Some specific problems in the annotation of semantic papers. Brief comparative study based on data from AnCorA, SenSem and ADESSE

被引：0

作者：

Vaamonde, Gael ^{[1
]}

机构：

[1] Univ Vigo, Dept Traduc & Linginst, Grp Invest Gramat & Lexico GIGRALEX, E-36200 Vigo, Spain

来源：

PROCESAMIENTO DEL LENGUAJE NATURAL | 2008年 / 41期

关键词：

Semantic roles; corpus annotation; verbal classification; argument structure;

D O I：

暂无

中图分类号：

H0 [语言学];

学科分类号：

030303 ; 0501 ; 050102 ;

摘要：

The labelling of semantic roles has become an important challenge both in the field of corpus linguistics and in the natural language processing. However, it is a hard task in which we have to deal with certain problems of annotation and in which different groups often take different solutions, regardless of the theoretical framework which supports the analysis. This paper outlines some of these problems and simultaneously compares the different solutions adopted by three research projects that have dealt with the syntactic-semantic analysis of a Spanish corpus.

引用

页码：233 / 240

页数：8