Automatic Evaluation of Alignments without using a Gold-Corpus - Archive ouverte HAL Access content directly
Conference Papers Year : 2018

Automatic Evaluation of Alignments without using a Gold-Corpus

Raoul Blin


This paper presents a test designed to automatically evaluate the alignment quality of a bilingual aligned corpus without comparing it to a reference/gold corpus. We justify bypassing a reference corpus and present the advantages of the test, particularly for language pairs where few bilingual aligned corpora exist. The test, which is based on the observation of single-translation words, is demonstrated using three bilingual French-Japanese corpora : a manually translated and aligned corpus, a freely translated and semi-manually aligned corpus, and a freely translated and automatically aligned corpus. This paper shows that the results achieved validate the proposed evaluation technique.
Fichier principal
Vignette du fichier
23_W29.pdf (167.11 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-01789350 , version 1 (10-05-2018)


  • HAL Id : hal-01789350 , version 1


Raoul Blin. Automatic Evaluation of Alignments without using a Gold-Corpus: Example with French-Japanese Aligned Corpora. Conference on Language Resources and Evaluation (LREC 2018), European Language Resources Association (ELRA), May 2018, Miyazaki, Japan. ⟨hal-01789350⟩
213 View
124 Download


Gmail Facebook Twitter LinkedIn More