Research map / Label-efficient and robust learning / Recognition and long-tail
Building Earth Mover's Distance on Bilingual Word Embeddings for Machine Translation
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace nearest neighbor retrieval with EMD matching | BilBOWA embeddings · EN-IT word translation | accuracy (%): 7.11 -> 28.20 | Helped |
| replace nearest neighbor retrieval with EMD matching | BilBOWA embeddings · ZH-EN word translation | accuracy (%): 26.14 -> 29.07 | Helped |
| replace nearest neighbor with globally-corrected retrieval (GC) | NN retrieval · ZH-EN word translation | accuracy (%): 26.14 -> 25.71 | No clear effect |
| use globally-corrected retrieval instead of NN | NN retrieval · EN-IT word translation | accuracy (%): 7.11 -> 16.55 | Helped |
| add IDF weighting to EMD sentence distance (ablation) | EMD corpus filter · Corpus filtering, noise 0.2 | filtering accuracy (%): 75.50 -> 83.07 | Helped |
| add IDF weighting to EMD sentence distance (ablation) | EMD corpus filter · Chinese-English MT NIST avg | BLEU: 27.32 -> 27.97 | Helped |
| filter training corpus with EMD vs IBM Model 1 | Moses MT training data · Chinese-English MT NIST avg | BLEU: 26.84 -> 27.97 | Helped |
| filter training corpus with IBM Model 1 | Moses MT training data · Chinese-English MT NIST avg | BLEU: 26.93 -> 26.84 | No clear effect |
Closest papers by meaning
- Word translation without parallel data (ICLR 2018 · similarity 0.89)
- Nearest Neighbor Machine Translation (ICLR 2021 · similarity 0.89)
- Bilingual Lexicon Induction from Non-Parallel Data with Minimal Supervision (AAAI 2017 · similarity 0.88)
- EMO: EARTH MOVER DISTANCE OPTIMIZATION FOR AUTO-REGRESSIVE LANGUAGE MODELING (ICLR 2024 · similarity 0.87)
- The Scandinavian Embedding Benchmarks: Comprehensive Assessment of Multilingual and Monolingual Text Embedding (NeurIPS 2024 · similarity 0.86)
- MMTEB: Massive Multilingual Text Embedding Benchmark (ICLR 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.