@inproceedings{5b17e3cd26f24196a6ef581d0e07fc65,
title = "An encoder-decoder architecture with graph convolutional networks for abstractive summarization",
abstract = "We propose a single-document abstractive summarization system that integrates token relation into a traditional RNN-based encoder-decoder architecture. We employ pointer-wise mutual information to represent the token relation and adopt Graph Convolutional Networks (GCN) to extract token representation from the relation graph. In our experiment on Gigaword, we consider importing two kinds of structural information: token (node) representation from the relation graph. Also, we implement two kinds of GCNs, a spectral-based one and a spatial-based one, to extract structural information. The result shows that the spatial based GCN-enhanced model with node representation outperforms the classical RNN-based encoder-decoder model.",
keywords = "GCN, Natural language processing, Seq2Seq, text summarization, natural language processing",
author = "Gangmin Li and QiAo Yuan and Pin Ni and Junru Liu and Xiangzhi Tong and Hanzhe Lu and Steven Guan",
note = "Publisher Copyright: {\textcopyright} 2021 IEEE.; 2021 IEEE 4th International Conference on Big Data and Artificial Intelligence (BDAI) ; Conference date: 02-07-2021 Through 04-07-2021",
year = "2021",
month = aug,
day = "20",
doi = "10.1109/BDAI52447.2021.9515256",
language = "English",
isbn = "9781665412704",
series = "2021 IEEE 4th International Conference on Big Data and Artificial Intelligence, BDAI 2021",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
pages = "91--97",
booktitle = "2021 IEEE 4th International Conference on Big Data and Artificial Intelligence, BDAI 2021",
address = "United States",
}