[Zurück]


Buchbeiträge:

G. Haneder, A. Rauber, M. Dittenbach, E. Schweighofer:
"Improvement of vector representations of legal documents with legal ontologies";
in: "Proceedings of the 5th International Conference on Business Information Systems (BIS 2002)", Springer, 2002, ISBN: 8391684202.



Kurzfassung englisch:
In previous research, a TFxIDF vector representation of legal documents
was the basis for structuring and labelling a collection of legal
documents.
Extensive experiments with three text corpora of European law of about
580 documents in three languages have shown, that binary or weighted
vector representation may not be sufficient. Even quite successful
approaches of similarity computation have problems in identifying the
best context of classification.
This paper deals with a new approach. The vector representation is
modified
using a simple ontology of the domain. The adaptation of the vector is
done using the newly developed Data Enrichment feature of the SOM/KONTERM
library. This approach was evaluated with a small text corpus of
253 documents of cyberspace law. The improvement was significant in the
quality of the clusters but also the labels.


Erstellt aus der Publikationsdatenbank der Technischen Universität Wien.