Homonymic adverbs and prepositions and the problematics of their disambiguation in the Czech national corpus
Abstract
The article deals with homonymic adverbs and prepositions such as blízko, dovnitř, vstříc, etc. There is a list of them in the article, besides the distribution of these homonyms in the corpus SYN2010 and success rate of their automatic disambiguation in this corpus are analysed - there is also classification and analysis of the mistakes in the disambiguation of these units.
It is concluded that problems of the automatic disambiguation of homonymic adverbs and prepositions arise on the background of wide grammatical homonymy in the Czech language and are mainly complicated by homonymy of case forms of nouns. This research is considered to be the basis for the follow-on development of particular linguistics rules, necessary for the improvement of automatic disambiguation of homonymic adverbs and prepositions in the Czech National Corpus.