=Paper=
{{Paper
|id=None
|storemode=property
|title=Extending Tagging Ontologies with Domain Specific Knowledge
|pdfUrl=https://ceur-ws.org/Vol-1272/paper_14.pdf
|volume=Vol-1272
|dblpUrl=https://dblp.org/rec/conf/semweb/FontOFS14
}}
==Extending Tagging Ontologies with Domain Specific Knowledge==
https://ceur-ws.org/Vol-1272/paper_14.pdf
Extending Tagging Ontologies with
Domain Specific Knowledge
Frederic Font1 , Sergio Oramas1 , György Fazekas2 , and Xavier Serra1
1
Music Technology Group, Universitat Pompeu Fabra, Barcelona, Spain
2
Centre for Digital Music, Queen Mary University of London, London, UK
{name.surname}@upf.edu, gyorgy.fazekas@eecs.qmul.ac.uk
Abstract. Currently proposed tagging ontologies are mostly focused on
the definition of a common schema for representing the agents involved in
a tagging process. In this paper we describe preliminary research around
the idea of extending tagging ontologies by incorporating some domain
specific class definitions and relations. We illustrate our idea with a par-
ticular use case where a tag recommendation system is driven by such
an ontology. Besides our use case, we believe that such extended tagging
ontologies can bring more meaningful structure into folksonomies and
improve browsing and organisation functionalities of online platforms
relying on tagging systems.
Keywords: Tagging ontology, Tag recommendation, Folksonomy, Freesound
1 Introduction
Tagging systems are extensively used in online sharing sites as a means of content
browsing and organisation. In general, tagging systems allow users to annotate
resources with free-form textual labels chosen by the users of the system. The re-
sulting set of associations between tags, users and resources that arise in tagging
systems is known as a folksonomy. Folksonomies suffer from a number of well-
known issues including tag scarcity, ambiguities with synonymy and polysemy,
typographical errors, the use of user-specific naming conventions, or even the use
of different languages [1]. Despite these issues, folksonomies have succeeded in
providing basic organisation and browsing functionalities to online sharing sites.
However, their unstructured nature makes it difficult to allow more advanced
capabilities such as hierarchical browsing or faceted searching.
In order to bring some structure to folksonomies, some studies have focused
on the analysis of folksonomies to automatically derive structured or semi-
structured representations of the knowledge of the domain, typically in the form
of lightweight ontologies or hierarchical taxonomies [2–4]. However, these meth-
ods still tend to require significant amount of manual effort to provide meaningful
representations. Some other studies have proposed modelling folksonomies and
the tagging process using ontologies [5]. These ontologies are focused on defining
a common schema for the agents involved in a tagging process. Current tagging
ontologies may enhance interoperability between folksonomies, but do not gen-
erally provide ways of structuring a folksonomy with domain-specific knowledge.
2 F. Font et al.
In this paper, we present some preliminary research on extending a tagging
ontology by including the possibility to represent the semantics of a specific
domain. The generic idea is presented and discussed in Sec. 2. In Sec. 3 we
describe a practical application in a real-world tagging system where the tagging
ontology is used to drive a tag recommendation system. Finally, in Sec. 4, we
discuss about possible future directions.
2 Extending a tagging ontology
Our starting point for the extension of the tagging ontology is the Modular
Unified Tagging Ontology (MUTO) [5]. In the core of the MUTO ontology, the
muto:Tagging class is defined which supports several relations to indicate, among
others, a resource that is tagged (muto:hasResource of type rdfs:Resource), the
tag assigned to the resource (muto:hasTag of type muto:Tag), and the user that
made the tag assignment (muto:hasCreator of type sioc:UserAccount).
We propose to extend the tagging ontology in two ways. First, we add a num-
ber of subclasses to the muto:Tag class which can be used instead of muto:Tag
(right side of Fig. 1). These subclasses represent different tag categories (i.e. with
a narrower scope than the generic muto:Tag class), similarly to the idea of TagSet
introduced in the SCOT ontology [6], but in a semantic sense. A tag category
represents a broad concept that groups a set of tags that share some seman-
tic characteristics related to the specific domain. The same principle is applied
to resources, and a number of rdfs:Resource subclasses are defined (left side
of Fig. 1). Resource subclasses (or resource categories) are used to organise re-
sources into groups with a narrower scope than the general rdfs:Resource class.
The particular definition of tag and resource categories would depend on the
particular application domain of the extended tagging ontology (an example is
given below). Also, in the diagram of Fig. 1, both tag and resource subclasses
are only shown as a flat hierarchy, but more complex class structures could be
explored. Moreover, existing domain ontologies and taxonomies may be reused
to extend the tagging ontology.
Second, we propose to extend the tagging ontology by adding object prop-
erties to model semantic relations among tag categories and resource categories
(dashed lines in Fig. 1). These object properties are useful to, for example, model
dependencies between categories of tags and resources. The specific meaning of
these semantic relations would also depend on the particular application domain
of the extended tagging ontology. In addition to semantic relations between tag
and resource categories, and given that the muto:Tag class inherits from the Sim-
ple Knowledge Organization System (SKOS) [7] class skos:Concept, semantic
relations between tag individuals can be also modelled [5].
3 Use case: tag recommendation in Freesound
We applied an extended tagging ontology as described above in a tag recommen-
dation task in the context of Freesound, an online collaborative sound database
with more than 200,000 uploaded sounds and 3,8 million registered users [8]. In
Extending Tagging Ontologies with Domain Specific Knowledge 3
rdfs:Resource muto:hasResource
muto:Tagging muto:hasTag
muto:Tag
rdfs:subClassOf rdfs:subClassOf
:ResCategory1 :TagCategory1
:ResCategory2 :TagCategory2
... ...
:ResCategoryN :TagCategoryN
Fig. 1. Diagram of the extended parts of the tagging ontology.
previous work by the authors, a tag recommendation system was proposed which,
given a set of input tags, is able to suggest other potentially relevant tags [9].
The system is based on the construction of five tag-tag similarity matrices tai-
lored to five manually defined and rather generic audio categories (e.g. “Music”,
“Effects”, etc.). The recommendation system uses a classifier to automatically
predict one of these five categories depending on the input tags, and then uses
the corresponding tag-tag similarity matrix for the recommendation process.
To improve that recommendation system, we used the extended tagging on-
tology to model the folksonomy and include some domain specific knowledge. On
the one side, we extended the tagging ontology by adding 5 resource subclasses
corresponding to the 5 sound categories mentioned above (e.g. :EffectsSound).
Moreover, we defined 26 tag subclasses that are intended to group the tags in
categories according to the type of information that they describe about sounds
(i.e. grouped in audio properties). These include categories like “instrument”,
“microphone”,“chord”, “material”, or “action” (e.g. :InstrumentTag). On the
other side, we extended the ontology by defining several object properties that
relate resource and tag categories. These object properties indicate that a par-
ticular tag category is relevant for one or more resource categories. For example,
:InstrumentTag is relevant for :MusicSound audio category, and this is indicated
with a :hasInstrument object property that relates instrument tag category with
music resource category. Furthermore, we populated the extended ontology by
manually classifying the 500 most used tags in Freesound into one of the 26
defined tag categories and added these tags as individuals (instances) of the
corresponding tag category. This last step was necessary to bootstrap the tag
recommendation system (see below).
Using this ontology we can extend the tag recommendation system in a way
that, given the audio category detected by the classifier and the object properties
that relate resource and tag categories, we can guide the annotation process by
suggesting tag categories that are relevant for a particular sound. For example,
for a sound belonging to the resource category :MusicSound, we can suggest tag
categories like :InstrumentTag or :TempoTag, which are particularly relevant for
musical sounds. Once tag categories are suggested, users can click on them and
get a list of tag recommendations for every category. This list is obtained by
computing the intersection of the tags provided by the aforementioned recom-
mendation system (based on the tag-tag similarity matrix), with those that have
been manually introduced in the ontology as tag instances of the selected tag
category. See Fig. 2 for an screenshot of a prototype interface for that system.
4 F. Font et al.
Fig. 2. Screenshot of the interface of a prototype tag recommendation system driven
by the extended tagging ontology.
4 Conclusions
In this paper we have shown some preliminary research on extending current
tagging ontologies with structured knowledge specific to the domain of applica-
tion of a tagging system. By incorporating domain specific knowledge in tagging
ontologies, we expect to be able to bring some semantically-meaningful structure
into folksonomies. We have illustrated the idea with a use case in the context
of an audio clip sharing site where a tag recommendation system is driven by
an extended tagging ontology. Formal evaluation of the ontology-driven tag rec-
ommendation system is planned for future work. Besides the described use case,
we think that using extended tagging ontologies can improve other aspects of
online platforms relying on tagging systems such as browsing and organisation
functionalities. The main limitation for such improvements is the population of
the ontology. In our use case, we use a manually populated ontology to bootstrap
the recommender, but the tagging system could further populate the ontology
by learning new “tag individuals-tag category” relations when users annotate
new sounds. Furthermore, other knowledge extraction techniques could be used
to automatically populate the ontology with information coming from other
user-generated data (e.g. in our case could be sound comments or textual de-
scriptions), and even from external data sources from linked open data.
References
1. H. Halpin, V. Robu, and H. Shepard, “The dynamics and semantics of collaborative tagging,”
in Proceedings of the 1st Semantic Authoring and Annotation Workshop, pp. 1–21, 2006.
2. P. Mika, “Ontologies are us: A unified model of social networks and semantics,” Web Semantics:
Science, Services and Agents on the World Wide Web, vol. 5, pp. 5–15, Mar. 2007.
3. P. Heymann and H. Garcia-Molina, “Collaborative Creation of Communal Hierarchical Tax-
onomies in Social Tagging Systems,” tech. rep., 2006.
4. F. Limpens, F. L. Gandon, and M. Buffa, “Linking folksonomies and ontologies for supporting
knowledge sharing: a state of the art,” 2009.
5. S. Lohmann, P. Dı́az, and I. Aedo, “MUTO: the modular unified tagging ontology,” Proceedings
of the 7th International Conference on Semantic Systems - I-Semantics ’11, pp. 95–104, 2011.
6. H. L. Kim, S. Scerri, J. G. Breslin, S. Decker, and H. G. Kim, “The State of the Art in Tag
Ontologies : A Semantic Model for Tagging and Folksonomies,” pp. 128–137, 2008.
7. SKOS: Simple Knowledge Organization System. http://www.w3.org/TR/skos-reference.
8. F. Font, G. Roma, and X. Serra, “Freesound Technical Demo,” in Proceedings of the 21st ACM
Conference on Multimedia (ACM MM 13), pp. 411–412, 2013.
9. F. Font, J. Serrà, and X. Serra, “Class-based tag recommendation and user-based eval-
uation in online audio clip sharing,” Journal on Knowledge Based Systems, 2014,
10.1016/j.knosys.2014.06.003.