Czech Word Sketch Relations with Full Syntax Parser
| Authors | |
|---|---|
| Year of publication | 2009 |
| Type | Chapter of a book |
| MU Faculty or unit | |
| Citation | |
| Description | This paper describes the exploitation of dependency relations obtained from syntactic parsing of Czech for building new Czech Word Sketch tables. Standard Word Sketch construction process usually uses so called Sketch grammars -- a simplified process of identifying dependency relations based on regular expressions. This may, of course, lead to errors, which should however not influence (so much) the overall numbers computed on a very big corpus. The paper presents an experiment of using relations resulting from full syntactic parsing -- will they perform better than the standard Sketch grammar or not? |
| Related projects: |