How to (Properly) Evaluate Cross-Lingual Word Embeddings: On Strong Baselines, Comparative Analyses, and Some Misconceptions

@inproceedings{Glavas2019HowT,
  title={How to (Properly) Evaluate Cross-Lingual Word Embeddings: On Strong Baselines, Comparative Analyses, and Some Misconceptions},
  author={Goran Glavas and Robert Litschko and Sebastian Ruder and Ivan Vulic},
  booktitle={ACL},
  year={2019}
}
Cross-lingual word embeddings (CLEs) enable multilingual modeling of meaning and facilitate cross-lingual transfer of NLP models. Despite their ubiquitous usage in downstream tasks, recent increasingly popular projection-based CLE models are almost exclusively evaluated on a single task only: bilingual lexicon induction (BLI). Even BLI evaluations vary greatly, hindering our ability to correctly interpret performance and properties of different CLE models. In this work, we make the first step… Expand
View on ACL
arxiv.org
91 Citations
Highly Influential Citations
12
Background Citations
62
Methods Citations
48
Results Citations
10

91 Citations

Citation Type

Citation Type

Do We Really Need Fully Unsupervised Cross-Lingual Embeddings?
  • 31
  • PDF
Robust Cross-lingual Embeddings from Parallel Sentences
  • 2
  • PDF
Investigating Cross-Lingual Alignment Methods for Contextualized Embeddings with Token-Level Evaluation
  • 12
  • PDF
Analyzing Zero-shot Cross-lingual Transfer in Supervised NLP Tasks
  • PDF
Why Overfitting Isn’t Always Bad: Retrofitting Cross-Lingual Word Embeddings to Dictionaries
  • 3
  • PDF
On the Robustness of Unsupervised and Semi-supervised Cross-lingual Word Embedding Learning
  • 12
  • Highly Influenced
  • PDF
Beyond Offline Mapping: Learning Cross Lingual Word Embeddings through Context Anchoring
  • Highly Influenced
  • PDF
Lost in Embedding Space: Explaining Cross-Lingual Task Performance with Eigenvalue Divergence
  • 2
Training Effective Neural CLIR by Bridging the Translation Gap
  • Highly Influenced
Evaluating Multilingual Text Encoders for Unsupervised Cross-Lingual Retrieval
  • 1
  • PDF
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 67 REFERENCES
Cross-lingual Models of Word Embeddings: An Empirical Comparison
  • 168
  • PDF
Improving Cross-Lingual Word Embeddings by Meeting in the Middle
  • 31
  • PDF
A Strong Baseline for Learning Cross-Lingual Word Embeddings from Sentence Alignments
  • 50
  • PDF
Unsupervised Cross-Lingual Information Retrieval Using Monolingual Data Only
  • 37
  • PDF
A Survey of Cross-lingual Word Embedding Models
  • 218
  • PDF
Ten Pairs to Tag - Multilingual POS Tagging via Coarse Mapping between Embeddings
  • 87
  • Highly Influential
  • PDF
XNLI: Evaluating Cross-lingual Sentence Representations
  • 281
  • Highly Influential
  • PDF
Unsupervised Cross-lingual Transfer of Word Embedding Spaces
  • 57
  • PDF
Monolingual and Cross-Lingual Information Retrieval Models Based on (Bilingual) Word Embeddings
  • 218
  • PDF
Massively Multilingual Sentence Embeddings for Zero-Shot Cross-Lingual Transfer and Beyond
  • 287
  • PDF
...
1
2
3
4
5
...