If two translation systems differ differ in performance on a test set, can we trust that this indicates a difference in true system quality? To answer this question, we describe bootstrap resampling methods to compute statistical significance of test results, and validate them on the concrete example of the BLEU score. Even for small test sizes of only 300 sentences, our methods may give us assurances that test result differences are real.
|Title of host publication||Proceedings of EMNLP 2004|
|Editors||Dekang Lin, Dekai Wu|
|Place of Publication||Barcelona, Spain|
|Publisher||Association for Computational Linguistics|
|Number of pages||8|
|Publication status||Published - 1 Jul 2004|