Intrinsic Evaluations of Word Embeddings: What Can We Do Better?

@inproceedings{Rogers2016IntrinsicEO,
  title={Intrinsic Evaluations of Word Embeddings: What Can We Do Better?},
  author={Anna Rogers and Aleksandr Drozd},
  booktitle={RepEval@ACL},
  year={2016}
}
This paper presents an analysis of existing methods for the intrinsic evaluation of word embeddings. We show that the main methodological premise of such evaluations is “interpretability” of word embeddings: a “good” embedding produces results that make sense in terms of traditional linguistic categories. This approach is not only of limited practical use, but also fails to do justice to the strengths of distributional meaning representations. We argue for a shift from abstract ratings of word… Expand
68 Citations
A Survey of Word Embeddings Evaluation Methods
  • 96
  • PDF
Geographical Evaluation of Word Embeddings
  • 6
  • PDF
Evaluating the Stability of Embedding-based Word Similarities
  • 80
  • PDF
Investigating Different Syntactic Context Types and Context Representations for Learning Word Embeddings
  • 28
  • PDF
Evaluating word embedding models: methods and experimental results
  • 51
  • PDF
Evaluation of Greek Word Embeddings
  • 4
  • PDF
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 34 REFERENCES
Evaluation methods for unsupervised word embeddings
  • 408
  • Highly Influential
  • PDF
How to Generate a Good Word Embedding
  • 229
  • PDF
Analogy-based detection of morphological and semantic relations with word embeddings: what works and what doesn't
  • 118
  • PDF
Evaluation of Word Vector Representations by Subspace Alignment
  • 135
  • PDF
Controlled Experiments for Word Embeddings
  • 15
  • PDF
The Role of Context Types and Dimensionality in Learning Word Embeddings
  • 95
  • PDF
Specializing Word Embeddings for Similarity or Relatedness
  • 124
  • PDF
How we BLESSed distributional semantic evaluation
  • 267
  • PDF
Efficient Non-parametric Estimation of Multiple Embeddings per Word in Vector Space
  • 386
  • PDF
...
1
2
3
4
...