information retrieval

This study investigates whether retrieval fusion techniques—commonly adopted in real-world retrieval-augmented generation (RAG) systems, such as multi-query retrieval and reciprocal rank fusion—consistently improve end-to-end answer quality under practical deployment constraints. Conducted within an enterprise knowledge-base RAG pipeline, the evaluation is performed under fixed retrieval depth, reranking budget, and latency limits. While retrieval fusion enhances initial recall, it fails to translate into improved Top-k accuracy after subsequent reranking and context truncation; notably, Hit@10 declines from 0.51 to 0.48 and incurs additional latency. These findings challenge the prevailing assumption of the default efficacy of recall-oriented fusion strategies, revealing diminishing returns in production settings where downstream processing and system constraints critically shape overall performance.

To address the semantic relevance modeling mismatch between traditional retrievers and generators in retrieval-augmented generation (RAG), this paper proposes the first end-to-end differentiable retriever joint optimization framework. It enables collaborative training of retriever and generator via gradient backpropagation and introduces context-aware relevance modeling, allowing the retriever to explicitly learn task-specific semantic matching required by the generator. This approach is the first to support full-pipeline differentiability and dynamic task adaptation in RAG. Experiments demonstrate an average 4.0% improvement over baselines across multiple tasks and consistent superiority—by 2.1%—over state-of-the-art retrievers. Notably, a lightweight 0.2B-parameter retriever outperforms an 8B-parameter instruction-tuned LLM-enhanced RAG system on several tasks, achieving significant gains in both efficiency and performance trade-offs.

To address low semantic recall accuracy and high latency in large-enterprise knowledge base retrieval, this paper proposes a retrieval-augmented framework leveraging large language models (LLMs) to dynamically generate structured metadata. Methodologically, it integrates semantic, recursive, and naive chunking strategies; enhances vector representations via TF-IDF–weighted embeddings and prefix fusion; and employs a cross-encoder for re-ranking to construct high-quality ground-truth labels. A novel metadata consistency metric is introduced to rigorously evaluate retrieval effectiveness. Experimental results demonstrate that recursive chunking combined with TF-IDF–weighted embeddings achieves 82.5% precision, while naive chunking with prefix fusion attains a Hit Rate@10 of 0.925—both significantly outperforming content-only baselines. The framework delivers superior accuracy, low latency, and strong scalability, making it suitable for enterprise-scale knowledge retrieval.

This study addresses the confounding effects of metadata, structured representations, and retrieval mechanisms in current RAG systems, which often combine multiple context-augmentation strategies, obscuring their individual contributions to answer quality. Through controlled experiments across six benchmarks, four models, and five augmentation levels—totaling over 24,000 evaluations—the work reveals that increased contextual richness does not necessarily improve accuracy. It introduces the “tractability hierarchy” theory, emphasizing that context must align with model capacity. The findings demonstrate that most augmentation strategies actually degrade performance; however, when metadata and retrieval strategies are carefully matched to a model’s capabilities, smaller models can outperform state-of-the-art large models by up to 19 F1 points on specific tasks, challenging the prevailing RAG design paradigm centered on stacking metadata.

The impact of key design decisions—RAG activation, retrieval granularity, and knowledge integration strategy—on RAG system performance remains poorly understood. Method: We conduct systematic ablation studies across three code/qa benchmarks and two state-of-the-art LLMs, quantitatively evaluating how document type, recall rate, document selection strategy, and prompt engineering jointly affect answer correctness and confidence via multi-dimensional analysis, cross-model/dataset comparison, and joint prompt-retrieval analysis. Contribution/Results: We identify precise interaction patterns and operational boundaries among these factors and propose nine actionable, empirically grounded guidelines for diagnosing and optimizing RAG failures. Our findings significantly improve RAG system stability, debuggability, and reliability, offering rigorous empirical evidence and a principled methodology to support the engineering deployment of LLM-augmented systems.

This work proposes an end-to-end graph-based retrieval-augmented generation (RAG) framework that addresses the limitations of traditional RAG methods in efficiently retrieving relevant information within unknown search spaces or when handling semi-structured and structured documents. By integrating labeled property graphs (LPGs) with the Resource Description Framework (RDF), the approach automatically converts JSON key-value pairs into RDF triples to incorporate semi-structured data. It further introduces a text-to-Cypher query generation mechanism, enabling real-time, high-precision graph retrieval without requiring a predefined number of source documents. Eliminating inefficient re-ranking steps, the method significantly enhances answer accuracy, reasoning capability, and overall response quality, demonstrating particularly strong performance in complex semi-structured tasks and online scenarios.

This work addresses the lack of an integrated Retrieval-Augmented Generation (RAG) development and evaluation toolkit in the R programming language, where existing solutions predominantly rely on the Python ecosystem. The authors propose ragR, the first native R framework enabling end-to-end RAG workflows—including document ingestion, vector storage, similarity-based retrieval, evidence synthesis, and structured question-answering logging—and faithfully reimplements the four core RAGAS evaluation metrics: context precision, context recall, faithfulness, and answer relevance. Experimental results demonstrate that ragR yields evaluation outcomes highly consistent with those produced by the original Python-based RAGAS implementation. This provides R users with a lightweight, reproducible, and self-contained environment for RAG research and education without requiring language switching.

This study addresses the lack of systematic evaluation of modern retrieval methods in Retrieval-Augmented Generation (RAG) systems for heterogeneous financial documents containing both text and tables. It presents the first large-scale benchmark comparing ten retrieval strategies—including sparse, dense, hybrid fusion, cross-encoder reranking, and query expansion—on a financial mixed-document question answering task. Evaluation employs multiple metrics (Recall@k, MRR, nDCG, and numerical match accuracy) alongside bootstrap significance testing. Results reveal that BM25 outperforms state-of-the-art dense retrievers in this domain; a two-stage hybrid retrieval pipeline with neural reranking achieves substantial gains (Recall@5 = 0.816, MRR@3 = 0.605); context-aware retrieval consistently improves performance, whereas query expansion yields limited benefits for numerical questions. The authors release a comprehensive open-source benchmark and provide practical guidance on cost–accuracy trade-offs.

This study investigates whether retrieval quality can serve as a reliable early indicator of information coverage in responses generated by Retrieval-Augmented Generation (RAG) systems. Through systematic experiments across three benchmarks—TREC NeuCLIR 2024, TREC RAG 2024, and WikiVideo—the authors evaluate 15 text-based and 10 multimodal retrieval systems using the Auto-ARGUE and MiRAGE assessment frameworks. The work provides the first empirical evidence of a strong correlation between coverage-oriented retrieval metrics and the informational coverage of generated outputs. Findings reveal that, at both topic and system levels, such metrics effectively predict RAG output coverage when retrieval and generation objectives are aligned, underscoring the critical role of goal consistency in optimizing RAG performance.

This work addresses the limitations of traditional Retrieval-Augmented Generation (RAG) systems, which organize evidence in a flat manner and struggle to effectively manage retrieval states—such as search trajectories, entity relationships, and intermediate reasoning artifacts—thereby hindering cross-task reasoning efficiency and accuracy. The paper formalizes RAG as a structured retrieval state management problem and introduces three core mechanisms: a Typed Hierarchical State Space (TAM), Multi-Agent Role-based Synergy verification (MARS), and a State-aware Memory Pool (SMP). Experimental results demonstrate that the proposed approach achieves state-of-the-art question-answering performance on three LongBench subsets, matches the strongest agent-based baseline in EM score on HotpotQA while reducing large model token consumption by 3.51×, and enables low-overhead inference on DocVQA with cross-query cache hit rates ranging from 3.77% to 23.18%.