@inproceedings{liu-etal-2021-hetformer,
title = "{HETFORMER}: Heterogeneous Transformer with Sparse Attention for Long-Text Extractive Summarization",
author = "Liu, Ye and
Zhang, Jianguo and
Wan, Yao and
Xia, Congying and
He, Lifang and
Yu, Philip",
booktitle = "Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing",
month = nov,
year = "2021",
address = "Online and Punta Cana, Dominican Republic",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.emnlp-main.13",
doi = "10.18653/v1/2021.emnlp-main.13",
pages = "146--154",
abstract = "To capture the semantic graph structure from raw text, most existing summarization approaches are built on GNNs with a pre-trained model. However, these methods suffer from cumbersome procedures and inefficient computations for long-text documents. To mitigate these issues, this paper proposes HetFormer, a Transformer-based pre-trained model with multi-granularity sparse attentions for long-text extractive summarization. Specifically, we model different types of semantic nodes in raw text as a potential heterogeneous graph and directly learn heterogeneous relationships (edges) among nodes by Transformer. Extensive experiments on both single- and multi-document summarization tasks show that HetFormer achieves state-of-the-art performance in Rouge F1 while using less memory and fewer parameters.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="liu-etal-2021-hetformer">
<titleInfo>
<title>HETFORMER: Heterogeneous Transformer with Sparse Attention for Long-Text Extractive Summarization</title>
</titleInfo>
<name type="personal">
<namePart type="given">Ye</namePart>
<namePart type="family">Liu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jianguo</namePart>
<namePart type="family">Zhang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yao</namePart>
<namePart type="family">Wan</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Congying</namePart>
<namePart type="family">Xia</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Lifang</namePart>
<namePart type="family">He</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Philip</namePart>
<namePart type="family">Yu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2021-nov</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing</title>
</titleInfo>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Online and Punta Cana, Dominican Republic</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>To capture the semantic graph structure from raw text, most existing summarization approaches are built on GNNs with a pre-trained model. However, these methods suffer from cumbersome procedures and inefficient computations for long-text documents. To mitigate these issues, this paper proposes HetFormer, a Transformer-based pre-trained model with multi-granularity sparse attentions for long-text extractive summarization. Specifically, we model different types of semantic nodes in raw text as a potential heterogeneous graph and directly learn heterogeneous relationships (edges) among nodes by Transformer. Extensive experiments on both single- and multi-document summarization tasks show that HetFormer achieves state-of-the-art performance in Rouge F1 while using less memory and fewer parameters.</abstract>
<identifier type="citekey">liu-etal-2021-hetformer</identifier>
<identifier type="doi">10.18653/v1/2021.emnlp-main.13</identifier>
<location>
<url>https://aclanthology.org/2021.emnlp-main.13</url>
</location>
<part>
<date>2021-nov</date>
<extent unit="page">
<start>146</start>
<end>154</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T HETFORMER: Heterogeneous Transformer with Sparse Attention for Long-Text Extractive Summarization
%A Liu, Ye
%A Zhang, Jianguo
%A Wan, Yao
%A Xia, Congying
%A He, Lifang
%A Yu, Philip
%S Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing
%D 2021
%8 nov
%I Association for Computational Linguistics
%C Online and Punta Cana, Dominican Republic
%F liu-etal-2021-hetformer
%X To capture the semantic graph structure from raw text, most existing summarization approaches are built on GNNs with a pre-trained model. However, these methods suffer from cumbersome procedures and inefficient computations for long-text documents. To mitigate these issues, this paper proposes HetFormer, a Transformer-based pre-trained model with multi-granularity sparse attentions for long-text extractive summarization. Specifically, we model different types of semantic nodes in raw text as a potential heterogeneous graph and directly learn heterogeneous relationships (edges) among nodes by Transformer. Extensive experiments on both single- and multi-document summarization tasks show that HetFormer achieves state-of-the-art performance in Rouge F1 while using less memory and fewer parameters.
%R 10.18653/v1/2021.emnlp-main.13
%U https://aclanthology.org/2021.emnlp-main.13
%U https://doi.org/10.18653/v1/2021.emnlp-main.13
%P 146-154
Markdown (Informal)
[HETFORMER: Heterogeneous Transformer with Sparse Attention for Long-Text Extractive Summarization](https://aclanthology.org/2021.emnlp-main.13) (Liu et al., EMNLP 2021)
ACL