Dependency-based Convolutional Neural Networks for Sentence Embedding

TL;DR

This paper introduces a tree-based convolutional neural network that captures long-distance dependencies in sentences, improving performance on classification tasks over sequential models.

Contribution

It presents a novel tree-based CNN model that leverages linguistic structures to enhance sentence embedding and classification accuracy.

Findings

Improved accuracy on sentiment and question classification tasks

Achieved highest published accuracy on TREC dataset

Outperformed sequential CNN baselines

Abstract

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.