=Paper=
{{Paper
|id=None
|storemode=property
|title=OntoVerbal: a Protégé plugin for verbalising ontology classes
|pdfUrl=https://ceur-ws.org/Vol-897/demo_5.pdf
|volume=Vol-897
|dblpUrl=https://dblp.org/rec/conf/icbo/LiangSSR12
}}
==OntoVerbal: a Protégé plugin for verbalising ontology classes==
https://ceur-ws.org/Vol-897/demo_5.pdf
OntoVerbal: a Protégé plugin for verbalising ontology classes
Shao Fen Liang 1∗, Robert Stevens 1 , Donia Scott 2 and Alan Rector 1
1
School of Computer Science, University of Manchester, Oxford Road, Manchester, M13 9PL, UK
2
School of Engineering and Informatics, University of Sussex, Falmer, Brighton, BN1 9QH, UK
ABSTRACT ontology1 that describes the anatomy of a human heart. The
OntoVerbal attempts to reduce the difficulties that non-ontology ontology’s axioms relating to the Valve class are :
experts face in ‘reading’ ontologies, and the burden that ontology (DisjointClasses )
authors face in writing natural language definitions of classes. It does (SubClassOf )
this by verbalising (i.e., automatically generating as natural language) (SubClassOf )
the axioms of OWL classes. Its method relies on presenting, (SubClassOf )
(SubClassOf )
through the use of natural language generation (NLG), naturalistic
(SubClassOf )
descriptions of ontology classes as textual paragraphs. OntoVerbal (SubClassOf )
has been implemented as a Protégé plugin that can offer an (EquivalentTo (and (
alternative ‘English’ view of a class and graphical views provided some ) and (some ))
by various other Protégé plugins. The plugin provides automatic )
RDF label generation for ontology entities and a natural language
description for each class, both for the asserted and ‘inferred’ forms OntoVerbal has structured and ordered these axioms into an
of the class. We have made OntoVerbal, version 1.0, available for English paragraph (Figure 1) according to Rhetorical Structure
Protégé 4.1 via http://swatproject.org/demos.asp. Theory as
A valve is a kind of anatomical concept. More specialised
1 INTRODUCTION kinds of valve are mitral valve, partial valve, semi lunar valve,
tricuspid valve and vestigial cardiac valve. Also, a valve is
Ontology development involves at least two ‘hard’ authoring
different from an anatomical cavity. Another relevant aspect
activities: creating axioms in a new ontology and editing existing
of a valve is that an atrio ventricular valve is defined as a valve
axioms. Thus it is fundamental to using an ontology that the author
that has valve input an atrium cavity and has valve output a
is able to understand its content. As a consequence, managing
ventricular cavity.
ontologies is a highly skilled task that tends to be carried out by
specialists. A richly axiomatised ontology can be hard to read,
either in a native OWL syntax or in some graphical presentation.
Given the growing importance and proliferation of ontologies in
the biomedical and other fields, the lack of ready access to their
content is a major stumbling block to wider use. Also, natural
language descriptions are a desirable feature of ontologies and
mandated by the Open Biomedical Ontologies consortium, and as
these are time-consuming to write, support for their production can
be valuable (Stevens et al. (2011)).
OntoVerbal has been developed to help address these problems
(Liang et al. (2011a)). It has applied methods from linguistics,
psycholinguistics and computational linguistics to achieve its
language generation (Liang et al. (2011b)). In particular,
OntoVerbal deploys axioms of a selected class into a discourse
structure. Thus axioms can be transformed into a set of sentences
and then into a structured and well ordered paragraph that represents
the class. OntoVerbal’s aim is not perfect natural language, but
a generic approach to producing acceptable English for a class’
axioms. Fig. 1. the OntoVerbal description of Valve
http:
2 ONTOVERBAL IN PROTÉGÉ //www.swatproject.org/publications/Valve.jpg
OntoVerbal generates a natural language paragraph for any selected
class. For illustrative examples in this paper, we use the heart 1 retrieved from http://owl.cs.manchester.ac.uk/
repository/download?ontology=http://smi.stanford.
edu/people/dameron/ontology/anatomy/heart\&format=
∗ To whom correspondence should be addressed: fennie.liang@cs.man.ac.uk RDF/XML downloaded April 2012.
1
Liang et al.
The names of the ontology’s classes provide much of the lexical 3 DISCUSSION
content of the generated English, and so if the ontology does not Currently, OntoVerbal uses lightweight linguistic approaches for its
use well formed labels, but only URI fragments, this will have NL paragraph generation. The main reason is that OntoVerbal is
a detrimental effect on OntoVerbal’s verbalisation; for example, intended as a real time application and employing heavy linguistic
instead of reading A valve is a kind of anatomical concept the methods will slow down its performance. Also, since the aim is not
reader will be confronted with A is a kind of to produce perfect English, but rather English that is acceptable
. In the latter case, OntoVerbal will for the purpose of revealing clearly the content of the ontology,
also lose some of its abilities in paragraph generation, such as the output of OntoVerbal will at times include incorrect articles
putting articles in the right places. For this reason, OntoVerbal will, (as seen above) and/or plurals, and be clumsy in places. Since
when necessary, make its own labels from URI fragments. The OntoVerbal is intended to be faithful to its input, in contexts where
natural language generation engine will supply labels for ontology the selected class contains many related axioms, it will sometimes
classes, object properties, data properties and individuals. It breaks produce excessively long paragraphs. Given that our aim is for rapid
entity URI fragments such as CamelCase, Under score or a mixture generation of coherent English text for any class, we feel that these
of both into separate words. compromises are acceptable.
OntoVerbal can also provide descriptions for classes after The OntoVerbal Description tab can generate paragraphs for
reasoning. The description for the Valve class after running a classes without RDF labels, but the text will be of reduced
reasoner becomes: quality compared to those with hand-crafted labels. The OntoVerbal
A valve is a kind of anatomical concept. A more specialised Description tab can also provide more specific descriptions for
kind of valve is partial valve. Also, a valve is different from a classes if a reasoner is used. OntoVerbal will not replace hand-
left atrium cavity, a coronary artery, a vestigial cardiac valve, crafted natural language descriptions, but can provide a substitute in
a conus artery, a right marginal artery, a pulmonary valve, ... their absence. It also provides an alternative view to an ontology’s
an apex of heart, an anterior part of wall of right ventricle, a axioms in a reasonably familiar natural language form that seeks to
valve of coronary sinus, a left circumflex artery and an aorta. ‘ease’ access to often complex ontologies.
Another relevant aspect of a valve is that an atrio ventricular
valve is defined as a valve that has valve input an atrium cavity
and has valve output a ventricular cavity.
After reasoning, much more is known about the class and this
obviously has an effect on the verbalisation. However, a reader
needs verbalisation of both views at different times—as is provided ACKNOWLEDGEMENTS
in tools such as Protégé. The inferred description (figure 2), in fact, This work is part of the Semantic Web Authoring Tool (SWAT)
contains 64 disjoint classes, and some of them are omitted in this project (see www.swatproject.org), which is supported by the UK
paper. The red coloured classes shown in Protégé are unsatisfiable, Engineering and Physical Sciences Research Council (EPSRC)
but OntoVerbal’s descriptions has ignored the red colour and still grant EP/G032459/1, to the University of Manchester, the
generated descriptions using the inferred axioms. University of Sussex and the Open University. We are grateful for
the comments received from our colleagues on the project.
REFERENCES
Liang, S. F., Stevens, R., Scott, D., and Rector, A. (2011a).
Automatic verbalisation of SNOMED classes using ontoverbal.
Proceedings of the 13th Conference on Artificial Intelligence in
Medicine, AIME 2011, pages 338–342.
Liang, S. F., Scott, D., Stevens, R., and Rector, A.
(2011b). Unlocking medical ontologies for non-ontology experts.
Proceedings of the 2011 Workshop on Biomedical Natural
Language Processing, ACL-HLT 2011, pages 174–181.
Stevens, R., Malone, J., Williams, S., Power, R., and Third, A.
Fig. 2. inferred OntoVerbal description of Valve (2011). Automating generation of textual class definitions from
http://www.swatproject.org/publications/ OWL to English. Journal of Biomedical Semantics, 2(Suppl 2),
InferredValve.jpg S5.
2