SODA

When SUC met CLE. Parsing tagged unrestricted text in the Swedish Core Language Engine

Lindberg, Nikolaj and Santamarta, Lena (1994) When SUC met CLE. Parsing tagged unrestricted text in the Swedish Core Language Engine. [SICS Report]

[img]
Preview
PDF
4Mb

Abstract

This paper describes a way of fielding part of speech tagged Swedish text to the syntactic parser of the Swedish Core Language Engine, in order to get automatically produced syntactic analyses of unrestricted written text. The idea is to later on manually disambiguate and correct the output of the parser, or, in other words, to start building a tree-bank - a corpus of syntactically analyzed text. After describing an already existing tree-bank for English and presenting the material used, a detailed account of the process of transforming the tagged text to a formal suitable for the parser is given.

Item Type:SICS Report
Additional Information:This report is also available as a Bachelor of Art Thesis in Computational Linguistics, Dept. of Linguistics, Stockholm University.
Uncontrolled Keywords:Tree-bank, corpora, computational lexica
ID Code:2490
Deposited By:Vicki Carleson
Deposited On:27 Jul 2009
Last Modified:18 Nov 2009 16:09

Repository Staff Only: item control page