Structural Embedding of Syntactic Trees for Machine Comprehension
This addresses the challenge of enhancing machine comprehension accuracy for natural language processing tasks, though it is incremental as it builds on existing neural attention models.
The paper tackles the problem of machine comprehension by incorporating structured linguistic information, such as syntactic trees, into neural networks, resulting in improved performance on the SQuAD dataset with more syntactically coherent answer extraction compared to baseline methods.
Deep neural networks for machine comprehension typically utilizes only word or character embeddings without explicitly taking advantage of structured linguistic information such as constituency trees and dependency trees. In this paper, we propose structural embedding of syntactic trees (SEST), an algorithm framework to utilize structured information and encode them into vector representations that can boost the performance of algorithms for the machine comprehension. We evaluate our approach using a state-of-the-art neural attention model on the SQuAD dataset. Experimental results demonstrate that our model can accurately identify the syntactic boundaries of the sentences and extract answers that are syntactically coherent over the baseline methods.