Sampling-Based Approximations to Minimum Bayes Risk Decoding for Neural Machine Translation

TL;DR

This paper explores sampling-based approximations to minimum Bayes risk decoding in neural machine translation, addressing limitations of beam search and enabling larger hypothesis spaces for improved translation quality.

Contribution

It introduces new approximations to MBR decoding that avoid the beam search curse and decouple exploration from utility estimation, allowing larger hypothesis sets.

Findings

MBR approximations do not suffer from the beam search curse.

Larger hypothesis spaces improve translation quality.

Mode-seeking strategies help identify promising hypotheses.

Abstract

In NMT we search for the mode of the model distribution to form predictions. The mode and other high-probability translations found by beam search have been shown to often be inadequate in a number of ways. This prevents improving translation quality through better search, as these idiosyncratic translations end up selected by the decoding algorithm, a problem known as the beam search curse. Recently, an approximation to minimum Bayes risk (MBR) decoding has been proposed as an alternative decision rule that would likely not suffer from the same problems. We analyse this approximation and establish that it has no equivalent to the beam search curse. We then design approximations that decouple the cost of exploration from the cost of robust estimation of expected utility. This allows for much larger hypothesis spaces, which we show to be beneficial. We also show that mode-seeking strategies can aid in constructing compact sets of promising hypotheses and that MBR is effective in identifying good translations in them. We conduct experiments on three language pairs varying in amounts of resources available: English into and from German, Romanian, and Nepali.