| Titel: |
Query-driven Dictionary Enhancement |
| Personen: | Jakopin, Primož/Lönneker, Birte |
| Jahr: |
2004 |
| Typ: |
Aufsatz |
| Verlag: |
Université de Bretagne Sud |
| Ortsangabe: |
Lorient |
| In: |
Williams, Geoffrey/Vessier, Sandra (Hgg.): Proceedings of the 11th EURALEX International Congress, Lorient, France, 6 - 10 July 2004 |
| Seiten: |
273-284 |
| Untersuchte Sprachen: |
Deutsch*German - Slowenisch*Slovenian |
| Schlagwörter: |
Benutzungsforschung*usage research
Internet-Lexikografie/Online-Lexikografie*internet lexicography/online lexicography
|
| URI: |
http://euralex.org/category/publications/euralex-2004/ |
| Zuletzt besucht: |
03.09.2018 |
| Abstract: |
With the log files of online dictionaries, in which all submitted user queries are stored, dictionary authors have for the first time in the history of dictionary building direct access to the users' requests. In this article, we show 1) how to use the log file to evaluate the current contents of an online dictionary and 2) how to choose the most promising corpus type for enlarging it according to the users' needs. We use the example of a German-Slovenian online dictionary for this work. As a result of the first evaluation, we detect that the dictionary does not fulfil the users' needs in coverage of colloquial and vulgar language, as well as in words and expressions used in everyday life. The result of the second evaluation confirms the importance of this part of the vocabulary. In an overall comparison of queries and corpora, a fiction corpus, which by its nature contains also colloquial language, yields a better result than a newspaper corpus and a non-fiction corpus. |