Overview of the Third BUCC Shared Task: Spotting Parallel Sentences in Comparable Corpora
Résumé
This paper presents the BUCC 2018 shared task on parallel sentence extraction from comparable corpora. This task used the same data as the BUCC 2017 shared task. 17 runs were submitted by 3 teams, covering all four proposed language pairs: German-English (3 runs), French-English (6 runs), Russian-English (3 runs), and Chinese-English (5 runs). The best F-scores as measured against the gold standard were 0.86 (German-English), 0.81 (French-English and Russian-English), and 0.77 (Chinese-English). All top scores improved over those of 2017.