@inproceedings{shin-etal-2026-planning,
title = "Planning Before Reasoning: A Vision-Language Dataset for Structured Stepwise Reasoning in Document Understanding",
author = "Shin, Dongjae and
Yoo, Hangyeol and
Nam, Yerin and
Shin, Hoyeon and
Song, SeungWoo and
Song, Seohyun and
Yang, Donghun and
Lim, KyungTae",
editor = "Lee, Hung-Yi and
Dev, Sunipa and
Li, Jessy and
Elliott, Desmond",
booktitle = "Findings of the {A}ssociation for {C}omputational {L}inguistics: {EMNLP} 2026",
month = oct,
year = "2026",
address = "Budapest, Hungary",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2026.findings-emnlp.996/",
doi = "10.18653/2026.findings-emnlp.996",
pages = "21022--21061",
ISBN = "979-8-89176-462-0",
abstract = "This paper introduces DocPlan, a vision-language training dataset designed to emulate the stepwise reasoning strategies employed by human during visual document understanding. DocPlan was constructed to enable vision-language models (VLMs) to effectively learn the structured reasoning process that humans typically follow when interpreting documents. We reorganized existing document understanding datasets into 133K examples annotated with the following four sequential stages: Recognition {\textrightarrow} Planning {\textrightarrow} Extraction and Reasoning {\textrightarrow} Conclusion. To evaluate the effectiveness of DocPlan, we fine-tuned the direct prediction-based VLMs to perform stepwise reasoning generation across these stages. The experimental results demonstrated that our approach achieved up to a 34.4{\%} improvement in accuracy on document-based benchmarks."
}<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="shin-etal-2026-planning">
<titleInfo>
<title>Planning Before Reasoning: A Vision-Language Dataset for Structured Stepwise Reasoning in Document Understanding</title>
</titleInfo>
<name type="personal">
<namePart type="given">Dongjae</namePart>
<namePart type="family">Shin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hangyeol</namePart>
<namePart type="family">Yoo</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yerin</namePart>
<namePart type="family">Nam</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Hoyeon</namePart>
<namePart type="family">Shin</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">SeungWoo</namePart>
<namePart type="family">Song</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Seohyun</namePart>
<namePart type="family">Song</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Donghun</namePart>
<namePart type="family">Yang</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">KyungTae</namePart>
<namePart type="family">Lim</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2026-10</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Findings of the Association for Computational Linguistics: EMNLP 2026</title>
</titleInfo>
<name type="personal">
<namePart type="given">Hung-Yi</namePart>
<namePart type="family">Lee</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Sunipa</namePart>
<namePart type="family">Dev</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Jessy</namePart>
<namePart type="family">Li</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Desmond</namePart>
<namePart type="family">Elliott</namePart>
<role>
<roleTerm authority="marcrelator" type="text">editor</roleTerm>
</role>
</name>
<originInfo>
<publisher>Association for Computational Linguistics</publisher>
<place>
<placeTerm type="text">Budapest, Hungary</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
<identifier type="isbn">979-8-89176-462-0</identifier>
</relatedItem>
<abstract>This paper introduces DocPlan, a vision-language training dataset designed to emulate the stepwise reasoning strategies employed by human during visual document understanding. DocPlan was constructed to enable vision-language models (VLMs) to effectively learn the structured reasoning process that humans typically follow when interpreting documents. We reorganized existing document understanding datasets into 133K examples annotated with the following four sequential stages: Recognition → Planning → Extraction and Reasoning → Conclusion. To evaluate the effectiveness of DocPlan, we fine-tuned the direct prediction-based VLMs to perform stepwise reasoning generation across these stages. The experimental results demonstrated that our approach achieved up to a 34.4% improvement in accuracy on document-based benchmarks.</abstract>
<identifier type="citekey">shin-etal-2026-planning</identifier>
<identifier type="doi">10.18653/2026.findings-emnlp.996</identifier>
<location>
<url>https://aclanthology.org/2026.findings-emnlp.996/</url>
</location>
<part>
<date>2026-10</date>
<extent unit="page">
<start>21022</start>
<end>21061</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T Planning Before Reasoning: A Vision-Language Dataset for Structured Stepwise Reasoning in Document Understanding
%A Shin, Dongjae
%A Yoo, Hangyeol
%A Nam, Yerin
%A Shin, Hoyeon
%A Song, SeungWoo
%A Song, Seohyun
%A Yang, Donghun
%A Lim, KyungTae
%Y Lee, Hung-Yi
%Y Dev, Sunipa
%Y Li, Jessy
%Y Elliott, Desmond
%S Findings of the Association for Computational Linguistics: EMNLP 2026
%D 2026
%8 October
%I Association for Computational Linguistics
%C Budapest, Hungary
%@ 979-8-89176-462-0
%F shin-etal-2026-planning
%X This paper introduces DocPlan, a vision-language training dataset designed to emulate the stepwise reasoning strategies employed by human during visual document understanding. DocPlan was constructed to enable vision-language models (VLMs) to effectively learn the structured reasoning process that humans typically follow when interpreting documents. We reorganized existing document understanding datasets into 133K examples annotated with the following four sequential stages: Recognition → Planning → Extraction and Reasoning → Conclusion. To evaluate the effectiveness of DocPlan, we fine-tuned the direct prediction-based VLMs to perform stepwise reasoning generation across these stages. The experimental results demonstrated that our approach achieved up to a 34.4% improvement in accuracy on document-based benchmarks.
%R 10.18653/2026.findings-emnlp.996
%U https://aclanthology.org/2026.findings-emnlp.996/
%U https://doi.org/10.18653/2026.findings-emnlp.996
%P 21022-21061
Markdown (Informal)
[Planning Before Reasoning: A Vision-Language Dataset for Structured Stepwise Reasoning in Document Understanding](https://aclanthology.org/2026.findings-emnlp.996/) (Shin et al., Findings 2026)
ACL
- Dongjae Shin, Hangyeol Yoo, Yerin Nam, Hoyeon Shin, SeungWoo Song, Seohyun Song, Donghun Yang, and KyungTae Lim. 2026. Planning Before Reasoning: A Vision-Language Dataset for Structured Stepwise Reasoning in Document Understanding. In Findings of the Association for Computational Linguistics: EMNLP 2026, pages 21022–21061, Budapest, Hungary. Association for Computational Linguistics.