EMNLP Findings2024
Translation of Multifaceted Data without Re-Training of Machine Translation Systems
Hyeonseok Moon, Seungyoon Lee, Seongtae Hong, Seungjun Lee, Chanjun Park, Heuiseok Lim
Abstract
Translating major language resources to build minor language resources becomes a widelyused approach. Particularly in translating complex data points composed of multiple components, it is common to translate each component separately. However, we argue that this practice often overlooks the interrelation between components within the same data point. To address this limitation, we propose a novel MT pipeline that considers the intra-data relation 1 in implementing MT for training data. In our MT pipeline, all the components in a data point are concatenated to form a single translation sequence and subsequently reconstructed to the data components after translation. We introduce a Catalyst Statement (CS) to enhance the intra-data relation, and Indicator Token (IT) to assist the decomposition of a translated sequence into its respective data components. Through our approach, we have achieved a considerable improvement in translation quality itself, along with its effectiveness as training data. Compared with the conventional approach that translates each data component separately, our method yields better training data that enhances the performance of the trained model by 2.690 points for the web page ranking (WPR) task, and 0.845 for the question generation (QG) task in the XGLUE benchmark.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
Lineage
- Frustratingly Easy Label Projection for Cross-lingual Transfer cites MLQA: Evaluating Cross-lingual Extractive Question Answering
- Frustratingly Easy Label Projection for Cross-lingual Transfer cites Unsupervised Cross-lingual Representation Learning at Scale
- MLQA: Evaluating Cross-lingual Extractive Question Answering cites XGLUE: A New Benchmark Datasetfor Cross-lingual Pre-training, Understanding and Generation
- XGLUE: A New Benchmark Datasetfor Cross-lingual Pre-training, Understanding and Generation cites MLQA: Evaluating Cross-lingual Extractive Question Answering
- Unsupervised Cross-lingual Representation Learning at Scale cites MLQA: Evaluating Cross-lingual Extractive Question Answering
Builds on7
- Unsupervised Cross-lingual Representation Learning at ScaleAlexis Conneau, Kartikay Khandelwal, Naman Goyal, Vishrav Chaudhary et al.ACL 2020 · 539 citations
- AlpaGasus: Training a Better Alpaca with Fewer DataLichang Chen, Shiyang Li, Jun Yan, Hai Wang et al.ICLR 2024 · 295 citations
- XGLUE: A New Benchmark Datasetfor Cross-lingual Pre-training, Understanding and GenerationYaobo Liang, Nan Duan, Yeyun Gong, Ning Wu et al.EMNLP 2020 · 232 citations
- MLQA: Evaluating Cross-lingual Extractive Question AnsweringPatrick Lewis, Barlas Oguz, Ruty Rinott, Sebastian Riedel et al.ACL 2020 · 52 citations
- Frustratingly Easy Label Projection for Cross-lingual TransferYang Chen, Chao Jiang, Alan Ritter, Wei XuACL Findings 2023 · 11 citations
Related papers
- CTQScorer: Combining Multiple Features for In-context Example Selection for Machine TranslationAswanth M., Ratish Puduppully, Raj Dabre, Anoop KunchukuttanEMNLP Findings 2023 · 4 citations
- PreQuEL: Quality Estimation of Machine Translation Outputs in AdvanceShachar Don-Yehiya, Leshem Choshen, Omri AbendEMNLP 2022 · 4 citations
- A Simple Multi-Modality Transfer Learning Baseline for Sign Language TranslationYutong Chen, Fangyun Wei, Xiao Sun, Zhirong Wu et al.CVPR 2022 · 137 citations
- Are Multilingual Models the Best Choice for Moderately Under-resourced Languages? A Comprehensive Assessment for CatalanJordi Armengol-Estapé, Casimiro Pio Carrino, Carlos Rodríguez Penagos, Ona De Gibert Bonet et al.ACL Findings 2021 · 21 citations
- Lost in Translation, Found in Spans: Identifying Claims in Multilingual Social MediaShubham Mittal, Megha Sundriyal, Preslav NakovEMNLP 2023 · 4 citations
