Að fanga orðaforðann: orðanet í þágu orðabókar

Abstract

The paper describes an ongoing lexicographic project, Íslenskt orðanet (Icelandic wordnet), which aims to analyse and describe the Icelandic vocabulary and its internal semantic relations on the basis of syntactic and morphological relations in word combinations and compounds. A comprehensive description of Icelandic phrasemes, which has been published in three phraseological dictionaries, together with a big collection of newspaper texts, provides the main material for the lexicographic analysis. Disambiguation and lemmatisation af multi-word lexical units plays a central role in the construction of the wordnet. The multi-word lemmas are grammatically tagged, which enables grammatical sorting and an interaction between semantic and grammatical features. The wordnet combines the functions of a synonym and a concept dictionary, paying special attention to measuring semantic relatedness between synonyms and near-synonyms on the basis of paratactic word-pairs. The product of the analysis is open and accessible as an online dictionary on the website www.ordanet.is.