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

Representation Learning with Large Language Models for Recommendation (WWW 2024)

Forum topic · 小凯 · 2026-07-05

Summary

This forum entry catalogs the WWW 2024 paper 'Representation Learning with Large Language Models for Recommendation' (DOI: 10.1145/3589334.3645458), a study at the intersection of large language models (LLMs) and recommender systems. The post situates the work within the broader evolution of search and personalization: from BM25 and dense retrieval to generative recommendation and agentic, LLM-driven pipelines. It outlines a typical technical pipeline—input representation, core modules (retriever, reranker, planner, generator), learning strategies (supervised fine-tuning, contrastive learning, distillation, reinforcement learning), and inference strategies such as iterative retrieval and budget control. It also discusses evaluation practices (nDCG@10, MRR, Recall@k, LLM-as-judge), engineering constraints for production deployment (latency, cost, safety, index versioning), and open problems including evaluation credibility, hallucination, cross-lingual and multimodal extension. The entry includes an engineering checklist, a glossary of IR/RAG terms, and cross-references to related papers such as 360Brew, DiffKG, and Netflix search augmentation. Readers should consult the original PDF for quantitative results, as the post is a curated summary rather than the primary source.

Representation Learning with Large Language Models for Recommendation (WWW 2024)

This post is a curated entry for the paper "Representation Learning with Large Language Models for Recommendation", published at WWW 2024.

  • Source: https://dl.acm.org/doi/abs/10.1145/3589334.3645458
  • Venue: The Web Conference (WWW) 2024
  • Area: Recommender systems × Large Language Models
  • > Note: The original forum post is largely a template-based summary. Quantitative results should be verified against the paper PDF.

    Context and Motivation

    Large-scale search and recommendation systems have long faced challenges in efficiency, scalability, and user-intent understanding. Traditional pipelines split retrieval, ranking, and generation into disjoint stages, which makes it hard to meet LLM-era expectations for natural-language interaction, multi-hop reasoning, and up-to-date knowledge. This paper addresses the intersection of representation learning with LLMs and recommendation, aiming to advance both theory and practice in this cross-domain area.

    Core Themes

  • A unified perspective for organizing scattered related work in LLM-based recommendation.
  • A decomposition of method components: representation learning, retrievers, rerankers, planners, generators, and feedback mechanisms.
  • Discussion of interfaces with emerging paradigms such as LLM tool calling, reinforcement learning, and multi-agent collaboration.
  • Open problems: evaluation credibility, latency and cost, hallucination and safety, cross-lingual and multimodal extension.
  • Typical Method Pipeline

    1. Input & representation — encode queries, documents, and user context as dense/sparse representations or structured prompts. 2. Core modules — retriever, reranker, planner, memory, tool interfaces, composed in series or parallel. 3. Learning strategies — supervised fine-tuning, contrastive learning, distillation, RL (including process rewards), synthetic data bootstrapping. 4. Inference strategies — single-pass retrieval, iterative retrieval, parallel sub-queries, early stopping and budget control.

    Evaluation Landscape

  • Datasets: MS MARCO, BEIR, Natural Questions, domain corpora, public recommendation benchmarks.
  • Metrics: nDCG@10, MRR, Recall@k, Hit@k, human preference, task success rate, latency and token cost.
  • Baselines: BM25, dense retrieval, cross-encoder reranking, retrieval-free LLMs, commercial search APIs.
  • Takeaways for Search / Rec / Personalization

    1. Architecture: cascade retrieval + rerank + generate remains mainstream, but agentic paradigms make "how many retrievals and what policy" itself learnable. 2. Data: high-quality instruction data and click/session logs matter equally; synthetic data must guard against leakage and distribution shift. 3. Evaluation: the gap between offline metrics and online satisfaction is widening; LLM-as-judge needs cross-validation with human evaluation. 4. Product: latency, cost, explainability, and safety are hard constraints for industrial deployment—not just academic benchmarks.

    Engineering Checklist

    | Item | Question | Suggestion | |------|----------|------------| | Data | PII in training/index? Versioning? | Partitioned indexes, anonymization, rollback-able embedding versions | | Latency | p99 budget? Retrieval steps? | Cascade + early stop, hot-query caching, async reranking | | Quality | Do offline gains transfer to online CTR/satisfaction? | Interleaving experiments, human audits, citation checks | | Safety | Does open retrieval introduce poisoning/bias? | Source whitelists, adversarial detection, output filtering | | Cost | Per-query tokens and GPU usage? | Small-model routing, distillation, hybrid sparse+dense |

    Glossary

  • IR — Information Retrieval
  • RAG — Retrieval-Augmented Generation
  • LTR — Learning to Rank
  • nDCG — Normalized Discounted Cumulative Gain
  • Agentic Search — search modeled as sequential decision-making with tool calls
  • Gen-IR — Generative Information Retrieval
  • Related Entries

  • 360Brew: A Decoder-only Foundation Model for Personalized Ranking
  • Actions Speak Louder than Words: Trillion-Parameter Sequential Transducers
  • Augmenting Netflix Search with In-Session Adapted Recommendations
  • Bridging Language and Items for Retrieval and Recommendation
  • Data-efficient Fine-tuning for LLM-based Recommendation, SIGIR 2024
  • DiffKG: Knowledge Graph Diffusion Model for Recommendation, WSDM 2024

Tags

#recommender-systems#large-language-models#representation-learning#information-retrieval#rag#www-2024#personalization#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/178208916