Multi-Sentence Resampling: A Simple Approach to Alleviate Dataset Length Bias and Beam-Search Degradation

TL;DR

This paper identifies dataset length-bias as a key factor in beam-search degradation in neural machine translation and proposes Multi-Sentence Resampling to improve translation quality for long sentences.

Contribution

The paper introduces Multi-Sentence Resampling, a novel data augmentation method that mitigates length-bias and improves beam-search performance in NMT.

Findings

MSR reduces quality degradation with larger beams

MSR improves translation quality on multiple datasets

Length-bias significantly impacts beam-search performance

Abstract

Neural Machine Translation (NMT) is known to suffer from a beam-search problem: after a certain point, increasing beam size causes an overall drop in translation quality. This effect is especially pronounced for long sentences. While much work was done analyzing this phenomenon, primarily for autoregressive NMT models, there is still no consensus on its underlying cause. In this work, we analyze errors that cause major quality degradation with large beams in NMT and Automatic Speech Recognition (ASR). We show that a factor that strongly contributes to the quality degradation with large beams is \textit{dataset length-bias} - \textit{NMT datasets are strongly biased towards short sentences}. To mitigate this issue, we propose a new data augmentation technique -- \textit{Multi-Sentence Resampling (MSR)}. This technique extends the training examples by concatenating several sentences from the original dataset to make a long training example. We demonstrate that MSR significantly reduces degradation with growing beam size and improves final translation quality on the IWSTL$15$ En-Vi, IWSTL$17$ En-Fr, and WMT$14$ En-De datasets.