[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-1-en-105":3,"doc-seo-564092-105":53,"doc-detail-564092-en":126},{"code":4,"msg":5,"data":6},0,"success",[7,14,19,24,29,34,39,44,49],{"id":8,"doc_module":9,"doc_module_name":10,"category_name":11,"show_sort_weight":12,"slug":13},11,1,"Template","Presentations",90,"presentations",{"id":15,"doc_module":9,"doc_module_name":10,"category_name":16,"show_sort_weight":17,"slug":18},12,"Resumes",80,"resumes",{"id":20,"doc_module":9,"doc_module_name":10,"category_name":21,"show_sort_weight":22,"slug":23},14,"Invoices",70,"invoices",{"id":25,"doc_module":9,"doc_module_name":10,"category_name":26,"show_sort_weight":27,"slug":28},15,"Posters",60,"posters",{"id":30,"doc_module":9,"doc_module_name":10,"category_name":31,"show_sort_weight":32,"slug":33},16,"Social Media",50,"social-media",{"id":35,"doc_module":9,"doc_module_name":10,"category_name":36,"show_sort_weight":37,"slug":38},17,"Forms",40,"forms",{"id":40,"doc_module":9,"doc_module_name":10,"category_name":41,"show_sort_weight":42,"slug":43},18,"Letters",30,"letters",{"id":45,"doc_module":9,"doc_module_name":10,"category_name":46,"show_sort_weight":47,"slug":48},21,"Paper Templates",5,"papers-templates",{"id":50,"doc_module":9,"doc_module_name":10,"category_name":51,"show_sort_weight":4,"slug":52},158,"General","general-158",{"code":4,"msg":54,"data":55},"ok",{"site_id":56,"language":57,"slug":58,"title":59,"keywords":60,"description":61,"schema_data":62,"social_meta":119,"head_meta":121,"extra_data":123,"updated_unix":125},105,"en","jsynflow-japanese-synthesised-flowchart-visual-question-answering-dataset-abstract","JSynFlow: Japanese Synthesised Flowchart Visual Question Answering Dataset - Abstract","","Vision-and-language models are expected to answer questions about complex documents such as flowcharts, yet reliable understanding depends on large-scale datasets of flowchart images and aligned text. JSynFlow addresses this by introducing a synthesized Japanese visual QA dataset generated with large language models. The dataset includes business-occupation task descriptions, flowchart images rendered from domain-specific language code, and corresponding question–answer pairs in JSON format. The paper also describes the synthesis procedure and shows fine-tuning with JSynFlow improves VLM performance on flowchart-based QA tasks.",{"@graph":63,"@context":118},[64,80,101],{"@type":65,"itemListElement":66},"BreadcrumbList",[67,71,74,77],{"item":68,"name":69,"@type":70,"position":9},"https://docshare.wps.com","Home","ListItem",{"item":72,"name":10,"@type":70,"position":73},"https://docshare.wps.com/template/",2,{"item":75,"name":11,"@type":70,"position":76},"https://docshare.wps.com/template/presentations/",3,{"item":78,"name":59,"@type":70,"position":79},"https://docshare.wps.com/template/jsynflow-japanese-synthesised-flowchart-visual-question-answering-dataset-abstract/564092/",4,{"url":78,"name":59,"@type":81,"image":82,"author":87,"headline":59,"publisher":90,"fileFormat":93,"inLanguage":57,"description":61,"dateModified":94,"datePublished":95,"encodingFormat":93,"isAccessibleForFree":96,"interactionStatistic":97},"DigitalDocument",{"url":83,"@type":84,"width":85,"height":86},"https://docshare.wps.com/thumbnails/jsynflow-japanese-synthesised-flowchart-visual-question-answering-dataset-abstract/564092.png","ImageObject",442,249,{"name":88,"@type":89},"adawds","Person",{"url":68,"name":91,"@type":92},"DocShare","Organization","application/pdf","2026-10-26","2026-10-12",true,{"@type":98,"interactionType":99,"userInteractionCount":79},"InteractionCounter",{"@type":100},"ViewAction",{"@type":102,"mainEntity":103},"FAQPage",[104,110,114],{"name":105,"@type":106,"acceptedAnswer":107},"What problem does JSynFlow aim to solve for flowchart VQA?","Question",{"text":108,"@type":109},"Accurate flowchart understanding by vision-language models requires large-scale, high-quality datasets of flowchart images aligned with text, which is time-consuming to create. JSynFlow synthesizes such aligned data using large language models.","Answer",{"name":111,"@type":106,"acceptedAnswer":112},"What does the JSynFlow dataset contain?",{"text":113,"@type":109},"It provides task descriptions for business occupations, flowchart images rendered from domain-specific language (DSL) code, and related question–answer pairs formatted in JSON.",{"name":115,"@type":106,"acceptedAnswer":116},"How is JSynFlow expected to improve model performance?",{"text":117,"@type":109},"Fine-tuning vision-language models with JSynFlow is shown to significantly improve performance on flowchart-based question answering tasks.","https://schema.org",{"og:url":78,"og:type":120,"og:title":59,"og:site_name":91,"og:description":61},"article",{"robots":122,"canonical":78},"index,follow",{"doc_id":124,"site_id":56},564092,1792366222,{"code":4,"msg":5,"data":127},{"doc_id":124,"user_id":128,"nickname":88,"user_avatar":129,"doc_module":9,"category_id":8,"category_name":11,"doc_title":59,"doc_description":61,"doc_content":130,"file_id":131,"file_url":132,"file_type":133,"file_size":134,"view_count":79,"is_deleted":4,"is_public":9,"is_downloadable":9,"audit_status":9,"page_count":135,"language":136,"language_code":57,"site_id":56,"html_lang":57,"table_of_contents":137,"faqs":138,"seo_title":139,"seo_description":61,"update_tm":140,"read_time":76},3985747872983,"https://ap-avatar.wpscdn.com/davatar_994ba38a5ba835b3df7d355c54d3ed8d","JSynFlow:  \nJapanese Synthesised Flowchart Visual Question Answering Dataset  \nbuilt with Large Language Models  \nHiroshi Sasaki / The Japan Research Institute, [Limited](Limited / sasaki.hiroshi@jri.co.jp)[ /](Limited / sasaki.hiroshi@jri.co.jp)[ sasaki.hiroshi@jri.co.jp](Limited / sasaki.hiroshi@jri.co.jp)  \narXiv :2602 .04 142v2 [ cs .CV] 5 Feb 2026  \nAbstract—Vision and language models (VLMs) are expected to analyse complex documents, such as those containing flowcharts, through a question-answering (QA) interface. The ability to recognise and interpret these flowcharts is in high demand, as they provide valuable insights unavailable in text-only explanations. However, developing VLMs with precise flowchart understanding requires large-scale datasets of flowchart images and corresponding text, the creation of which is highly time-consuming. To address this challenge, we introduce “JSynFlow”, a synthesised visual QA dataset for Japanese flowcharts, generated using large language models (LLMs) . Our dataset comprises task descriptions for various business occupations, the corresponding flowchart images rendered from domainspecific language (DSL) code, and related QA pairs. This paper details the dataset’s synthesis procedure and demonstrates that fine-tuning with JSynFlow significantly improves VLM performance on flowchartbased QA tasks. Our dataset is publicly available at [https://huggingface.co/datasets/jri-advtechlab/jsynfl](https://huggingface.co/datasets/jri-advtechlab/jsynfl)ow.  \nIndex Terms—Dataset, Vision and Language Models, Large Language Models, Visual Question and Answering, Flowchart.  \nI. Introduction  \nIn recent years, vision-and-language models (VLMs), a class of large language models (LLMs) endowed with the capability to comprehend and interpret visual information, have witnessed rapid advancements. These models have enabled many advanced computer vision (CV) tasks such as visual content description through a simple question-and-answer (QA) style. Such CV tasks include the comprehension of flowchart images, which serve as visual representations of complex procedures and branching conditions. Flowcharts are extensively employed for business purposes, including process understanding, knowledge sharing, and workflow optimisation, making the ability of VLMs to recognise them in high demand. Nevertheless, substantial challenges persist in the development of VLMs  \nThis article is a revised and translated version of a paper originally published in Japanese [12] .  \nNotice for the use of this material. The copyright of this material is retained by the Japanese Society for Artificial Intelligence (JSAI) . This material is published here with the agreement of JSAI. Please be complied with Copyright Law of Japan if any users wish to reproduce, make derivative work, distribute or make available to the public any part or whole thereof. All Rights Reserved, Copyright (C) The Japanese Society for Artificial Intelligence.  \nTABLE I: Statistics of the JSynFlow dataset  \n\n|  | Training | Test | Total |\n| --- | --- | --- | --- |\n| Industries\u003Cbr>Occupations\u003Cbr>Tasks\u003Cbr>QAs | 1,359\u003Cbr>10,072 | 152 1,065 | 9\u003Cbr>115 1,511\u003Cbr>11,137 |\n\ncapable of accurately interpreting the fine-grained visual details indispensable for the comprehension of flowcharts. Moreover, the acquisition of sufficiently large volumes of high-quality data, such as text-image aligned pairs for VLM training, remains a significant issue [4] .  \nThis paper introduces a Japanese flowchart dataset tailored to the visual question answering (VQA) task, constructed by leveraging LLMs (Fig. 1) . We detail both the dataset itself and its construction methodology. Furthermore, to demonstrate its practical application, we present experimental results from fine-tuning VLMs with the dataset, illustrating its positive impact on QA performance.  \nII. Related Work  \nDomain-specific languages (DSLs) enable LLMs to produce text-based diagram definitions and generate the corresp","cbCaioPyomjrgqb4","https://ap.wps.com/l/cbCaioPyomjrgqb4","pdf",469852,7,"English","# Introduction\n# Related Work\n# JSynFlow Dataset\n## Overview\n## Generation Procedure","[{\"question\":\"What problem does JSynFlow aim to solve for flowchart VQA?\",\"answer\":\"Accurate flowchart understanding by vision-language models requires large-scale, high-quality datasets of flowchart images aligned with text, which is time-consuming to create. JSynFlow synthesizes such aligned data using large language models.\"},{\"question\":\"What does the JSynFlow dataset contain?\",\"answer\":\"It provides task descriptions for business occupations, flowchart images rendered from domain-specific language (DSL) code, and related question–answer pairs formatted in JSON.\"},{\"question\":\"How is JSynFlow expected to improve model performance?\",\"answer\":\"Fine-tuning vision-language models with JSynFlow is shown to significantly improve performance on flowchart-based question answering tasks.\"}]","JSynFlow: Japanese Synthesised Flowchart Visual Question Answering Dataset - Abstract | PDF",1791840223]