The Viability of Crowdsourcing for RAG Evaluation

Abstract

How good are humans at writing and judging responses in retrieval-augmented generation (RAG) scenarios? To answer this question, we investigate the efficacy of crowdsourcing for RAG through two complementary studies: response writing and response utility judgment. We present the Crowd RAG Corpus 2025 (CrowdRAG-25), which consists of 903 human-written and 903 LLM-generated responses for the 301 topics of the TREC RAG'24 track, across the three discourse styles 'bulleted list', 'essay', and 'news'. For a selection of 65 topics, the corpus further contains 47,320 pairwise human judgments and 10,556 pairwise LLM judgments across seven utility dimensions (e.g., coverage and coherence). Our analyses give insights into human writing behavior for RAG and the viability of crowdsourcing for RAG evaluation. Human pairwise judgments provide reliable and cost-effective results compared to LLM-based pairwise or human/LLM-based pointwise judgments, as well as automated comparisons with human-written reference responses. All our data and tools are freely available.

11 pages, 9 tables, 5 figures. Accepted at SIGIR'25

References in corpus (11)

• LLMs-as-Judges: A Comprehensive Survey on LLM-based Evaluation Methods
• HAGRID: A Human-LLM Collaborative Dataset for Generative Information-Seeking with Attribution
• MultiHop-RAG: Benchmarking Retrieval-Augmented Generation for Multi-Hop Queries
• A Large-Scale Study of Relevance Assessments with Large Language Models: An Initial Look
• Initial Nugget Evaluation Results for the TREC 2024 RAG Track with the AutoNuggetizer Framework
• RECALL: A Benchmark for LLMs Robustness against External Counterfactual Knowledge
• DomainRAG: A Chinese Benchmark for Evaluating Domain-specific Retrieval-Augmented Generation
• LLMs Can Patch Up Missing Relevance Judgments in Evaluation
• GenTREC: The First Test Collection Generated by Large Language Models for Evaluating Information Retrieval Systems
• Ragnarök: A Reusable RAG Framework and Baselines for TREC 2024 Retrieval-Augmented Generation Track
• JudgeBlender: Ensembling Judgments for Automatic Relevance Assessment