Contents and evaluation of the first Slovenian-German online dictionary

Contents and evaluation of the first Slovenian-German online dictionary Birte Liinneker Institute for Romance Languages Hamburg University Von-Melle-Park 6 20146 Hamburg, Germany birte.loenneker@uni -hamburg.de Prima Jakopin Corpus Laboratory F. R. Institute of Slovenian language ZRC SAZU, Gosposka ulica 13 1000 Ljubljana, Slovenia primoz.jakopin@uni—lj.si Abstract This paper presents the first Slovenian- German and German-Slovenian online dictionary and contains evaluation fig- ures for its Slovenian part. Evaluations are based on coverage of a Slovenian newspaper corpus as well as on user queries. 1 Introduction The first Slovenian-German and German- Slovenian online dictionary is available at http ://www. stud.uni -h amburg. de/users/1.)i rte/slo Its current version, which was completed in November 2002, contains more than 4,800 entries covering the content of a beginners' textbook for Slovenian. Section 2 gives some information about the con- tents and structure of the dictionary. Section 3 evaluates the Slovenian part, based on its coverage of lemmas in DELO, a Slovenian newspaper con- tained in the Nova beseda corpus at ZRC SAZU (Beseda, 2000) as well as on its ability to fulfill user requests. Section 4 is the conclusion. 2 Contents and structure of the dictionary In November 2002, the Slovenian-German- Slovenian online dictionary contained more than 4,800 entries which correspond to words, selected 'http://bos.zrc-sazu.si/a_beseda.html word forms and expressions appearing in the text- book Odkrivajmo slovenkino, a beginners' text- book for Slovenian (C' uk et al., 1996). The entire content of the textbook is covered. The textbook is used in teaching Slovenian es- pecially in German, French or Italian speaking communities. 2 It contains Slovenian-only text, ex- planations and exercises. As there is neither an in- dex nor a vocabulary list, the online dictionary is a valuable completion of this educational material. However, it can be — and actually is — used inde- pendently of the textbook. The current version of the online dictionary contains the following elements, based on Odkri- vajmo slovenkino: • all lemmas appearing in the texts, explana- tions, instructions and exercises; • irregular inflected forms as well as the first person singular form of verbs; • common conversational phrases and multi- word expressions as well as some contextual examples of words and grammatical forms. Grammar information for both languages and information on stressed syllables for the Slovenian entries are contained in additional fields. For both languages, three different kinds of search are pos- sible: 2Personal communication from Meta Lokar, Centre for Slovenian as a Second/Foreign Language, University of Ljubljana. 119 1. exact match3 ; 2. match a text string as a part of the dictionary entry; words� distinct word forms�60,843,505 distinct lemmas or lemma sets�25,598 10,250 Table 1: Lemmatized word list for evaluation. 3. match a text string at the beginning of the dic- tionary entry. In the current version, the internal structure of the dictionary is a table containing one-to-one cor- respondences of words, word forms, and phrases. If an item has more than one equivalent in the other language, as many entries as necessary are created. 3 Evaluation The evaluations of the …

Content truncated to 500 words