Representations of Syntax [MASK] Useful: Effects of Constituency and Dependency Structure in Recursive LSTMs
This work addresses the challenge of enhancing syntactic biases in sequential models for natural language processing, offering an incremental improvement over existing methods.
The paper tackled the problem of improving syntactic task performance in sequence-based neural networks by evaluating the effectiveness of constituency and dependency parse structures in recursive LSTMs. The result showed that constituency-based networks generalize more robustly than dependency-based ones, and data augmentation via fine-tuning on constructed data substantially improved syntactic robustness.
Sequence-based neural networks show significant sensitivity to syntactic structure, but they still perform less well on syntactic tasks than tree-based networks. Such tree-based networks can be provided with a constituency parse, a dependency parse, or both. We evaluate which of these two representational schemes more effectively introduces biases for syntactic structure that increase performance on the subject-verb agreement prediction task. We find that a constituency-based network generalizes more robustly than a dependency-based one, and that combining the two types of structure does not yield further improvement. Finally, we show that the syntactic robustness of sequential models can be substantially improved by fine-tuning on a small amount of constructed data, suggesting that data augmentation is a viable alternative to explicit constituency structure for imparting the syntactic biases that sequential models are lacking.