Dependency-based Convolutional Neural Networks for Sentence Embedding
Computation and Language
2015-08-04 v2 Artificial Intelligence
Machine Learning
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.
Cite
@article{arxiv.1507.01839,
title = {Dependency-based Convolutional Neural Networks for Sentence Embedding},
author = {Mingbo Ma and Liang Huang and Bing Xiang and Bowen Zhou},
journal= {arXiv preprint arXiv:1507.01839},
year = {2015}
}
Comments
this paper has been accepted by ACL 2015