Konferenzbeitrag
YAWN: A Semantically Annotated Wikipedia XML Corpus
Lade...
Volltext URI
Dokumententyp
Text/Conference Paper
Dateien
Zusatzinformation
Datum
2007
Autor:innen
Zeitschriftentitel
ISSN der Zeitschrift
Bandtitel
Verlag
Gesellschaft für Informatik e. V.
Zusammenfassung
The paper presents YAWN, a system to convert the well-known and widely used Wikipedia collection into an XML corpus with semantically rich, self-explaining tags. We introduce algorithms to annotate pages and links with concepts from the WordNet thesaurus. This annotation process exploits categorical information in Wikipedia, which is a high-quality, manually assigned source of information, extracts additional information from lists, and utilizes the invocations of templates with named parameters. We give examples how such annotations can be exploited for high-precision queries.