Hybrid Retrieval

Follow this topic Ask about this topic

Hybrid retrieval studies how to combine complementary search methods, especially keyword-based methods such as BM25 with dense vector search, and sometimes graphs, metadata, or visual signals. Systems commonly merge ranked lists with reciprocal rank fusion, calibrate or weight scores, and use rerankers, query expansion, or adaptive routing to refine results. The approach is used in settings such as question answering, document search, recommendations, and retrieval-augmented generation, where missing evidence or irrelevant context can undermine downstream answers. There is no universally best combination: gains depend on the domain, query type, and benchmark, while fusion and reranking can add latency or reduce precision.

Latest papers

91 papers9 papers in the last four weeks, down 10% on the four weeks before. 0.1% of all new papers.

A Domain-Aware Pipeline for Structured Extraction and Retrieval-Augmented Question Answering over arXiv Superconductivity Literature
Oct 25, 2026Ivan Iorsh, Aleksandr Kapitonov, Ivan Filianin +2Scientific Document Retrieval Scientific QA
ITMO University, Saint Petersburg, Russia · New Uzbekistan University, Tashkent, Uzbekistan
How Ground Truth Shapes Retrieval Scores: Evidence-Aware Evaluation of a Retrieval-Augmented Generation System for Breast Lesions
Oct 22, 2026Miguel Mochizuki Silva, Samuel Ismael, Mateus C. Dias +3RAG Evaluation Biomedical IR
Universidade Federal da Paraíba (UFPB), João Pessoa, PB, Brazil · Department of Radiology, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
The Fusion Tax: A Diagnostic Analysis of When Hybrid Retrieval Hurts
Oct 22, 2026Aaditya Chauhan, Arun AgarwalHybrid Retrieval Data Fusion
Independent Researcher, Bothell, WA 98012 USA · Independent Researcher, Issaquah, WA 98027 USA
Scan or Index? large Evidence Coverage, Answer Quality, and Cost in Long-Document RAG
Oct 20, 2026Juan Terven, Diana-Margarita Córdova-Esparza, Julio-Alejandro Romero-González +5Long-Context Retrieval Evidence Retrieval
CICATA Qro., Instituto Politécnico Nacional, México · Facultad de Informática, Universidad Autónoma de Querétaro, México · CIDETEC, Instituto Politécnico Nacional, México +3
Improving the Expressive Power of Hybrid Sequence Models via Multi-Hop Retrieval
Oct 18, 2026Yam Eitan, Gilad Yehudai, Haggai MaronMulti-Hop Retrieval Hybrid Attention
Technion – Israel Institute of Technology · NVIDIA
Graph-free Retrieval with Graph-Modified Embeddings
Oct 13, 2026Federica Santi, Vasilije Marković, Daniel Scott +1Evidence Retrieval Graph-Based Retrieval
Cognee Inc
Tree Navigation Without LLM Summaries: A Matched-Cost Study of Hierarchical Retrieval for Long-Document QA
Sep 30, 2026Priyank Jayraj, Poonam Goyal, Navneet GoyalLong-Context Retrieval Information Retrieval
Birla Institute of Technology and Science, Pilani Pilani, Rajasthan, India
PILLAR: Private Inverted-Index Lexical Lookup for Augmented Retrieval
Sep 28, 2026Truong Son Nguyen, Daniel Blackley, Ni Trieu +1Retrieval-Augmented Generation Privacy-Preserving RAG
Arizona State University · Department of Computer Science Arizona State University Tempe, AZ 85281, USA · George Mason University +1
Page-Aware Retrieval-Augmented Generation for EvalLLM 2026: A Five-Variant Study on French PDFs
Sep 28, 2026Abdelhak keliousRetrieval-Augmented Generation Question Decomposition
Capsens, France
Retail Product Search: A Practical Approach at Target
Sep 25, 2026Darshan Sonagara, Qujiaheng Zhang, Ankit Singh +1E-Commerce Search Information Retrieval
Data Sciences, Target Corporation Sunnyvale, CA, USA
Semantic Candidate-Job Matching: A Comparative Evaluation of Dense Embedding Models in Hybrid Retrieval
Sep 20, 2026Sai Yashwant, Siddhartha Jain, Anurag Dubey +2Job-Candidate Matching Text Embedding Models
ManpowerGroup Services India Pvt. Ltd. · Indian Institute of Technology, Hyderabad
Parameterized Dense-Sparse Fusion for Hybrid Retrieval: Tuning a Rank-Score Mix on BEIR SciFact with Qdrant
Sep 19, 2026Satyanarayan Pati, Srikanth PatilHybrid Retrieval Rank Aggregation
Involead Services Pvt. Ltd.
Hybrid GPU-CPU Retrieval for Personalized Search at Ultra-Large Scale
Sep 18, 2026Hao Fu, Jichao Sun, Baiting Zhu +15Information Retrieval Hybrid Retrieval
Meta Platforms, Inc. Menlo Park, California, USA
VoiceTrace: A Benchmark and Retrieval Framework for Who-Said-What Speech Retrieval
Sep 16, 2026Aaron Yee, Fengjie Lu, Jiarui Hai +5Audio-Text Retrieval Hybrid Retrieval
Zhejiang University · Johns Hopkins University · independent researcher
Quanta: A Self-Contained Python Library for Hybrid Retrieval over Quantised Embeddings, Lexical Indexes, and Knowledge Graphs
Sep 16, 2026Ioannis E. LivierisGraph-Based Retrieval Hybrid Retrieval
Novelcore, Athens, GR 10436 · Department of Business Administration & Organization Administration, University of Peloponnese, Kalamata, GR 24100
IROH: Insightful Ranking Of Humor using Multi-Stage Hybrid Retrieval with Rationale-Distilled LLM Judges for JOKER 2026 Track Task 1 English
Sep 14, 2026Ana-Maria Luisa Mocanu, Sebastian Mocanu, Ciprian-Octavian Truica +1LLM-as-a-Judge Hybrid Retrieval
National University of Science and Technology POLITEHNICA Bucharest, Splaiul Independenţei 313, Bucureşti 060042, Romania · Academy of Romanian Scientists, Ilfov 3, Bucharest, 050044, Romania
HARP: Agentic Hybrid Retrieval and Analysis for Long-Form Audio
Sep 12, 2026Chin-Jou Li, Masao Someki, Woojeong Jin +4Audio QA Hybrid Retrieval
Carnegie Mellon University · Hippocratic AI
Comparing Retrieval Methods for Academic Advisor Discovery: A Six-Method Study of 768 CS Faculty Profiles Across 9 US Universities
Sep 3, 2026Biraj SubediLearning to Rank Hybrid Retrieval
Independent Researcher
Annotated Surrogate Retrieval for Polish Statutory Law
Aug 31, 2026Orkun Yiğit CengizLegal IR LLM Reranking
Independent Researcher
SQLite is Enough. Lexical, Semantic, and Hybrid Search with scrydb
Aug 25, 2026Timo BreuerHybrid Retrieval Semantic Search
TH Köln – University of Applied Sciences, 50968 Köln, Germany
EidosDoc: Implicit Structure Encoding for Cost-Effective Semi-Structured Document QA
Aug 17, 2026Teng Lin, Yuyu Luo, Nan TangDocument QA Evidence Retrieval
DSA Thrust HKUST(GZ) Guangzhou, China
Query Expansion Should Be Coordinated: Dense Expands, Sparse Anchors
Aug 16, 2026Chunran ZhangRetrieval-Augmented Generation RAG Evaluation
School of Computing and Artificial Intelligence Southwest Jiaotong University Chengdu, China
Where Does Retrieval Fail? Evaluating RAG Architectures for Agricultural Advisory
Aug 14, 2026Khan Raiyan Ibne Reza, Sanjana Aktar Maria, Sumaiya Tabassum NimiInformation Retrieval RAG Evaluation
North South University Dhaka, Bangladesh
ABAI at COLIEE 2026 Task 1: Multi-Stage Retrieval with GraphRAG-Enhanced Meta-Learning, and a Post-Hoc Study of the Cross-Validation-to-Test Gap
Aug 12, 2026Minhan Cho, Soyoung Park, Daejin Choi +1Legal IR IR Evaluation
AlphaBridge Seongnam-si, Gyeonggi-do, Republic of Korea · Sungkyunkwan University Seoul, Republic of Korea · National Assembly Research Service Seoul, Republic of Korea +1
AkasicDB: Demonstrating Omni RAG with a Unified Vector-Graph-Relational DBMS
Aug 10, 2026Geonho Lee, Jeongho Park, Donghyoung Han +1Retrieval-Augmented Generation Graph Retrieval-Augmented Generation
KAIST Daejeon, Republic of Korea · GraphAI Daejeon, Republic of Korea
VDGR-RAG: Vectors, Directories, Graphs, and Reflection Are All You Need for Unified Reasoning over Hierarchical Enterprise Knowledge
Aug 8, 2026Wenqi Chen, Haofei Yang, Rui Yang +1KG Construction Graph Retrieval-Augmented Generation
Huawei Technologies, Shanghai, China
Comparative Approaches to Agent Retrieval over Large Skill Libraries
Aug 6, 2026Indivara Kolluru, Nathan SportsmanGraph-Based Retrieval LLM Agent Skill Retrieval
Praetorian
SciRet: A Compute-Aware Empirical Study of Retrieval and Reranking for Scientific RAG
Aug 4, 2026Kaysarul Anas Apurba, Md. Hasibul Hasan, Rofiqul Alam Shehab +1Cross-Encoder Reranking Scientific QA
Laurentian University · North South University
Field-Aware Agent Skill Retrieval
Aug 3, 2026Paimon Goulart, Liang Wu, Kelly Wan +2LLM Agent Skill Retrieval Hybrid Retrieval
University of California, Riverside · Nokia
UEmbed: Unified Sparse and Dense Multimodal Embeddings
Aug 3, 2026Tingyu Song, Mingxin Li, Yanzhao Zhang +5Multimodal IR Learned Sparse Retrieval
CASIA · University of Chinese Academy of Sciences · Alibaba Group +1

Oct 25, 2026J/K move · Enter open · S save

A Domain-Aware Pipeline for Structured Extraction and Retrieval-Augmented Question Answering over arXiv Superconductivity Literature

Ivan Iorsh, Aleksandr Kapitonov, Ivan Filianin, Matvey Lesonen+1

ITMO University, Saint Petersburg, Russia · New Uzbekistan University, Tashkent, Uzbekistan

Scientific Document Retrieval Scientific QA Retrieval-Augmented Generation Hybrid Retrieval Materials Science

Open paper AskPDF arXiv

Scientific literature in materials physics is growing faster than researchers can manually inspect it, yet most retrieval tools rely on keyword matching or generic question answering that ignores domain-specific structure in PDF files. This paper presents an end-to-end pipeline that transforms arXiv superconductivity publications into a searchable corpus with structured metadata and retrieval-augmented question answering. The system combines thematic corpus construction, domain-aware extraction of materials and superconducting transition temperatures, and hybrid retrieval that merges dense chunk embeddings with structured publication fields. We test whether normalized domain attributes improve evidence retrieval and answer groundedness relative to lexical-only and dense text-only methods. Extraction is evaluated on a manually verified gold set of 59 superconducting materials, 46 of which have confirmed experimental critical temperatures. On the 33-material primary positive subset, domain-aware pair accuracy reaches 45.5% versus 9.1% for a regex baseline. Retrieval and generation are compared under a fixed language model (qwen2.5:7b-instruct) on a 100-question benchmark, varying only the retrieval mechanism: BM25, dense, hybrid, and hybrid with domain metadata. BM25 gives the highest source recall (Recall@5 of 47%), and adding extracted metadata to hybrid retrieval does not improve it (39% versus 41%), which we attribute to noise in the extracted fields; the main bottleneck is locating the supporting span inside a correctly retrieved paper. As a materials-informatics application, literature-derived temperatures are matched to an independent theoretical reference set: 200 materials appear in both sources, and 20 have a median experimental critical temperature above the theoretical value.