=Paper=
{{Paper
|id=Vol-551/paper-32
|storemode=property
|title=Matching natural language data on ontologies
|pdfUrl=https://ceur-ws.org/Vol-551/om2009_poster11.pdf
|volume=Vol-551
|dblpUrl=https://dblp.org/rec/conf/semweb/Heinecke08
}}
==Matching natural language data on ontologies==
https://ceur-ws.org/Vol-551/om2009_poster11.pdf
Matching Natural Language Data with Ontologies
Johannes Heinecke
Orange Labs, F-22307 Lannion cedex – France
johannes.heinecke@orange-ftgroup.com
Abstract. Ontologies and natural languages are complementary. Whereas ontologies
are used to model knowledge formally, natural language is primarily used by users to
communicate with ontology based systems. In order to transform information or queries
in natural language into valid ontological expressions, the meaning of natural language
entities have to be matched with the given ontologies. In contrast to pure ontology
matching, the matching with natural language data poses some problems linked to their
ambiguities (synonymy, homonymy/polysemy, redundancy, to name but a few).
1 Introduction, context and related work
In the context of the Semantic Web, the interfacing of ontological representation and natural
language is an important issue. Since much information on the Web exists (only) in tex-
tual form, the usage of this information in ontology based tools is not possible unless these
texts are made accessible or comprehensible by such tools. This means that texts and user
queries have to be “translated” into an ontological representation language such as the W3C
languages RDF/RDFS and OWL.
The need for the work described here came from the aceMedia project (http:// www.
acemedia.org/ ) [1,2] (cf. also [3]). In this project, the two tasks are
transforming textual annotations of multimedia contents into an ontological representa-
tion (based on an existing ontology) in order to make them available for a knowledge-base;
and translating English and French user queries into an ontological query language (in our
case SPARQL). The matching of linguistic data (lexicons, thesauri) with ontologies is similar
but not identical to ontology matching or ontology alignment, i.e. trying to find correspond-
ing classes of ontology A in ontology B [4]. Different methods of matching are discussed
in detail by [5, p. 65]. Following this classification the present approach can be considered
being terminological and linguistic, since we use relationships found in the lexicon (via a
semantic thesaurus, [6]) and the taxonomic hierarchies of both, the lexical semantic data and
the ontologies notably for the disambiguation of polysemous words. Similar work describe
[7] and [8]. In contrast to their results we do not have a classification at hand.
2 Linguistic-ontology matching
Apart from the ontologies, the matching requires a complete lexicon of the language used to
label or describe the ontological classes and properties (= entities). Our lexicon is lso linked
to a semantic thesaurus. The ontologies, on the other hand, usually have non-ambiguous
entity labels (like http: //www.acemedia.org/ontos/tennis#Player1 ) or a comment, explain-
ing the entity. This is especially necessary if the entity labels are not self-explanatory like
1 We shorten name spaces like http://www.acemedia.org/ontos/tennis# to “tennis:” etc.
tennis:C12 (a fortunately rare case). Further, the semantic thesaurus contains a thematic hi-
erarchy of all semantic concepts to help disambiguation. These are grouped into 880 themes
which in turn are organized in 80 domains. Domains are divided into about 10 macro-do-
mains. The matching itself comprises several
lexicon NLP
semantic
(b) Semantic analysis of (f) Creation of semantic
steps (cf. fig. 1). Apart from a (more or less
classes (taking into clustering rules for
thesaurus account their taxonomic
context)
complex classes manual) preparation in order to correct possi-
(e) adding ontologies to
semantic thesaurus (c) Semantic analysis of (g) Creation of RDFS
ble labeling errors in the ontologies, the other
properties (taking into rules (predicateclass
domain account their range and
domain)
transformation) steps do not need any intervention: (a) extract-
ontologies
ing the “ontological context” of entities and as-
(a) extract and correct (d) Detecting synonyms
class and property labels and hyponyms signing eventual reformulations of entity labels;
(b) natural language processing passes: detect-
Fig. 1. linguistic-ontological matching
ing meanings for classes using their ontological
context (direct sub-classes); (c) and for properties using their ontological context (domain
and range classes); (d) determining the application depending synonyms and co-hyponyms;
(e) adding the ontological hierarchy to the semantic taxonomy; (f) creating semantic trans-
formation rules for “complex class labels”2 ; (g) creating transformation rules for the creation
of ontological representation (from semantic graphs. Synonyms (defined in our multilingual
thesaurus, [9]) are all matched onto the same ontological class (e.g. “river”, “stream”, “creek”
etc. → holidays:River). If a class has no sub-classes, we also match the co-hyponyms of the
label to the class (e.g. in our case “car”, “bus”, “truck”, “motorbike” . . . → general:Vehicle.
The resulting linguistic data is successfully used the aceMedia prototype, similarly produced
data is used in an industrial application to create and access ontological based information
from/via natural language.
New perspectives are offered by structured semantic data which is getting more and more
available. Databases like Wikipedia (especially the categorization schema used within) or
RDF or ontology based information systems like DBpedia or freebase3 (both initialized by
Wikipedia contents) will help to improve the linking of natural languages and formally mod-
eled ontologies.
References
1. Heinecke, J.: Génération automatique des représentations ontologiques. In: TALN. (2006) 502–511
2. Dasiopoulou, S., Heinecke, J., Saathoff, C., Strintzis, M.G.: Multimedia reasoning with natural
language support. In: IEEE-ICSC. (2007) 413–420
3. Heinecke, J., Toumani, F.: A Natural Language Mediation System for E-Commerce applications.
In: Workshop HLT for the Semantic Web and Web Services. ISWC. (2003) 39–50
4. Ehrig, M., Staab, S.: QOM - quick ontology mapping. In: ISWC. (2004) 683–697
5. Euzenat, J., Shvaiko, P.: Ontology Matching. Springer, Heidelberg (2007)
6. Heinecke, J., Smits, G., Chardenon, C., Guimier De Neef, E., Maillebuau, E., Boualem, M.: TiLT :
plate-forme pour le traitement automatique des langues naturelles. TAL 49:2 (2008)
7. Giunchiglia, F., Marchese, M., Zaihrayeu, I.: Encoding classifications into lightweight ontologies.
Journal of Data Semantics VIII (2007) 57–81
8. Reiter, N., Hartung, M., Frank, A.: A resource-poor approach for linking ontology classes to
Wikipedia articles. In: STEP. (2008) 381–387
9. Chagnoux, M., Heinecke, J.: Aligner ontologies et langues naturelles. gérer la synonymie. In:
Plateforme AFIA, Grenoble (2007) 87–94
2 Labels which use multi-word expressions like tennis:ExhibitionMatch instead of simple words.
3 http:// dbpedia.org/ , http:// freebase.com/