# Improving Neural Machine Translation through Phrase-based Forced Decoding

arXiv: https://arxiv.org/abs/1711.00309 · HTML: https://mlsanity.com/paper/1711.00309/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 1 Nov 2017.

## Built on by (most cited first, 2 of 2)

- [Language Modeling, Lexical Translation, Reordering: The Training Process of NMT through the Lens of Classical SMT](https://mlsanity.com/paper/2109.01396/) (2021)
- [Neural Rule-Execution Tracking Machine For Transformer-Based Text Generation](https://mlsanity.com/paper/2107.13077/) (2021)

## Abstract (arXiv)

Compared to traditional statistical machine translation (SMT), neural machine translation (NMT) often sacrifices adequacy for the sake of fluency. We propose a method to combine the advantages of traditional SMT and NMT by exploiting an existing phrase-based SMT model to compute the phrase-based decoding cost for an NMT output and then using this cost to rerank the n-best NMT outputs. The main challenge in implementing this approach is that NMT outputs may not be in the search space of the standard phrase-based decoding algorithm, because the search space of phrase-based SMT is limited by the phrase-based translation rule table. We propose a soft forced decoding algorithm, which can always successfully find a decoding path for any NMT output. We show that using the forced decoding cost to rerank the NMT outputs can successfully improve translation quality on four different language pairs.