Dependency-based Convolutional Neural Networks for Sentence Embedding
This work addresses the limitation of sequential processing in sentence embedding for NLP tasks, offering a domain-specific advancement.
The paper tackled the problem of sentence modeling and classification by proposing a tree-based convolutional neural network that incorporates long-distance dependencies, achieving state-of-the-art accuracy on TREC and improvements on sentiment and question classification tasks.
In sentence modeling and classification, convolutional neural network approaches have recently achieved state-of-the-art results, but all such efforts process word vectors sequentially and neglect long-distance dependencies. To exploit both deep learning and linguistic structures, we propose a tree-based convolutional neural network model which exploit various long-distance relationships between words. Our model improves the sequential baselines on all three sentiment and question classification tasks, and achieves the highest published accuracy on TREC.