The Expressive Power of Word Embeddings

TL;DR

This paper investigates the semantic capabilities of various word embeddings, revealing significant differences in their quality and the influence of their dimensionality on capturing nuanced meanings for NLP tasks.

Contribution

It introduces tasks to evaluate embeddings' semantic properties and analyzes how dimensionality affects their effectiveness, providing insights into embedding quality.

Findings

Embeddings capture nuanced semantics even without sentence structure.

Significant variance exists in the quality of different embeddings.

Dimensionality impacts the features and effectiveness of embeddings.

Abstract

We seek to better understand the difference in quality of the several publicly released embeddings. We propose several tasks that help to distinguish the characteristics of different embeddings. Our evaluation of sentiment polarity and synonym/antonym relations shows that embeddings are able to capture surprisingly nuanced semantics even in the absence of sentence structure. Moreover, benchmarking the embeddings shows great variance in quality and characteristics of the semantics captured by the tested embeddings. Finally, we show the impact of varying the number of dimensions and the resolution of each dimension on the effective useful features captured by the embedding space. Our contributions highlight the importance of embeddings for NLP tasks and the effect of their quality on the final results.