectives, the selection
philosophical discipline), of educational strategies, the implementation of the curriculum,
concluded by evaluation and feedback.
3. logicians and mathematicians as experts both of the mathema-
Based on a general requirements analysis as given in the back-
tical formalisms used for building ontology axioms and their
ground section, we identified the targeted learners of the curriculum
computational properties; and
as employees or students in the life sciences with a background
in computer science, either as a minor subject or part of a bio-
informatics curriculum. The needs of these learners with regard to
1 http://owl.cs.manchester.ac.uk/tutorials/protegeowltutorial/ ontology development were assessed.
2 http://www.co-ode.org/ontologies/pizza/
3 http://protege.stanford.edu/publications/ontology development/ontology101- Content selection
noy-mcguinness.html Prior to the curriculum development, a guideline had been deve-
4 http://www.meteck.org/teaching/SA/MOWS10OntoEngCouse.html loped, in which the authors elucidated the principles of good
2
Teaching Ontology
biomedical ontology design using a decidable description logics in Module 16 – Spatial disjointness ODP
Design Patterns
OWL format. The objective of the guideline is to provide practical Representing organ parts with the spatial disjointness ODP
Ontology
guidance for novices and experts on how to use the abovementioned Module 15 – Closure ODP
Representing an animal taxonomy and use the closure ODP
representation framework, and how to address ontology engineering
Module 14 – Introduction in Ontology Design Patterns (ODP)
projects using top-level ontologies and ontology design patterns.
Understang ODPs and using the exception ODP
Based on this guideline and in view of the curriculum time con-
Module 13 – Information objects
straints the most appropriate content for the learning objectives was Representing plans and documents on animals
Using top-level
selected. Module 12 – Non-material physical objects
ontologies
The main step in curricular development is the formulation of Representing habitats and enclosures of animals
goals for the complete curriculum and specific educational objecti- Module 11 – Collective entities
ves. Based on the problem analysis, and the general and targeted Representing feed and groups for animals
learners needs assessment, learning objectives and goals for the Module 10 – Process and participation
curriculum were specified. Although literature-based, the final sele- Representing locomotion and development of animals
ction of educational objectives was led by personal experience in a Module 9 – Introduction in the BioTop domain top-level ontology
Using the basic features of BioTop
series of life science ontology development projects (Boeker et al.,
Module 8 – Typical ontology design errors
2007; Beisswanger et al., 2008; Schober et al., 2010; Schulz et al.,
Practical ontology
Learn to ask the right questions in building an ontology
2011). Module 7 – Description Logic reasoning
We decided to use the Protégé editor because of its free availa- Using a DL reasoner in the editing cycle in Protégé
design
bility and its support for OWL and automatic reasoners. In view of Module 6 – Introduction in OWL and the Manchester syntax
our guidelines we decided, however, not to use all available featu- Use the restrictions editor in Protégé
res of Protégé. For example, we did not include individuals and data Module 5 – Relations and mereology
type properties, because they are of no relevance for the modeling of Using relations and implementing a partonomy
proper ontological facts and even apt to mislead a novice developer. Module 4 – Disjoints and polyhierarchies
Implementing disjoints and polyhierarchies with Protégé
A small sample ontology (Zoo ontology) was developed as a run-
Module 3 – The ontology editor Protégé
ning example throughout most lectures and exercises. We decided
principles
Implementing an is_a hierarchy with Protégé
to use the BioTop domain top level ontology (Beisswanger et al.,
Basic
Module 2 – Classification and Taxonomy
2008) in the modules concerned with top-level ontologies and cre- Building an is_a hierarchy
ated a reduced version of it (BioTopLite) to restrict the information
load to the essential information needed for the solving of the tasks Module 1 – Introduction in ontology and philosophical background
and learning objectives.
Fig. 1. Structure of the curriculum. Modules are arranged as placed in the
Instructional format curriculum in ascending order, each relying on the previous one.
As the most appropriate instructional format for the training of skills
practical exercises were developed, so that the curriculum structure
was designed to be based on 16 training modules (each lasting 2-3 on their uninterrupted participation, but not on results in any of the
hours) which consisted in a short introductory oral presentation of associated tests and questionnaires.
about 15 min followed by one or more practical tasks. Hands-on The condensed main curriculum took place in the first five days of
exercises were shaped for pair-wise execution, whereas paper-based the summer school. In the remaining three days a quantitative study
practical exercises were adapted to groups of up to six students. on ontology development was performed.
Curriculum evaluation was performed with a questionnaire con-
sisting of 48 closed, and 12 open questions, where the first ones
assessed how students judged the educational principles, the dif- 3 RESULTS
ficulty of the content, and the relevance of each module using a Curriculum structure and contents
five-point Likert scale. Additionally opinions and attitudes with
Figure 1 shows the structure of the curriculum and the main contents
regard to problems and advantages of the course were collected.
of the modules. Sixteen modules with a length between two and
three hours followed each other with increasing complexity. These
Implementation could be grouped in four sections without sharp borders: Basic pri-
The curriculum was implemented in Summer 2011 as an elective nciples, Practical ontology design, Using top-level ontologies, and
summer school at Freiburg University, Germany. 24 students from Using ontology design patterns.
Austria, Germany and Slovenia participated who either studied bio- Especially in the first phase of the curriculum, many exercises
logy as major subject and computer science/ bio-informatics as were conducted in group work without a computer, the aim being to
minor subject or studied computer science/ mathematics as major demonstrate that major parts of the development process consist of
subject and biology as minor subject. The participants agreed in an cognitive decision steps, which are independent of technical skills
informed consent to the scientifically analysis of results of the sum- or software programs.
mer school and to their participation in a subsequent educational
study. Each participant received an expense allowance of e 500 after Module structure
completing the summer school and the study. It was clearly commu- Most modules were organized in a similar structure, consisting of
nicated that the payment of the expense allowance would depend an introductory short presentation (at most 20 min), followed by a
3
Boeker et al
training phase in which the students worked on practical tasks alone Mininum Maximum Mean (SD)
or in pairs. In a plenary session results, suggested solutions and pro- (module) (module) n=24
blems were presented and discussed, and a take-home message was 1.5 2.7
Didactics 2.1 (0.84)
formulated. All material and all exercises were laid out in a brief Taxonomy Immaterial object
document, which focused on most important details without distur- 1.8 3.3
Difficulty 2.5 (0.92)
bing the students’ creativity. Module 10 Process and participation Taxonomy Immaterial object
is presented here as an example. 1.5 2.4
Relevance 1.8 (0.96)
The module starts with a presentation which introduces processes Process, Closure Design errors
as an important ontological category, relating it to the correspon- Table 1. Results of the curriculum evaluation with questionnaire on a
ding BioTop classes and relations: Processes (Process) are things 5-point Likert scale (1 = very good; 5 = very poor). The range as minimum
which have parts in time following each other in a sequence, so that and maximum of average module ratings is given with the corresponding
all process parts which are preceded by the next part in sequence module names. The mean and standard deviation includes the ratings of all
are existentially related to the latter (preceded-by). Processes are 24 students on the complete set of 16 modules.
only fully instantiated when they are finished. They have at least
one participant (expressed by the relation pair hasParticipant, par-
ticipatesIn), and they have a duration (hasDuration). The location
(expressed by the relation pair hasLocus, locusOf) of a process has
in little time and more transfer tasks were demanded from the
to be differentiated from its participant. The process participation
participants.
can be further distinguished, which is expressed by the subrelati-
ons hasAgent, hasPatient, and hasOutcome. A plan (Plan) to do • Even after eight days of hand-on work with Protégé , many
something can only be realized (hasRealization) by processes. students expressed lack of confidence in applying ome of its
In the four exercises of this module, practical skills and imple- functionalities appropriately.
mentation issues had been theoretically introduced beforehand, and • Many of the students were highly motivated by the curricu-
were then issued as practical Protégé editing task. To provide stu- lum and expressed interest allocating further time to ontology
dents with a framework of basic classes and relations, they were engineering.
requested to use a custom-tailored BioTopLite version, that inclu- • Some students wanted to be trained more thoroughly in formal
ded basic constraints during practical ontology building tasks, and and logical backgrounds. Many students complained of “philo-
limiting potential variants to facilitate the subsequent comparison of sophical deviations” during presentations and wanted to have a
the resulting OWL artifacts. more straight forward teaching focusing on clearer statements
In the first exercise of this module students had to provide a taxo- on what had to be done practically.
nomy of animal locomotion: Swimming, Flying, Running, Riding
and Digging. These processes should be represented with at least • In the opinion of a few students the practical usage of ontolo-
one participant and should be correctly localized in a suitable envi- gies in different scenarios should have been shown (and even
ronment, e.g. (PortionOfAir, PortionOfLand, PortionOfWater). The practiced). For them the benefit, meaning and technical stru-
definition of locomotive processes should then be used to define cture of an “ontology driven semantic framework” remained
FlyingAnimal and the other classes described along their locomo- unclear and should have been incorporated in the curriculum.
tion modality. In other exercises of this module the sequence of
processual parts in a developmental process had to be represented
with the precededBy relation or different roles had to be assigned 4 DISCUSSION
to participants in a hunting action (Hunting, Hunter, Prey), so that This work is based on the assumption that better educational pro-
individuals that are members of the same class can have different grams for the training of life sciences ontology engineering are
roles. necessary to improve the quality of the ontological artifacts pro-
duced in this domain. Consequently, we developed a curriculum
Curriculum evaluation which covered the most important aspects of ontologies in the bio-
medical domain according to the literature, based on prior work on
The quantitative evaluation of students on the curriculum was a ontology design guideline and our practical experience as onto-
performed on a 5-point Likert scale (1 = very good; 5 = very poor). logy developers. Moreover, the curriculum was specifically targeted
As shown in table 1, students evaluated the overall didactic princi- to learners with a biomedical background, additionally trained in
ples of the curriculum modules and their relevance positively. The basic and practical computer science. This profile characterizes in
ratings for difficulty were slightly lower. our view the most important stakeholders in biomedical ontology
As anticipated, the qualitative answers of the students showed a building, maintenance, deployment, implementation, and use. The
large variety of partly contradictory opinions and feelings on the curriculum has consequently been developed to serve as a hands-on
curriculum. However, they purport a clear view on the following guide to real-life ontology design problems, and supports the future
aspects of the curriculum: ontology engineer throughout the ontology development life cycle.
Although students were generally very satisfied with the curricu-
• The curriculum started too slowly regarding the amount and lum (as shown in the evaluation), they also signaled some major and
the difficulty of the module content. In the mid-curriculum the minor curriculum improvements. If their assessment on time allo-
pace was evaluated to be just right, but accelerated too steeply cation (in the beginning too long, in the end too short) is correlated
at the end, where the more difficult issues had been addressed with the curriculum structure (see Fig. 1) this could mean that the
4
Teaching Ontology
introductory modules, especially on taxonomy, should be conden- REFERENCES
sed. On the other hand, the modules on complex real life scenarios, Ashburner, M., Ball, C. A., Blake, J. A., Botstein, D., Butler, H., Cherry, J. M., Davis,
beginning in the middle of the curriculum, should have been alloca- A. P., Dolinski, K., Dwight, S. S., Eppig, J. T., Harris, M. A., Hill, D. P., Issel-
ted more time. Some students also wanted to gain deeper insight Tarver, L., Kasarskis, A., Lewis, S., Matese, J. C., Richardson, J. E., Ringwald, M.,
Rubin, G. M., and Sherlock, G. (2000). Gene ontology: tool for the unification of
in formal and logical aspects, which were deliberately presented
biology. Nature genetics, 25(1), 25–29. PMID: 10802651 PMCID: 3037419.
superficially, as we did not consider them important for a practical
Baader, F., Calvanese, D., McGuiness, D., Nardi, D., and Patel-Schneider, P., editors
approach under the given time constraints. In the future this could (2007). The description logic handbook : theory, implementation, and applications.
be addressed by providing additional readings for those interested Cambridge University Press, Cambridge, New York, 2nd ed. edition.
in these topics. As a consequence of the survey, we will modify the Beisswanger, E., Schulz, S., Stenzhorn, H., and Hahn, U. (2008). Biotop: An upper
curriculum by integrating short tutorial sections about the principle domain ontology for the life sciences. a description of its current structure, contents
and interfaces to obo ontologies. Applied Ontology, 3(4), 205–212.
and basic subject matter into modules with more practical impact.
Boeker, M., Stenzhorn, H., Kumpf, K., Bijlenga, P., Schulz, S., and Hanser, S. (2007).
In addition to shifting the focus more towards practical design pro- The @neurist ontology of intracranial aneurysms: Providing terminological services
blems, we consider developing a new module which addresses the for an integrated it infrastructure. In J. M. Teich, J. Suermondt, and G. Hripcsak,
logical and formal foundations of ontology and description logic. editors, American Medical Informatics Association 2007 Proceedings. Biomedical
The modularized approach of the curriculum gives the educator and Health Informatics: From Foundations to Applications to Policy, pages 56–60.
Boeker, M., Tudose, I., Hastings, J., Schober, D., and Schulz, S. (2011). Uninten-
freedom for customization, aligned to either specific requirements ded consequences of existential quantifications in biomedical ontologies. BMC
of certain user groups or to curriculum length. It is relatively easy to Bioinformatics, 12(1), 456.
focus the modules on certain topics or to compress them for students Brochhausen, M., Burgun, A., Ceusters, W., Hasman, A., Leong, T. Y., Musen, M., Oli-
with more pre-existing skills and knowledge. veira, J. L., Peleg, M., Rector, A., and Schulz, S. (2011). Discussion of ”biomedical
ontologies: toward scientific debate”. Methods of Information in Medicine, 50(3),
There are two major sources for further development and impro-
217–236.
vement of the curriculum. Students evaluations as described above Kern, D. E., Thomas, P. A., Howard, D. M., and Bass, E. B. (1998). Curriculum Deve-
and typical mistakes and errors made in the exercises during the lopment for Medical Education: A Six-Step Approach. The John Hopkins University
summer school. The latter were documented in detail and are cur- Press, Baltimore and Maryland.
rently subject of a systematic analysis. Future work will also include Liskov, B. (1987). Keynote address - data abstraction and hierarchy. SIGPLAN Not.,
23(5), 1734.
the design of online modules on the basis of the current curriculum, Mizoguchi, R. and Kozaki (2009). Ontology engineering environments. In S. Staab
which enables the design of flexible courses in a blended lear- and R. Studer, editors, Handbook on Ontologies. International Handbook on
ning approach, i.e. as a combination of electronic with face-to-face Information Systems, pages 315–336. Springer-Verlag, s.l.
teaching. Munn, K. and Smith, B. (2008). Applied ontology: An introduction, volume 9 of
Metaphysical research. Ontos-Verl, Frankfurt.
Rector, A. L., Brandt, S., and Schneider, T. (2011). Getting the foot out of the
5 CONCLUSION pelvis: modeling problems affecting use of SNOMED CT hierarchies in practi-
The dissemination of good ontological practice in the life sciences cal applications. Journal of the American Medical Informatics Association, 18(4),
432–440.
is not only a matter of research, but also of the availability of Schober, D., Boeker, M., Bullenkamp, J., Huszka, C., Depraetere, K., Teodoro, D.,
professional trainers who impart the knowledge in this highly inter- Nadah, N., Choquet, R., Daniel, C., and Schulz, S. (2010). The debugit core onto-
disciplinary area. On this basis, a curriculum of 16 modules was logy: semantic integration of antibiotics resistance patterns. In C. Safran, S. Reti,
developed to train students in the biomedical domain with a back- and H. F. Marin, editors, MEDINFO 2010 - Proceedings of the 13th World Con-
gress on Medical Informatics, volume 160 of Studies in Health Technology and
ground in computer sciences how to practically develop and build
Informatics, pages 1060–1064, Amsterdam. IOS Press.
ontologies in OWL using the Protégé editor. The implementation Schulz, S., Suntisrivaraporn, B., Baader, F., and Boeker, M. (2009). Snomed reaching
of this curriculum in a summer school setting including 24 students its adolescence: ontologists’ and logicians’ health check. International journal of
clearly showed that it is adequate to convey and train the complex medical informatics, 78 Suppl 1, S86–94.
knowledge and skills of ontology engineering in a week long course. Schulz, S., Spackman, K., James, A., Cocos, C., and Boeker, M. (2011). Scalable repre-
sentations of diseases in biomedical ontologies. Journal of Biomedical Semantics,
This curriculum represents an outline, how to successfully train 2(Suppl 2), S6.
students and researchers in the life sciences with the abstract matters Smith, B., Köhler, J., and Kumar, A. (2004). On the application of formal principles to
of ontology engineering, enabling them to formally represent their life science data: a case study in the gene ontology. In E. Rahm, editor, Data Inte-
domain knowledge with maximum profit. gration in the Life Sciences, volume 2994, pages 79–94. Springer Berlin Heidelberg,
Berlin, Heidelberg.
Smith, B., Ashburner, M., Rosse, C., Bard, J., Bug, W., Ceusters, W., Goldberg, L. J.,
ACKNOWLEDGEMENTS Eilbeck, K., Ireland, A., Mungall, C. J., Leontis, N., Rocca-Serra, P., Ruttenberg, A.,
Sansone, S., Scheuermann, R. H., Shah, N., Whetzel, P. L., and Lewis, S. (2007).
This work was supported by the Deutsche Forschungsgemein-
The OBO foundry: coordinated evolution of ontologies to support biomedical data
schaft (DFG) grant JA 1904/2-1, SCHU 2515/1-1 within the integration. Nat Biotech, 25(11), 1251–1255.
project GoodOD (Good Ontology Design), http://www.iph.uni-
rostock.de/Good-Ontology-Design.902.0.html.
5