[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-82778-en":3,"doc-seo-82778-105":29,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":13,"seo_description":14,"update_tm":27,"read_time":28},82778,137441390410,"Hazel","https://ap-avatar.wpscdn.com/avatar/2000252f4ab5702993?_k=1776741390130283984",8,"Research & Report","Candidate Constrained Retrieval Augmented Generation for LongEval RAG System Design and Empirical Analysis","Candidate-constrained retrieval-augmented generation for LongEval-RAG builds query-time retrieval around an organizer-provided candidate set, requiring that all retrieved evidence and final citations remain inside the set. The pipeline integrates deterministic provenance tracking, passage-based retrieval, deterministic query expansion, pseudo-relevance feedback, reciprocal rank fusion, lightweight reranking, citation-aware evidence aggregation, and optional MiniLM sentence reranking. Ten variants are evaluated via primary organizer scoring and a supplementary diagnostic LLM-judge protocol, with rule-minilm achieving the best overall balance.","Candidate-Constrained Retrieval-Augmented Generation for LongEval-RAG: System Design and Empirical Analysis  \nLongEval-RAG at CLEF 2026 Yingdong Yang1, *,†, Haijian Wu1,†  \nAbstract  \nWe present a candidate-constrained retrieval-augmented generation system for LongEval-RAG, where each query is associated with an organizer-provided candidate set and all retrieved evidence and final citations must remain within that set. The system combines deterministic provenance tracking with passage-based retrieval, deterministic query expansion, pseudo-relevance feedback (PRF), reciprocal rank fusion (RRF), lightweight evidence reranking, citation-aware evidence aggregation, and optional MiniLM sentence reranking. We evaluate ten pipeline variants using a primary organizer evaluation and a supplementary self-generated diagnostic protocol. The primary evaluation shows that the strongest balanced variant is rule-minilm: a rule-based chunking pipeline with query expansion, PRF, RRF, reranking, citation prior, and late MiniLM sentence selection. This variant obtains the highest BERTScore, retrieval precision, nugget coverage, and average grade among our submissions. The result suggests that the main gain does not come from more complex semantic or topic-shift chunking, but from pairing stable rule-based passages with sentence-level neural selection before generation. The supplementary LLM-judge evaluation remains useful for early diagnosis and additional analysis, but it emphasizes different systems than the primary gold-answer and nugget-based evaluation, highlighting the need for multi-metric RAG evaluation.  \nKeywords  \nRetrieval-augmented generation, LongEval-RAG, Evidence ranking, Citation-grounded generation  \n1. Introduction  \nLongEval-RAG studies answer generation from an organizer-provided candidate set rather than unrestricted corpus search [1, 2, 3]. This setting differs from open-domain RAG, where retrieval is typically performed against a large external corpus such as Wikipedia [4, 5] . Instead, each query is paired with an official list of candidate document IDs, and every retrieved passage, selected evidence item, and final citation must remain inside that list. In our system, the main retrieval-time selection unit is the passage: the pipeline ranks and selects passages derived from candidate documents, then selects sentence-level evidence from those passages for answer generation, while the final references list is derived from the source document IDs ofthe selected evidence.  \nThis constraint makes the task especially suitable for analyzing the internal quality of a RAG system. Since broad corpus recall is removed from the problem definition, performance depends more directly on how well the system structures documents into passages, ranks evidence, allocates answer space, and preserves provenance from final claims back to source documents.  \nThis paper makes three main contributions. First, we describe a complete candidate-constrained RAG pipeline for LongEval-RAG, covering data preparation, passage construction, retrieval, evidence selection, answer generation, and deterministic provenance tracking. Second, we report an ablation study often pipeline variants under the primary multi-metric organizer evaluation, while retaining a supplementary cross-provider LLM-judge protocol as a diagnostic view. Third, we identify the strongest primary-evaluation configuration: rule-minilm, which combines deterministic rule-based chunks  \nCLEF 2026 Working Notes, 21 – 24 September 2026, Jena, Germany  \n* Corresponding author.  \n†  \nThese authors contributed equally and share first authorship.  \n$ [yingdongyang0305@outlook.com](yingdongyang0305@outlook.com) (Y. Yang); [willwuhj@gmail.com](willwuhj@gmail.com) (H. Wu)  \n􀂀 https://github.com/yyd859/ (Y. Yang); [https://wijowill.github.io/](https://wijowill.github.io/) (H. Wu)  \n􀀚 0009-0007-5546-5293 (Y. Yang); 0009-0007-0311-6235 (H. Wu)  \n © 2026 Copyright for this paper by its authors. Use per","cbCaiq95IK9ifxYy","https://ap.wps.com/l/cbCaiq95IK9ifxYy","pdf",1319615,1,15,"English","en",105,"# Introduction\n## Task Setting and Data Preparation","[{\"question\":\"What is the main constraint in LongEval-RAG candidate-constrained generation?\",\"answer\":\"Each query is paired with an organizer-provided candidate set, and every retrieved passage, selected evidence item, and final citation must stay within that set.\"},{\"question\":\"Which system variant performed best in the primary evaluation?\",\"answer\":\"The rule-minilm variant achieved the highest BERTScore, retrieval precision, nugget coverage, and average grade among the submitted systems.\"},{\"question\":\"What role does MiniLM reranking play in the best-performing configuration?\",\"answer\":\"MiniLM is applied at the late sentence-selection stage, improving evidence budget allocation before generation rather than changing upstream chunk boundaries.\"}]",1784182881,38,{"code":4,"msg":30,"data":31},"ok",{"site_id":24,"language":23,"slug":32,"title":13,"keywords":33,"description":14,"schema_data":34,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":27},"candidate-constrained-retrieval-augmented-generation-for-longeval-rag-system-design-and-empirical-analysis","",{"@graph":35,"@context":85},[36,53,68],{"@type":37,"itemListElement":38},"BreadcrumbList",[39,43,47,50],{"item":40,"name":41,"@type":42,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":44,"name":45,"@type":42,"position":46},"https://docshare.wps.com/document/","Document",2,{"item":48,"name":12,"@type":42,"position":49},"https://docshare.wps.com/document/research-report/",3,{"item":51,"name":13,"@type":42,"position":52},"https://docshare.wps.com/document/candidate-constrained-retrieval-augmented-generation-for-longeval-rag-system-design-and-empirical-analysis/82778/",4,{"url":51,"name":13,"@type":54,"author":55,"headline":13,"publisher":57,"fileFormat":60,"inLanguage":23,"description":14,"dateModified":61,"datePublished":62,"encodingFormat":60,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":56},"Person",{"url":40,"name":58,"@type":59},"DocShare","Organization","application/pdf","2026-07-17","2026-07-16",true,{"@type":65,"interactionType":66,"userInteractionCount":20},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"What is the main constraint in LongEval-RAG candidate-constrained generation?","Question",{"text":75,"@type":76},"Each query is paired with an organizer-provided candidate set, and every retrieved passage, selected evidence item, and final citation must stay within that set.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Which system variant performed best in the primary evaluation?",{"text":80,"@type":76},"The rule-minilm variant achieved the highest BERTScore, retrieval precision, nugget coverage, and average grade among the submitted systems.",{"name":82,"@type":73,"acceptedAnswer":83},"What role does MiniLM reranking play in the best-performing configuration?",{"text":84,"@type":76},"MiniLM is applied at the late sentence-selection stage, improving evidence budget allocation before generation rather than changing upstream chunk boundaries.","https://schema.org",{"og:url":51,"og:type":87,"og:title":13,"og:site_name":58,"og:description":14},"article",{"robots":89,"canonical":51},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":45,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":46,"doc_module":4,"doc_module_name":45,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":52,"doc_module":4,"doc_module_name":45,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":45,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":45,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":45,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":45,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":45,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":45,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":45,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":45,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]