Research map / Large language models / Language, safety and interpretability
Learning Better Name Translation for Cross-Lingual Wikification
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 |
|---|---|---|---|
| use joint word alignment + word transliteration model instead of best baseline (P&R with f-align) | name translation · Wikipedia title translation avg | fuzzy F1: 75.45 -> 78.22 | Helped |
| joint model vs P&R f-align on organization names | name translation · Wikipedia title translation, ORG avg | fuzzy F1: 69.22 -> 73.14 | Helped |
| joint model vs P&R f-align on person names | name translation · Wikipedia title translation, PER avg | fuzzy F1: 83.54 -> 84.45 | Helped |
| joint model vs P&R f-align for candidate generation | cross-lingual wikification candidate generation · avg of 9 languages | % gold title in candidates: 57.19 -> 60.12 | Helped |
| joint model vs P&R f-align for candidate generation on German | candidate generation · German | % gold title in candidates: 61.63 -> 61.37 | Hurt |
| add proposed name translation model to base TAC 2016 EDL system | cross-lingual wikifier · TAC 2016 EDL Spanish | F1: 53.71 -> 60.67 | Helped |
| add proposed name translation model to base TAC 2016 EDL system | cross-lingual wikifier · TAC 2016 EDL Chinese | F1: 63.62 -> 65.53 | Helped |
| align words with a word alignment model (f-align) instead of position (p-align) for baseline transliteration models | P&R · Wikipedia title translation avg | fuzzy F1: 73.30 -> 75.45 | Helped |
| higher fuzzy F1 does not imply better candidate generation (NMT-char on Spanish LOC) | NMT-char · Spanish LOC | fuzzy F1 vs candidate coverage: 73.56 vs 27.86 | Mixed |
Closest papers by meaning
- Neural Cross-Lingual Entity Linking (AAAI 2018 · similarity 0.87)
- Enhanced Meta-Learning for Cross-Lingual Named Entity Recognition with Minimal Resources (AAAI 2020 · similarity 0.86)
- MCL-NER: Cross-Lingual Named Entity Recognition via Multi-View Contrastive Learning (AAAI 2024 · similarity 0.84)
- Constrained Decoding for Cross-lingual Label Projection (ICLR 2024 · similarity 0.84)
- Three Heads Are Better than One: Improving Cross-Domain NER with Progressive Decomposed Network (AAAI 2024 · similarity 0.83)
- Enhancing Machine Translation Experiences with Multilingual Knowledge Graphs (AAAI 2024 · similarity 0.82)
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.