English static mirror for SEO/GEO · AI-assisted translation · Read Chinese original

CLUE: Using Large Language Models for Judging Document Usefulness in Web Search Evaluation (CIKM 2025)

Forum topic · 小凯 · 2026-07-05

Summary

This forum entry on zhichai.net indexes the CIKM 2025 paper "CLUE: Using Large Language Models for Judging Document Usefulness in Web Search Evaluation" (ACM DL: 10.1145/3746252.3761158). The paper addresses relevance judging, a core bottleneck in web search evaluation where human assessors must judge whether retrieved documents are useful for a query. CLUE explores replacing or augmenting manual judging with large language models, fitting the broader shift toward LLM-as-judge evaluation in information retrieval. The forum post situates the work within the search/recommendation/personalization landscape, discussing the classic pipeline of retrieval, reranking, and generation; the growing agentic search paradigm where the number and strategy of retrieval steps become learnable; and evaluation gaps between offline metrics like nDCG@10 and online user satisfaction. It also provides an engineering checklist covering PII handling, p99 latency budgets, online-vs-offline quality transfer, retrieval poisoning risks, and per-query token cost, plus a glossary of IR terms (RAG, LTR, nDCG, agentic search). Cross-referenced entries include RAG evaluation surveys, the ARES evaluation framework, and AgentBoard. Quantitative results should be verified against the original PDF, as the post itself notes it summarizes from abstract and public metadata.

CLUE: Using Large Language Models for Judging Document Usefulness in Web Search Evaluation (CIKM 2025)

  • Paper: CLUE: Using Large Language Models for Judging Document Usefulness in Web Search Evaluation
  • Venue: CIKM 2025
  • Link: https://dl.acm.org/doi/abs/10.1145/3746252.3761158
  • Category: Evaluation of Search Engines
  • Key points

  • The paper tackles document usefulness judging, one of the most labor-intensive steps in web search evaluation, and investigates whether large language models can perform (or assist with) this task at scale.
  • It belongs to the growing LLM-as-judge line of work, where model-based assessors are cross-validated against human judgments rather than trusted blindly.
  • Context: search evaluation in the LLM era

    The forum entry situates CLUE within a broader shift in information retrieval. Classic search stacks form a funnel — recall (candidate generation), precision (reranking), and presentation — but LLM-era systems add two new variables: inference budget and action space (whether to retrieve, how many times, and which tools to call). Evaluation is consequently moving from static metrics like nDCG toward process-oriented measures such as task success rate, citation accuracy, and multi-hop reasoning chain completeness.

    Key observations from the entry:

    1. Architecture: cascaded retrieve + rerank + generate remains mainstream, but agentic paradigms increasingly make retrieval strategy itself learnable. 2. Data: high-quality instruction data and click/session logs matter as much as model design; synthetic data risks knowledge leakage and distribution shift. 3. Evaluation: the gap between offline metrics and online satisfaction is widening; LLM judges need human cross-validation — precisely the problem space CLUE addresses. 4. Product: latency, cost, explainability, and safety are hard constraints that academic benchmarks alone do not capture.

    Engineering checklist (from the post)

    | Item | Question | Suggestion | |------|----------|------------| | Data | PII in training/index? Versioning? | Partitioned indexes, anonymization, rollback-capable embedding versions | | Latency | p99 budget? Retrieval steps? | Cascades + early stopping, hot-query caching, async reranking | | Quality | Does offline gain transfer to online CTR/satisfaction? | Interleaving experiments, human audit samples, citation checks | | Safety | Poisoning/bias via open retrieval? | Source allowlists, adversarial detection, output filtering | | Cost | Token and GPU cost per query? | Small-model routing, distillation, hybrid sparse+dense |

    Glossary

  • IR: Information Retrieval
  • RAG: Retrieval-Augmented Generation
  • LTR: Learning to Rank
  • nDCG: Normalized Discounted Cumulative Gain, a ranking quality metric
  • Agentic Search: modeling search as sequential decision-making and tool calling
  • Related entries

  • ARES: An Automated Evaluation Framework for RAG (arXiv:2311.09476)
  • Evaluation of Retrieval-Augmented Generation: A Survey (arXiv:2405.07437)
  • AgentBoard: An Analytical Evaluation Board of Multi-turn LLM Agents (arXiv:2401.13178)
  • AI Search Has A Citation Problem (CJR, March 2025)
> Note: This post is an annotated index entry based on the paper's abstract and public metadata. Quantitative results should be verified against the original PDF.

Tags

#information-retrieval#llm-as-judge#search-evaluation#relevance-judging#web-search#cikm-2025#rag#llm

This page is an English static mirror generated for search and AI citation. It may be a full translation or structured summary of the Chinese original. Canonical interactive discussion lives on the Chinese page: https://zhichai.net/topic/178208663