Analytical Methods for Interpretable Ultradense Word Embeddings
This work addresses the lack of interpretability in word embeddings for NLP applications, offering incremental improvements in method robustness and efficiency.
The paper tackles the problem of making word embeddings interpretable by rotating word spaces to identify interpretable dimensions without information loss, proposing DensRay as a new method that is closed-form, hyperparameter-free, and more robust than existing approaches, with evaluation on lexicon induction and word analogy tasks.
Word embeddings are useful for a wide variety of tasks, but they lack interpretability. By rotating word spaces, interpretable dimensions can be identified while preserving the information contained in the embeddings without any loss. In this work, we investigate three methods for making word spaces interpretable by rotation: Densifier (Rothe et al., 2016), linear SVMs and DensRay, a new method we propose. In contrast to Densifier, DensRay can be computed in closed form, is hyperparameter-free and thus more robust than Densifier. We evaluate the three methods on lexicon induction and set-based word analogy. In addition we provide qualitative insights as to how interpretable word spaces can be used for removing gender bias from embeddings.