Skip to Main content Skip to Navigation
Conference papers

Analyse et classification d'exemples illustratifs dans le dictionnaire "Cesselin" en utilisant Google Traduction et un dictionnaire UNL-UWs

Abstract : Knowing whether an example in a bilingual dictionary is a proverb, an idiomatic expression, a collocation, or a simple usage example would be quite helpful to users, but is rarely indicated. In this paper, we present a manual classification and annotation of examples in a Japanese-French bilingual dictionary, the "Cesselin" (145000 entries), derived from the manual analysis of 500 entries, and outline steps towards automating the annotation process. As for the proverbs, we simply use proverb dictionaries. For the other cases, our method relies on comparing the (automatic) translation into English of an original (Japanese) example and its (French) translation in the dictionary. The comparison is based first on the number of common lexemes in the two (English) outputs, and second on the number of synonymous lexemes, which can be found using a large UNL-UW dictionary. We show our classification procedure in detail and its future automatization.
Document type :
Conference papers
Complete list of metadatas

Cited literature [25 references]  Display  Hide  Download

https://hal.archives-ouvertes.fr/hal-01992882
Contributor : Mathieu Mangeot <>
Submitted on : Thursday, January 24, 2019 - 4:16:43 PM
Last modification on : Friday, July 17, 2020 - 11:10:27 AM

File

LTT2018_MTCBMM.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01992882, version 1

Collections

Citation

Mutsuko Tomokiyo, Mathieu Mangeot-Nagata, Christian Boitet. Analyse et classification d'exemples illustratifs dans le dictionnaire "Cesselin" en utilisant Google Traduction et un dictionnaire UNL-UWs. 11es journées du réseau Lexicologie Terminologie Traduction 2018, Sep 2018, Grenoble, France. ⟨hal-01992882⟩

Share

Metrics

Record views

123

Files downloads

60