RMBR: A Regularized Minimum Bayes Risk Reranking Framework for Machine Translation

TL;DR

This paper introduces RMBR, a reranking framework for neural machine translation that improves translation quality by incorporating semantic similarity, translation quality, and model uncertainty into the MBR decoding process.

Contribution

The paper proposes a novel regularized MBR reranking framework that addresses limitations of traditional MBR by integrating semantic similarity, quality, and uncertainty measures.

Findings

RMBR outperforms standard beam search and traditional MBR in translation quality.

Incorporating semantic and uncertainty regularizers improves translation accuracy.

Extensive experiments validate the effectiveness of the proposed framework.

Abstract

Beam search is the most widely used decoding method for neural machine translation (NMT). In practice, the top-1 candidate with the highest log-probability among the n candidates is selected as the preferred one. However, this top-1 candidate may not be the best overall translation among the n-best list. Recently, Minimum Bayes Risk (MBR) decoding has been proposed to improve the quality for NMT, which seeks for a consensus translation that is closest on average to other candidates from the n-best list. We argue that MBR still suffers from the following problems: The utility function only considers the lexical-level similarity between candidates; The expected utility considers the entire n-best list which is time-consuming and inadequate candidates in the tail list may hurt the performance; Only the relationship between candidates is considered. To solve these issues, we design a regularized MBR reranking framework (RMBR), which considers semantic-based similarity and computes the expected utility for each candidate by truncating the list. We expect the proposed framework to further consider the translation quality and model uncertainty of each candidate. Thus the proposed quality regularizer and uncertainty regularizer are incorporated into the framework. Extensive experiments on multiple translation tasks demonstrate the effectiveness of our method.