We study a problem with pairwise ranking optimization (PRO): that it tends to yield too short translations. We find that this is partially due to the inadequate smoothing in PRO’s BLEU+1, which boosts the precision component of BLEU but leaves the brevity penalty unchanged, thus destroying the balance between the two, compared to BLEU. It is also partially due to PRO optimizing for a sentence-level score without a global view on the overall length, which introducing a bias towards short translations; we show that letting PRO optimize a corpus-level BLEU yields a perfect length. Finally, we find some residual bias due to the interaction of PRO with BLEU+1: such a bias does not exist for a version of MIRA with sentence-level BLEU+1. We propose several ways to fix the length problem of PRO, including smoothing the brevity penalty, scaling the effective reference length, grounding the precision component, and unclipping the brevity penalty, which yield sizable improvements in test BLEU on two Arabic-English datasets: IWSLT (+0.65) and NIST (+0.37).

@inproceedings{nakov-guzman-vogel:2012:PAPERS, author = {Nakov, Preslav and Guzm{\'a}n, Francisco and Vogel, Stephan}, title = {Optimizing for Sentence-Level {BLEU}+1 Yields Short Translations}, booktitle = {Proceedings of the 24rd International Conference on Computational Linguistics (COLING) 2012 }, month = {December}, year = {2012}, address = {Mumbai, India}, publisher = {The {COLING} 2012 Organizing Committee}, pages = {1979--1994}, url = {http://www.aclweb.org/anthology/C12-1121}, abstract = {We study a problem with pairwise ranking optimization (PRO): that it tends to yield too short translations. We find that this is partially due to the inadequate smoothing in PRO’s BLEU+1, which boosts the precision component of BLEU but leaves the brevity penalty unchanged, thus destroying the balance between the two, compared to BLEU. It is also partially due to PRO optimizing for a sentence-level score without a global view on the overall length, which introducing a bias towards short translations; we show that letting PRO optimize a corpus-level BLEU yields a perfect length. Finally, we find some residual bias due to the interaction of PRO with BLEU+1: such a bias does not exist for a version of MIRA with sentence-level BLEU+1. We propose several ways to fix the length problem of PRO, including smoothing the brevity penalty, scaling the effective reference length, grounding the precision component, and unclipping the brevity penalty, which yield sizable improvements in test BLEU on two Arabic-English datasets: IWSLT (+0.65) and NIST (+0.37).} }