Finding Better Subword Segmentation for Neural Machine Translation

By Yingting Wu, Hai Zhao. 2018

24 citations recorded by the source corpus.

For different language pairs, word-level neural machine translation (NMT) models with a fixed-size vocabulary suffer from the same problem of representing out-of-vocabulary (OOV) words. The common practice usually replaces all these rare or unknown words with a \(\langle \)UNK\(\rangle \) token, which limits the translation performance to some extent. Most of recent work handled such a problem by splitting words into characters or other specially extracted subword units to enable open-vocabulary translation. Byte pair encoding (BPE) is one of the successful attempts that has been shown extremely competitive by providing effective subword segmentation for NMT systems. In this paper, we extend the BPE style segmentation to a general unsupervised framework with three statistical measures: frequency (FRQ), accessor variety (AV) and description length gain (DLG). We test our approach on two translation tasks: German to English and Chinese to English. The experimental results show that AV and DLG enhanced systems outperform the FRQ baseline in the frequency weighted schemes at different significant levels.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Rui Wang
2. Zhao Hai
3. Felix Stahlberg
4. Fengshun Xiao
5. Kehai Chen
6. Jiangtong Li
7. Zhuosheng Zhang
8. Bill Byrne

In Computer Vision and Pattern Recognition

1. Felix Stahlberg
2. Rui Wang
3. Kehai Chen
4. Zhao Hai
5. Jiangtong Li

Citations in the corpus

• Neural Machine Translation: A Review (2019, background): “11Wu and Zhao [257] proposed alternatives to the co-occurrence counts.”
• Lattice-Based Transformer Encoder for Neural Machine Translation (2019, background): “…OOV problems, recent research has modeled source sentences at character level (Ling et al., 2015; Costa-Jussa and Fonollosa, 2016; Yang et al., 2016; Lee et al., 2016), subword level (Sennrich et al., 2015; Kudo, 2018; Wu and Zhao, 2018) and mixed character-word level (Luong and Manning, 2016).”