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

LLMs for User Interest Exploration in Large-scale Recommendation Systems (Google, KDD 2024 GenAIRecP Workshop)

Forum topic · 小凯 · 2026-07-05

Summary

This forum post indexes a Google research paper titled "LLMs for User Interest Exploration in Large-scale Recommendation Systems," presented at the Generative AI and Recommender Systems (GenAIRecP) Workshop at KDD 2024. The work sits at the intersection of large language models and industrial recommendation systems, addressing how LLMs can help explore and expand user interests beyond what traditional retrieval and ranking pipelines capture. The post frames the paper within broader trends in generative recommendation (Gen-Rec), retrieval-augmented generation (RAG), and agentic search, and discusses the core engineering tensions in deploying LLMs at scale: latency and cost budgets, hallucination risks, sparse user behavior signals, massive item catalogs, and cold-start handling. It also provides an engineering checklist covering data privacy, p99 latency, online A/B validation, safety filtering, and per-query compute cost, along with a glossary of terms such as IR, LTR, nDCG, and agentic search. The direct paper PDF is linked, and readers are advised to treat paper results as directional evidence to be validated via shadow experiments and A/B tests before full production rollout. Suitable for researchers and engineers working on LLM-based recommendation, semantic user interest modeling, and production recommender architecture.

LLMs for User Interest Exploration in Large-scale Recommendation Systems (Google, KDD 2024 GenAIRecP Workshop)

This post indexes a Google research paper presented at the Generative AI and Recommender Systems (GenAIRecP) Workshop at KDD 2024:

  • Title: LLMs for User Interest Exploration in Large-scale Recommendation Systems
  • Affiliation: Google
  • Paper PDF: https://genai-personalization.github.io/assets/papers/GenAIRecP2024/Wang.pdf
  • Section: Discovery
  • Context and Motivation

    Large-scale search and recommendation systems have long faced challenges around efficiency, scalability, and understanding user intent. Traditional pipeline approaches separate retrieval, ranking, and generation, which makes it hard to meet modern expectations for natural-language interaction, multi-hop reasoning, and fresh knowledge. This work explores how LLMs can be used to explore and expand user interests in an industrial-scale recommender, complementing conventional retrieval with semantic priors and cold-start capability.

    The core tension highlighted in the post: LLMs bring strong semantic understanding, but online inference cost, latency budgets, and hallucination risks require careful system design when item catalogs are massive and user behavior is sparse.

    Where It Fits Technically

    The work sits at the intersection of recommendation and large-scale search. In the classic search stack — recall (coverage) → ranking (discrimination) → generation (presentation) — the LLM era adds two new variables: the inference budget and the action space (whether to retrieve, how many times, and which tools to call).

    Related evolution across the field:

  • Retrieval: BM25 → BERT cross-encoders → dense two-tower retrieval → late interaction → generative retrieval and LLM agents.
  • Recommendation: matrix factorization → deep CTR models → sequential Transformers → LLM instruction-following and generative recommendation (Gen-Rec).
  • RAG / Agentic Search: external knowledge access becomes iterative, verifiable, and plannable; evaluation shifts from static nDCG toward task success rate, citation accuracy, and multi-hop reasoning chain completeness.
  • Engineering Deployment Checklist

    | Area | Question | Suggestion | |--------|----------|------------| | Data | Does training/indexing contain PII? How are versions managed? | Partitioned indexes, anonymization, rollback-capable embedding versions | | Latency | What is the p99 budget? How many retrieval steps? | Cascades with early stopping, caching hot queries, async re-ranking | | Quality | Do offline gains translate to online CTR/satisfaction? | Interleaving experiments, human audits, citation verification | | Safety | Does open retrieval introduce poisoning/bias? | Source allow-lists, adversarial detection, output filtering | | Cost | Token and GPU cost per query? | Route to smaller models, distillation, hybrid sparse+dense retrieval |

    Practical Advice

    1. Researchers: Reproduce the core comparisons; check whether statistical significance and compute costs are reported. 2. Engineers: Extract pluggable modules (encoders, re-rankers, planners) and evaluate integration cost with your existing stack. 3. Product managers: Focus on user-perceivable benefits (latency, answer trustworthiness, multi-turn consistency) rather than offline nDCG alone.

    Treat paper conclusions as directional evidence: distribution drift, seasonality, new-item cold start, and multilingual user mixtures can change optimal strategies. Validate with small-scale shadow experiments and A/B tests before full rollout.

    Glossary

    | Term | Meaning | |------|---------| | 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 calls | | Gen-IR | Generative Information Retrieval |

    Reference

  • Original paper PDF: <https://genai-personalization.github.io/assets/papers/GenAIRecP2024/Wang.pdf> (GenAIRecP Workshop @ KDD 2024, Google)

Tags

#llm#recommendation-systems#kdd-2024#google#generative-ai#user-interest-modeling#rag#industrial-ml

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/178208620