DbAPE: Denoising-Based APE System for Improving English-Myanmar NMT
May Myo Zin, Teeradaj Racharak, Minh Le Nguyen
IEEE Access · 2022
doi:10.1109/access.2022.3185415 (opens in a new tab)
We email you when a new notice or a question about this paper appears.
Abstract
Automatic post-editing (APE) research aims to investigate methods for correcting systematic errors in machine translation (MT) results. Recent work has shown successful practices of APE for improving MT output quality; however, their effectiveness strongly relies on the availability of large-scale human-created APE triplets. The high production cost of human post-edited data has led to the absence of APE triplets for most language pairs, including English-Myanmar, which has become a limiting factor for the applicability of the APE task. This work investigates how to conduct the APE task on the English-Myanmar MT where human-edited APE triplets are unavailable. We build a denoising-based APE (DbAPE) system using only the monolingual and parallel MT corpora. The system takes the source sentence (<italic>src</italic>) and the MT output (<italic>mt</italic>) as inputs and produces the <italic>post-edited mt</italic> as output by operating the three processes together, including word alignment extraction, enriching <italic>mt</italic> using the extracted word alignment information, and denoising the enriched-version of <italic>mt</italic>. We conduct extensive experiments by applying our APE system as a post-processor to the raw output of the existing English-Myanmar MT systems. APE translations produced by DbAPE show statistically significant improvements of at least +4% BLEU and −16% TER points absolute over the original NMT. Moreover, DbAPE can improve the quality of the texts generated by state-of-the-art systems such as mT5 and Google Translate. In addition, we perform word alignment experiments with four types of alignment methods and demonstrate that the proposed multilingual word aligner can achieve robust performance over previous state-of-the-art models.
Abstract licence: CC BY-NC-ND
Reads
Read by people on this page
None yet.
Checked or read by AI through naulon
None yet.
A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.
Are you an author?
Is this your paper?
Claim it to add your note and be paid when agents cite it. Free.