Obtaining Better Static Word Embeddings Using Contextual Embedding Models
This work addresses efficiency and interpretability issues in NLP for practitioners, though it is incremental as it builds on existing distillation and CBOW-based methods.
The authors tackled the problem of high computational cost and poor interpretability of contextual word embeddings by proposing a distillation method that improves computational efficiency and outperforms existing static embeddings, achieving better quality in NLP tasks.
The advent of contextual word embeddings -- representations of words which incorporate semantic and syntactic information from their context -- has led to tremendous improvements on a wide variety of NLP tasks. However, recent contextual models have prohibitively high computational cost in many use-cases and are often hard to interpret. In this work, we demonstrate that our proposed distillation method, which is a simple extension of CBOW-based training, allows to significantly improve computational efficiency of NLP applications, while outperforming the quality of existing static embeddings trained from scratch as well as those distilled from previously proposed methods. As a side-effect, our approach also allows a fair comparison of both contextual and static embeddings via standard lexical evaluation tasks.