{"ID":2892069,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2507.22917","arxiv_id":"2507.22917","title":"Reading Between the Timelines: RAG for Answering Diachronic Questions","abstract":"While Retrieval-Augmented Generation (RAG) excels at injecting static, factual knowledge into Large Language Models (LLMs), it exhibits a critical deficit in handling longitudinal queries that require tracking entities and phenomena across time. This blind spot arises because conventional, semantically-driven retrieval methods are not equipped to gather evidence that is both topically relevant and temporally coherent for a specified duration. We address this challenge by proposing a new framework that fundamentally redesigns the RAG pipeline to infuse temporal logic. Our methodology begins by disentangling a user's query into its core subject and its temporal window. It then employs a specialized retriever that calibrates semantic matching against temporal relevance, ensuring the collection of a contiguous evidence set that spans the entire queried period. To enable rigorous evaluation of this capability, we also introduce the Analytical Diachronic Question Answering Benchmark (ADQAB), a challenging evaluation suite grounded in a hybrid corpus of real and synthetic financial news. Empirical results on ADQAB show that our approach yields substantial gains in answer accuracy, surpassing standard RAG implementations by 13% to 27%. This work provides a validated pathway toward RAG systems capable of performing the nuanced, evolutionary analysis required for complex, real-world questions. The dataset and code for this study are publicly available at https://github.com/kwunhang/TA-RAG.","short_abstract":"While Retrieval-Augmented Generation (RAG) excels at injecting static, factual knowledge into Large Language Models (LLMs), it exhibits a critical deficit in handling longitudinal queries that require tracking entities and phenomena across time. This blind spot arises because conventional, semantically-driven retrieval...","url_abs":"https://arxiv.org/abs/2507.22917","url_pdf":"https://arxiv.org/pdf/2507.22917v1","authors":"[\"Kwun Hang Lau\",\"Ruiyuan Zhang\",\"Weijie Shi\",\"Xiaofang Zhou\",\"Xiaojun Cheng\"]","published":"2025-07-21T05:19:41Z","proceeding":"cs.CL","tasks":"[\"cs.CL\",\"cs.AI\",\"cs.IR\"]","methods":"[\"RAG\",\"Large Language Model\",\"Language Model\"]","has_code":false,"code_links":[{"ID":611951,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_id":2892069,"paper_url":"https://arxiv.org/abs/2507.22917","paper_title":"Reading Between the Timelines: RAG for Answering Diachronic Questions","repo_url":"https://github.com/kwunhang/TA-RAG","is_official":false,"mentioned_in_paper":false,"mentioned_in_github":true,"github_stars":0}]}