mlsanityData to 14 Oct 2026

On the Limitations of Unsupervised Bilingual Dictionary Induction

arXiv 1805.03620 ↗First posted 9 May 2018cs.CL cs.LG stat.MLcited by 61

Not on the map. Why?

Abstract (arXiv)

Unsupervised machine translation---i.e., not assuming any cross-lingual supervision signal, whether a dictionary, translations, or comparable corpora---seems impossible, but nevertheless, Lample et al. (2018) recently proposed a fully unsupervised machine translation (MT) model. The model relies heavily on an adversarial, unsupervised alignment of word embedding spaces for bilingual dictionary induction (Conneau et al., 2018), which we examine here. Our results identify the limitations of current unsupervised MT: unsupervised bilingual dictionary induction performs much worse on morphologically rich languages that are not dependent marking, when monolingual corpora from different domains or different embedding algorithms are used. We show that a simple trick, exploiting a weak supervision signal from identical words, enables more robust induction, and establish a near-perfect correlation between unsupervised bilingual dictionary induction performance and a previously unexplored graph similarity metric.

Key facts. Cited by 61 papers in the map (data through 14 Oct 2026). First posted on arXiv 9 May 2018.

Builds on

No reference list yet.

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 18; 2022: 18; 2023: 11; 2024: 10; 2025: 2; 2026: 2.

61 papers.

Sort:

By area:

  1. 2022Challenges and Strategies in Cross-Cultural NLPcited by 218
  2. 2021XTREME-R: Towards More Challenging and Nuanced Multilingual Evaluationcited by 168
  3. 2021Neural Machine Translation for Low-Resource Languages: A Surveycited by 109
  4. 2024A Survey on Multilingual Large Language Models: Corpora, Alignment, and Biascited by 84
  5. 2024Understanding Cross-Lingual Alignment -- A Surveycited by 34
  6. 2023Do Vision and Language Models Share Concepts? A Vector Space Alignment Studycited by 27
  7. 2021Bilingual Lexicon Induction via Unsupervised Bitext Construction and Word Alignmentcited by 19
  8. 2021Do Language Models Know the Way to Rome?cited by 18
  9. 2022Improving Word Translation via Two-Stage Contrastive Learningcited by 18
  10. 2023Transfer to a Low-Resource Language via Close Relatives: The Case Study on Faroesecited by 15
  11. 2021AM2iCo: Evaluating Word Meaning in Context across Low-Resource Languages with Adversarial Examplescited by 14
  12. 2021Itihasa: A large-scale corpus for Sanskrit to English translationcited by 13
38 more
  1. 2024Concept Space Alignment in Multilingual LLMscited by 13
  2. 2021Combining Static Word Embeddings and Contextual Representations for Bilingual Lexicon Inductioncited by 12
  3. 2021Subword Mapping and Anchoring across Languagescited by 12
  4. 2021Multi-source Learning via Completion of Block-wise Overlapping Noisy Matricescited by 10
  5. 2023Beyond Shared Vocabulary: Increasing Representational Word Similarities across Languages for Multilingual Machine Translationcited by 9
  6. 2022Delving Deeper into Cross-lingual Visual Question Answeringcited by 8
  7. 2022IsoVec: Controlling the Relative Isomorphism of Word Embedding Spacescited by 7
  8. 2023Neural Machine Translation for the Indigenous Languages of the Americas: An Introductioncited by 7
  9. 2023Cultural Adaptation of Recipescited by 7
  10. 2021Cross-Lingual BERT Contextual Embedding Space Mapping with Isotropic and Isometric Conditionscited by 6
  11. 2022Bilingual Lexicon Induction for Low-Resource Languages using Graph Matching via Optimal Transportcited by 6
  12. 2022The Impact of Cross-Lingual Adjustment of Contextual Word Representations on Zero-Shot Transfercited by 5
  13. 2021indicnlp@kgp at DravidianLangTech-EACL2021: Offensive Language Identification in Dravidian Languagescited by 4
  14. 2022Did AI get more negative recently?cited by 4
  15. 2024The Shape of Word Embeddings: Quantifying Non-Isometry With Topological Data Analysiscited by 4
  16. 2021Towards Multi-Sense Cross-Lingual Alignment of Contextual Embeddingscited by 3
  17. 2021Cross-lingual alignments of ELMo contextual embeddingscited by 3
  18. 2022Systematic Investigation of Strategies Tailored for Low-Resource Settings for Low-Resource Dependency Parsingcited by 3
  19. 2022Sub-Word Alignment Is Still Useful: A Vest-Pocket Method for Enhancing Low-Resource Machine Translationcited by 3
  20. 2022RAPO: An Adaptive Ranking Paradigm for Bilingual Lexicon Inductioncited by 3
  21. 2023GRI: Graph-based Relative Isomorphism of Word Embedding Spacescited by 3
  22. 2023GARI: Graph Attention for Relative Isomorphism of Arabic Word Embeddingscited by 3
  23. 2021Word Embedding Transformation for Robust Unsupervised Bilingual Lexicon Inductioncited by 2
  24. 2021Learning a Reversible Embedding Mapping using Bi-Directional Manifold Alignmentcited by 2
  25. 2021An Analysis of Euclidean vs. Graph-Based Framing for Bilingual Lexicon Induction from Word Embedding Spacescited by 2
  26. 2022AutoMap: Automatic Medical Code Mapping for Clinical Prediction Model Deploymentcited by 2
  27. 2023Low-resource Bilingual Dialect Lexicon Induction with Large Language Modelscited by 2
  28. 2023ProMap: Effective Bilingual Lexicon Induction via Language Model Promptingcited by 2
  29. 2024Modular Sentence Encoders: Separating Language Specialization from Cross-Lingual Alignmentcited by 2
  30. 2021Examining Cross-lingual Contextual Embeddings with Orthogonal Structural Probescited by 1
  31. 2022Constrained Density Matching and Modeling for Cross-lingual Alignment of Contextualized Representationscited by 1
  32. 2022Isomorphic Cross-lingual Embeddings for Low-Resource Languagescited by 1
  33. 2022Don't Forget Cheap Training Signals Before Building Unsupervised Bilingual Word Embeddingscited by 1
  34. 2022Robust Unsupervised Cross-Lingual Word Embedding using Domain Flow Interpolationcited by 1
  35. 2022Domain Mismatch Doesn't Always Prevent Cross-Lingual Transfer Learningcited by 1
  36. 2023Multilingual Word Embeddings for Low-Resource Languages using Anchors and a Chain of Related Languagescited by 1
  37. 2024Cross-lingual Contextualized Phrase Retrievalcited by 1
  38. 2026An Efficient and Privacy-Preserving Architecture for Cross-Institutional Collaborative RAGcited by 1

50 most cited of 61 All 61 (JSON)