[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"detail-sidebar-cat-0-en-105":3,"doc-seo-273868-105":59,"doc-detail-273868-en":131},{"code":4,"msg":5,"data":6},0,"success",[7,13,18,23,28,33,38,43,48,51,55],{"id":8,"doc_module":4,"doc_module_name":9,"category_name":10,"show_sort_weight":11,"slug":12},1,"Document","Story & Novel",90,"story-novel",{"id":14,"doc_module":4,"doc_module_name":9,"category_name":15,"show_sort_weight":16,"slug":17},2,"Literature",80,"literature",{"id":19,"doc_module":4,"doc_module_name":9,"category_name":20,"show_sort_weight":21,"slug":22},4,"Exam",70,"exam",{"id":24,"doc_module":4,"doc_module_name":9,"category_name":25,"show_sort_weight":26,"slug":27},5,"Comic",60,"comic",{"id":29,"doc_module":4,"doc_module_name":9,"category_name":30,"show_sort_weight":31,"slug":32},6,"Technology",50,"technology",{"id":34,"doc_module":4,"doc_module_name":9,"category_name":35,"show_sort_weight":36,"slug":37},7,"Healthcare",40,"healthcare",{"id":39,"doc_module":4,"doc_module_name":9,"category_name":40,"show_sort_weight":41,"slug":42},8,"Research & Report",30,"research-report",{"id":44,"doc_module":4,"doc_module_name":9,"category_name":45,"show_sort_weight":46,"slug":47},9,"Religion & Spirituality",20,"religion-spirituality",{"id":46,"doc_module":4,"doc_module_name":9,"category_name":49,"show_sort_weight":46,"slug":50},"World Cup","world-cup",{"id":52,"doc_module":4,"doc_module_name":9,"category_name":53,"show_sort_weight":52,"slug":54},10,"Lifestyle","lifestyle",{"id":56,"doc_module":4,"doc_module_name":9,"category_name":57,"show_sort_weight":24,"slug":58},19,"General","general",{"code":4,"msg":60,"data":61},"ok",{"site_id":62,"language":63,"slug":64,"title":65,"keywords":66,"description":67,"schema_data":68,"social_meta":124,"head_meta":126,"extra_data":128,"updated_unix":130},105,"en","ensi-rag-entity-structure-indexed-retrieval-augmented-generation-for-long-document-question-answering","EnSI-RAG: Entity-Structure-Indexed Retrieval-Augmented Generation for Long-Document Question Answering","","Long-document question answering is difficult when relevant evidence is dispersed across multiple entities and their relationships. Common retrieval-augmented generation (RAG) methods index fixed raw chunks and retrieve by embedding similarity, causing performance drops when entity boundaries split evidence or when questions need multi-hop reasoning. EnSI-RAG builds a query-independent, entity-centered index, using entity records ⟨e,t,k,v⟩ linked to original passages, separating evidence localization from answer synthesis. On Loong and Oolong it reaches 78.24 average accuracy, 6.62 points above references, with traceable sources.",{"@graph":69,"@context":123},[70,84,106],{"@type":71,"itemListElement":72},"BreadcrumbList",[73,77,79,82],{"item":74,"name":75,"@type":76,"position":8},"https://docshare.wps.com","Home","ListItem",{"item":78,"name":9,"@type":76,"position":14},"https://docshare.wps.com/document/",{"item":80,"name":40,"@type":76,"position":81},"https://docshare.wps.com/document/research-report/",3,{"item":83,"name":65,"@type":76,"position":19},"https://docshare.wps.com/document/ensi-rag-entity-structure-indexed-retrieval-augmented-generation-for-long-document-question-answering/273868/",{"url":83,"name":65,"@type":85,"image":86,"author":91,"headline":65,"publisher":94,"fileFormat":97,"inLanguage":63,"description":67,"dateModified":98,"datePublished":99,"encodingFormat":97,"isAccessibleForFree":100,"interactionStatistic":101},"DigitalDocument",{"url":87,"@type":88,"width":89,"height":90},"https://docshare.wps.com/thumbnails/ensi-rag-entity-structure-indexed-retrieval-augmented-generation-for-long-document-question-answering/273868.png","ImageObject",300,407,{"name":92,"@type":93},"Ophelia","Person",{"url":74,"name":95,"@type":96},"DocShare","Organization","application/pdf","2026-10-30","2026-09-15",true,{"@type":102,"interactionType":103,"userInteractionCount":105},"InteractionCounter",{"@type":104},"ViewAction",26,{"@type":107,"mainEntity":108},"FAQPage",[109,115,119],{"name":110,"@type":111,"acceptedAnswer":112},"Why do existing RAG methods struggle with long-document question answering?","Question",{"text":113,"@type":114},"They often index fixed-length chunks, so entity evidence may be split by chunk boundaries or blended with unrelated topics, leading to incomplete evidence retrieval for multi-hop questions.","Answer",{"name":116,"@type":111,"acceptedAnswer":117},"What is EnSI-RAG’s main contribution to improve retrieval?",{"text":118,"@type":114},"EnSI-RAG constructs entity-centered passages and a query-independent entity-structure index with retrieval handles linked back to original source passages.",{"name":120,"@type":111,"acceptedAnswer":121},"How does EnSI-RAG connect structured records to final answers?",{"text":122,"@type":114},"At query time, the index guides retrieval of supporting passages, and an LLM synthesizes the retrieved evidence into the final response while preserving traceable sources.","https://schema.org",{"og:url":83,"og:type":125,"og:title":65,"og:site_name":95,"og:description":67},"article",{"robots":127,"canonical":83},"index,follow",{"doc_id":129,"site_id":62},273868,1789473076,{"code":4,"msg":5,"data":132},{"doc_id":129,"user_id":133,"nickname":92,"user_avatar":134,"doc_module":4,"category_id":39,"category_name":40,"doc_title":65,"doc_description":67,"doc_content":135,"file_id":136,"file_url":137,"file_type":138,"file_size":139,"view_count":105,"is_deleted":4,"is_public":8,"is_downloadable":8,"audit_status":8,"page_count":140,"language":141,"language_code":63,"site_id":62,"html_lang":63,"table_of_contents":142,"faqs":143,"seo_title":144,"seo_description":67,"update_tm":130,"read_time":145,"preview_page_count":4},7971461741311,"https://ap-avatar.wpscdn.com/avatar/74000253aff267980c6?x-image-process=image/resize,m_fixed,w_180,h_180&k=1779345379180704826","EnSI-RAG: Entity-Structure-Indexed Retrieval-Augmented Generation for  \nLong-Document Question Answering  \nXuanyu Meng, Jiashuo Sun, Jash Rajesh Parekh, Jiawei Han  \nUniversity of Illinois Urbana-Champaign  \nUrbana, Illinois, USA  \n{xuanyum2, jiashuo5, jashrp2, [hanj}@illinois.edu](hanj}@illinois.edu)  \narXiv : 2608 .2 1252v 1 [cs .CL] 2 1 Aug 2026  \nAbstract  \nQuestion answering (QA) over long, connected documents remains challenging because relevant evidence may span multiple entities and their relationships. Existing retrievalaugmented generation (RAG) methods typically index documents as raw chunks and retrieve them through embedding similarity. Their performance degrades when chunk boundaries separate entities from supporting evidence or when a question requires multi-hop reasoning across the corpus. We propose EnSI-RAG (Entity-Structure-Indexed RetrievalAugmented Generation), a framework that constructs a queryindependent, entity-centered index. Each record ⟨e, t, k, v⟩ represents an entity e, its type t, a semantic category k ∈{property, relation, aspect}, and a value v, while retaining links to the original source passages. At query time, these records serve as retrieval handles, and an LLM synthesizes the retrieved passages into the final answer. This design separates evidence localization from answer synthesis while preserving traceable source evidence. Across Loong and Oolong, EnSIRAG achieves an average accuracy of 78.24 . Relative to the published baseline scores used as references, this is 6.62 points higher, suggesting its effectiveness across these settings. The code is available at [https://github.com/RamonMeng/EnSI](https://github.com/RamonMeng/EnSI)RAG.  \nIntroduction  \nAnswering questions over long documents remains a fundamental challenge for large language models (LLMs). In many real-world applications, the information required to answer a question is rarely contained within a single passage. Financial analysts may need to compare values across multiple reports, legal researchers may need to classify or aggregate evidence from many judgments, and scientists may need to trace citations, methods, or findings across collections of papers. These tasks require systems not only to locate relevant evidence, but also to preserve entity identity, document structure, and relationships among dispersed pieces of information.  \nRecent advances in LLMs have made long-document question answering increasingly feasible. One prominent direction is to extend the context window, allowing models to consume longer inputs directly (Anthropic 2024; Gemini 2024) . However, larger context windows alone do not guarantee reliable evidence utilization: models remain sensitive to the position of relevant information, exhibiting the well-known  \n“lost in the middle” phenomenon (Liu et al. 2024). Moreover, recent long-context benchmarks demonstrate that even frontier models continue to struggle with questions requiring evidence aggregation and multi-hop reasoning across long documents rather than localized information lookup (Wanget al. 2024; Bertsch et al. 2025; Bai et al. 2025) . Another direction is retrieval-augmented generation (RAG), which retrieves external evidence before generation (Lewis et al. 2020; Guu et al. 2020) . While RAG substantially reduces the amount of text processed by the LLM, it shifts the burden to the retriever, since answer quality is ultimately limited by the evidence it retrieves.  \nMost existing RAG systems treat documents as collections of fixed-length chunks and retrieve them using lexical or dense similarity. Although simple and scalable, fixed-size chunking is often poorly aligned with the semantic organization of long documents. Chunk boundaries may separate entities from their supporting evidence, detach table rows from their headers, or merge multiple unrelated topics into the same retrieval unit. Consequently, retrieval may return passages that are locally similar to the query but insufficient for ","cbCaism4IfAJ1WGA","https://ap.wps.com/l/cbCaism4IfAJ1WGA","pdf",847571,21,"English","# Abstract\n# Introduction\n## Key Challenges in Long-Document QA\n## Limitations of Existing Approaches (Context Extension and RAG)\n## Core Idea: Entity-Centered Retrieval Units\n## Proposed Framework: EnSI-RAG\n## Evaluation Benchmarks and Results","[{\"question\":\"Why do existing RAG methods struggle with long-document question answering?\",\"answer\":\"They often index fixed-length chunks, so entity evidence may be split by chunk boundaries or blended with unrelated topics, leading to incomplete evidence retrieval for multi-hop questions.\"},{\"question\":\"What is EnSI-RAG’s main contribution to improve retrieval?\",\"answer\":\"EnSI-RAG constructs entity-centered passages and a query-independent entity-structure index with retrieval handles linked back to original source passages.\"},{\"question\":\"How does EnSI-RAG connect structured records to final answers?\",\"answer\":\"At query time, the index guides retrieval of supporting passages, and an LLM synthesizes the retrieved evidence into the final response while preserving traceable sources.\"}]","EnSI-RAG: Entity-Structure-Indexed Retrieval-Augmented Generation for Long-Document Question Answering | PDF",53]