Subword-augmented Embedding for Cloze Reading Comprehension
This work addresses a specific bottleneck in reading comprehension models for NLP researchers, offering an incremental improvement in embedding strategies.
The paper tackled the problem of suboptimal word embeddings in cloze-style reading comprehension by proposing subword-augmented embeddings instead of character-level ones, resulting in significant performance improvements over state-of-the-art baselines on various public datasets.
Representation learning is the foundation of machine reading comprehension. In state-of-the-art models, deep learning methods broadly use word and character level representations. However, character is not naturally the minimal linguistic unit. In addition, with a simple concatenation of character and word embedding, previous models actually give suboptimal solution. In this paper, we propose to use subword rather than character for word embedding enhancement. We also empirically explore different augmentation strategies on subword-augmented embedding to enhance the cloze-style reading comprehension model reader. In detail, we present a reader that uses subword-level representation to augment word embedding with a short list to handle rare words effectively. A thorough examination is conducted to evaluate the comprehensive performance and generalization ability of the proposed reader. Experimental results show that the proposed approach helps the reader significantly outperform the state-of-the-art baselines on various public datasets.