Titel: What is needed for automatic production of simple and complex dictionary entries in the first Slovene online dictionary of abbreviations using Termania website
Personen:Kompara, Mojca/Holozan, Peter
Jahr: 2011
Typ: Aufsatz
Verlag: Trojina, Institute for Applied Slovene Studies/ Lexical Computing Ltd.
Ortsangabe: Ljubljana/ Brighton
In: Kosem, Iztok/Kosem, Karmen (Hgg.): Electronic lexicography in the 21st Century: New Applications for New Users. Proceedings of eLex2011, Bled, Slowenien, 10 - 12 November 2011
Seiten: 140-146
Untersuchte Sprachen: Slowenisch*Slovenian
Schlagwörter: Internet-Lexikografie/Online-Lexikografie*internet lexicography/online lexicography
lexikografischer Prozess*lexicographic process
Redaktionssystem*lexicographic editor
URI: http://elex2011.trojina.si/Vsebine/proceedings.html
Zuletzt besucht: 10.09.2018
Abstract: The paper presents what is needed for automatic production of simple and complex dictionary entries in the first Slovene online dictionary of abbreviations using Termania web site. At a first step an algorithm for the automatic recognition of abbreviation-expansion pairs in newspaper texts has been used. After manual cleaning genuine pairs obtained with the algorithm were automatically included in Termania editing software. The editing process was automatised, non-nominal expansions were converted to nominal and language qualifiers were added automatically. The precision of the algorithm for recognition of abbreviation-expansion pairs is 96%, the algorithm for lemmatisation is successful in 70.9% of cases, the algorithm for adding language qualifiers in 95.9% of cases. Out of the obtained results we can say that simple entries are produced automatically, whereas complex "semi" automatically, because translations or explanations of abbreviation-expansion pairs can be added, at the present stage, just manually. The same goes for encyclopaedic data. For that reason we can say that the first Slovene online dictionary of abbreviations is produced "semi" automatically but further attempts to automatize translations and encyclopaedic data will be done in the future.