PEneo: Unifying Line Extraction, Line Grouping, and Entity Linking for End-to-end Document Pair Extraction

Document pair extraction aims to identify key and value entities as well as their relationships from visually-rich documents. Most existing methods divide it into two separate tasks: semantic entity recognition (SER) and relation extraction (RE). However, simply concatenating SER and RE serially can...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Lin, Zening, Wang, Jiapeng, Li, Teng, Liao, Wenhui, Huang, Dayi, Xiong, Longfei, Jin, Lianwen
Format: Artikel
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page
container_issue
container_start_page
container_title
container_volume
creator Lin, Zening
Wang, Jiapeng
Li, Teng
Liao, Wenhui
Huang, Dayi
Xiong, Longfei
Jin, Lianwen
description Document pair extraction aims to identify key and value entities as well as their relationships from visually-rich documents. Most existing methods divide it into two separate tasks: semantic entity recognition (SER) and relation extraction (RE). However, simply concatenating SER and RE serially can lead to severe error propagation, and it fails to handle cases like multi-line entities in real scenarios. To address these issues, this paper introduces a novel framework, PEneo (Pair Extraction new decoder option), which performs document pair extraction in a unified pipeline, incorporating three concurrent sub-tasks: line extraction, line grouping, and entity linking. This approach alleviates the error accumulation problem and can handle the case of multi-line entities. Furthermore, to better evaluate the model's performance and to facilitate future research on pair extraction, we introduce RFUND, a re-annotated version of the commonly used FUNSD and XFUND datasets, to make them more accurate and cover realistic situations. Experiments on various benchmarks demonstrate PEneo's superiority over previous pipelines, boosting the performance by a large margin (e.g., 19.89%-22.91% F1 score on RFUND-EN) when combined with various backbones like LiLT and LayoutLMv3, showing its effectiveness and generality. Codes and the new annotations are available at https://github.com/ZeningLin/PEneo.
doi_str_mv 10.48550/arxiv.2401.03472
format Article
fullrecord <record><control><sourceid>arxiv_GOX</sourceid><recordid>TN_cdi_arxiv_primary_2401_03472</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>2401_03472</sourcerecordid><originalsourceid>FETCH-LOGICAL-a672-e19e841fc6863fc5ad97b78089e1cbc05e7db593b3f27e49bf1e615504b8e7393</originalsourceid><addsrcrecordid>eNpNj7FOwzAYhL0woMIDMOEHaIIdO7HNhkooSJHoUObIdn4jA7Ur46Lm7UlaBqaT7tOd9CF0Q0nJZV2TO52O_qesOKElYVxUl-hj0waI9_gteDf68I47HwC3x5y0zT6G5blYp3jYT3iJdRhwG7LP40w-54mLaaqGIscCJvoY7WEHIeON9unf1RW6cPrrG67_coG2T-129Vx0r-uX1UNX6EZUBVAFklNnG9kwZ2s9KGGEJFIBtcaSGsRgasUMc5UAroyj0NBJjhsJgim2QLfn25Nsv09-p9PYz9L9SZr9ArB9Ugo</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>PEneo: Unifying Line Extraction, Line Grouping, and Entity Linking for End-to-end Document Pair Extraction</title><source>arXiv.org</source><creator>Lin, Zening ; Wang, Jiapeng ; Li, Teng ; Liao, Wenhui ; Huang, Dayi ; Xiong, Longfei ; Jin, Lianwen</creator><creatorcontrib>Lin, Zening ; Wang, Jiapeng ; Li, Teng ; Liao, Wenhui ; Huang, Dayi ; Xiong, Longfei ; Jin, Lianwen</creatorcontrib><description>Document pair extraction aims to identify key and value entities as well as their relationships from visually-rich documents. Most existing methods divide it into two separate tasks: semantic entity recognition (SER) and relation extraction (RE). However, simply concatenating SER and RE serially can lead to severe error propagation, and it fails to handle cases like multi-line entities in real scenarios. To address these issues, this paper introduces a novel framework, PEneo (Pair Extraction new decoder option), which performs document pair extraction in a unified pipeline, incorporating three concurrent sub-tasks: line extraction, line grouping, and entity linking. This approach alleviates the error accumulation problem and can handle the case of multi-line entities. Furthermore, to better evaluate the model's performance and to facilitate future research on pair extraction, we introduce RFUND, a re-annotated version of the commonly used FUNSD and XFUND datasets, to make them more accurate and cover realistic situations. Experiments on various benchmarks demonstrate PEneo's superiority over previous pipelines, boosting the performance by a large margin (e.g., 19.89%-22.91% F1 score on RFUND-EN) when combined with various backbones like LiLT and LayoutLMv3, showing its effectiveness and generality. Codes and the new annotations are available at https://github.com/ZeningLin/PEneo.</description><identifier>DOI: 10.48550/arxiv.2401.03472</identifier><language>eng</language><subject>Computer Science - Computation and Language</subject><creationdate>2024-01</creationdate><rights>http://creativecommons.org/licenses/by-nc-sa/4.0</rights><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><link.rule.ids>228,230,776,881</link.rule.ids><linktorsrc>$$Uhttps://arxiv.org/abs/2401.03472$$EView_record_in_Cornell_University$$FView_record_in_$$GCornell_University$$Hfree_for_read</linktorsrc><backlink>$$Uhttps://doi.org/10.48550/arXiv.2401.03472$$DView paper in arXiv$$Hfree_for_read</backlink></links><search><creatorcontrib>Lin, Zening</creatorcontrib><creatorcontrib>Wang, Jiapeng</creatorcontrib><creatorcontrib>Li, Teng</creatorcontrib><creatorcontrib>Liao, Wenhui</creatorcontrib><creatorcontrib>Huang, Dayi</creatorcontrib><creatorcontrib>Xiong, Longfei</creatorcontrib><creatorcontrib>Jin, Lianwen</creatorcontrib><title>PEneo: Unifying Line Extraction, Line Grouping, and Entity Linking for End-to-end Document Pair Extraction</title><description>Document pair extraction aims to identify key and value entities as well as their relationships from visually-rich documents. Most existing methods divide it into two separate tasks: semantic entity recognition (SER) and relation extraction (RE). However, simply concatenating SER and RE serially can lead to severe error propagation, and it fails to handle cases like multi-line entities in real scenarios. To address these issues, this paper introduces a novel framework, PEneo (Pair Extraction new decoder option), which performs document pair extraction in a unified pipeline, incorporating three concurrent sub-tasks: line extraction, line grouping, and entity linking. This approach alleviates the error accumulation problem and can handle the case of multi-line entities. Furthermore, to better evaluate the model's performance and to facilitate future research on pair extraction, we introduce RFUND, a re-annotated version of the commonly used FUNSD and XFUND datasets, to make them more accurate and cover realistic situations. Experiments on various benchmarks demonstrate PEneo's superiority over previous pipelines, boosting the performance by a large margin (e.g., 19.89%-22.91% F1 score on RFUND-EN) when combined with various backbones like LiLT and LayoutLMv3, showing its effectiveness and generality. Codes and the new annotations are available at https://github.com/ZeningLin/PEneo.</description><subject>Computer Science - Computation and Language</subject><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2024</creationdate><recordtype>article</recordtype><sourceid>GOX</sourceid><recordid>eNpNj7FOwzAYhL0woMIDMOEHaIIdO7HNhkooSJHoUObIdn4jA7Ur46Lm7UlaBqaT7tOd9CF0Q0nJZV2TO52O_qesOKElYVxUl-hj0waI9_gteDf68I47HwC3x5y0zT6G5blYp3jYT3iJdRhwG7LP40w-54mLaaqGIscCJvoY7WEHIeON9unf1RW6cPrrG67_coG2T-129Vx0r-uX1UNX6EZUBVAFklNnG9kwZ2s9KGGEJFIBtcaSGsRgasUMc5UAroyj0NBJjhsJgim2QLfn25Nsv09-p9PYz9L9SZr9ArB9Ugo</recordid><startdate>20240107</startdate><enddate>20240107</enddate><creator>Lin, Zening</creator><creator>Wang, Jiapeng</creator><creator>Li, Teng</creator><creator>Liao, Wenhui</creator><creator>Huang, Dayi</creator><creator>Xiong, Longfei</creator><creator>Jin, Lianwen</creator><scope>AKY</scope><scope>GOX</scope></search><sort><creationdate>20240107</creationdate><title>PEneo: Unifying Line Extraction, Line Grouping, and Entity Linking for End-to-end Document Pair Extraction</title><author>Lin, Zening ; Wang, Jiapeng ; Li, Teng ; Liao, Wenhui ; Huang, Dayi ; Xiong, Longfei ; Jin, Lianwen</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-a672-e19e841fc6863fc5ad97b78089e1cbc05e7db593b3f27e49bf1e615504b8e7393</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2024</creationdate><topic>Computer Science - Computation and Language</topic><toplevel>online_resources</toplevel><creatorcontrib>Lin, Zening</creatorcontrib><creatorcontrib>Wang, Jiapeng</creatorcontrib><creatorcontrib>Li, Teng</creatorcontrib><creatorcontrib>Liao, Wenhui</creatorcontrib><creatorcontrib>Huang, Dayi</creatorcontrib><creatorcontrib>Xiong, Longfei</creatorcontrib><creatorcontrib>Jin, Lianwen</creatorcontrib><collection>arXiv Computer Science</collection><collection>arXiv.org</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Lin, Zening</au><au>Wang, Jiapeng</au><au>Li, Teng</au><au>Liao, Wenhui</au><au>Huang, Dayi</au><au>Xiong, Longfei</au><au>Jin, Lianwen</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>PEneo: Unifying Line Extraction, Line Grouping, and Entity Linking for End-to-end Document Pair Extraction</atitle><date>2024-01-07</date><risdate>2024</risdate><abstract>Document pair extraction aims to identify key and value entities as well as their relationships from visually-rich documents. Most existing methods divide it into two separate tasks: semantic entity recognition (SER) and relation extraction (RE). However, simply concatenating SER and RE serially can lead to severe error propagation, and it fails to handle cases like multi-line entities in real scenarios. To address these issues, this paper introduces a novel framework, PEneo (Pair Extraction new decoder option), which performs document pair extraction in a unified pipeline, incorporating three concurrent sub-tasks: line extraction, line grouping, and entity linking. This approach alleviates the error accumulation problem and can handle the case of multi-line entities. Furthermore, to better evaluate the model's performance and to facilitate future research on pair extraction, we introduce RFUND, a re-annotated version of the commonly used FUNSD and XFUND datasets, to make them more accurate and cover realistic situations. Experiments on various benchmarks demonstrate PEneo's superiority over previous pipelines, boosting the performance by a large margin (e.g., 19.89%-22.91% F1 score on RFUND-EN) when combined with various backbones like LiLT and LayoutLMv3, showing its effectiveness and generality. Codes and the new annotations are available at https://github.com/ZeningLin/PEneo.</abstract><doi>10.48550/arxiv.2401.03472</doi><oa>free_for_read</oa></addata></record>
fulltext fulltext_linktorsrc
identifier DOI: 10.48550/arxiv.2401.03472
ispartof
issn
language eng
recordid cdi_arxiv_primary_2401_03472
source arXiv.org
subjects Computer Science - Computation and Language
title PEneo: Unifying Line Extraction, Line Grouping, and Entity Linking for End-to-end Document Pair Extraction
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-19T10%3A50%3A42IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-arxiv_GOX&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=PEneo:%20Unifying%20Line%20Extraction,%20Line%20Grouping,%20and%20Entity%20Linking%20for%20End-to-end%20Document%20Pair%20Extraction&rft.au=Lin,%20Zening&rft.date=2024-01-07&rft_id=info:doi/10.48550/arxiv.2401.03472&rft_dat=%3Carxiv_GOX%3E2401_03472%3C/arxiv_GOX%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true