A Methodology for the Comparison of Human Judgments With Metrics for Coreference Resolution

Archive ouverte

Borovikova, Mariya | Grobol, Loïc | Halftermeyer, Anaïs | Billot, Sylvie

Edité par CCSD ; Association for Computational Linguistics -

International audience. We propose a method for investigating the interpretability of metrics used for the coreference resolution task through comparisons with human judgments. We provide a corpus with annotations of different error types and human evaluations of their gravity. Our preliminary analysis shows that metrics considerably overlook several error types and overlook errors in general in comparison to humans. This study is conducted on French texts, but the methodology should be language-independent.

Suggestions

Du même auteur

TaeC: A manually annotated text dataset for trait and phenotype extraction and entity linking in wheat breeding literature

Archive ouverte | Nédellec, Claire | CCSD

International audience. Wheat varieties show a large diversity of traits and phenotypes. Linking them to genetic variability is essential for shorter and more efficient wheat breeding programs. A growing number of p...

Annoter la parole spontanée en arbres de constituants pour les besoins de l’analyse temporelle : résultats et comparaison français parlé / français écrit

Archive ouverte | Wang, Ilaine | CCSD

International audience. This paper presents the main results drawn from the syntactic part of Temporal@ODIL, a project whose objective is the construction of a temporally annotated corpus of spontaneous speech for F...

ODIL Syntax : a Free Spontaneous Spoken French Treebank Annotated with Constituent Trees

Archive ouverte | Wang, Ilaine | CCSD

International audience. This paper describes ODIL Syntax, a French treebank built on spontaneous speech transcripts. The syntactic structure of every speech turn is represented by constituent trees, through a proced...

Chargement des enrichissements...