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

Multimodal Pretraining, Adaptation, and Generation for Recommendation: A Survey

Forum topic · 小凯 · 2026-07-05

Summary

This arXiv survey (2404.00621, March 2024) by Qijiong Liu, Jieming Zhu, and colleagues reviews how multimodal pretraining, adaptation, and generation techniques can improve recommender systems. Traditional recommenders rely on unique IDs and categorical features, overlooking raw item content across text, image, audio, and video modalities—a limitation especially acute in multimedia services such as news, music, and short-video platforms. The survey organizes the field along modeling paradigms (discriminative vs. generative retrieval), LLM integration patterns (RAG, agents, tool use), optimization objectives, and evaluation protocols, and traces the evolution from BERT-era neural retrieval through RAG to agentic and generative recommendation systems. It covers benchmark datasets (MS MARCO, BEIR, Natural Questions), common metrics (nDCG, MRR, Recall@k), and open challenges including evaluation trustworthiness, latency and cost constraints, hallucination and safety risks, and cross-lingual or multimodal generalization. The paper is a useful reference for researchers and engineers building content-aware, LLM-powered recommendation pipelines.

Multimodal Pretraining, Adaptation, and Generation for Recommendation: A Survey

Source: https://arxiv.org/abs/2404.00621 · Type: Survey · Posted: 2024-03-31 · Authors: Qijiong Liu, Jieming Zhu, Yanting Yang, Quanyu Dai, Zhaocheng Du, Xiao-Ming Wu, et al. (9 authors)

Overview

Personalized recommendation is a ubiquitous channel for users to discover content tailored to their interests. However, traditional recommendation models primarily rely on unique IDs and categorical features for user-item matching, overlooking the nuanced essence of raw item contents across modalities such as text, image, audio, and video. This underutilization is especially limiting for multimedia services like news, music, and short-video platforms. Recent advances in large multimodal models create new opportunities—and challenges—for building content-aware recommender systems. This survey provides a comprehensive exploration of multimodal pretraining, adaptation, and generation techniques and their applications to recommendation.

Key points

  • Unified perspective: Consolidates scattered work on multimodal recommendation into a comparable framework spanning representation learning, retrieval, reranking, planning, generation, and feedback mechanisms.
  • Motivation: ID-based matching misses semantic signals in raw content; multimodal models offer semantic priors and cold-start capabilities, at the cost of inference expense and hallucination risk.
  • Taxonomy

    | Dimension | Subcategories | Representative ideas | Strengths | Limitations | |---|---|---|---|---| | Modeling paradigm | Discriminative / generative retrieval | Two-tower, cross-encoders, DSI, GPT indexing | Mature, scalable | Semantic drift, update cost | | LLM integration | RAG / Agent / Tool-use | Retrieval augmentation, search agents, API calls | Flexible, interpretable | Latency, error propagation | | Objectives | Relevance / diversity / freshness | Multi-objective LTR, RLHF, online learning | Business-aligned | Scarce annotations | | Evaluation | Offline / online / human | nDCG, MRR, LLM-as-judge, A/B | Comparable | Deviates from real satisfaction |

    Four research threads

  • Dense retrieval: high recall, low latency; suited to first-stage retrieval.
  • Late interaction (e.g., ColBERT): higher precision but larger indexes.
  • Generative IR: directly "generates" documents via tokens or docids, simplifying cascades.
  • Agentic search: frames search as sequential decision-making, enabling multi-hop reasoning and self-reflection.
  • Timeline

    2019–2021: BERT rerankers and DPR established neural retrieval; 2022–2023: RAG and FreshLLM fused retrieval with generation; 2024 onward: conversational/agentic search and Gen-RecSys exploded; 2025–2026: RL-trained search agents, Deep Research, and GraphRAG are emerging growth areas.

    Evaluation landscape

  • Datasets: MS MARCO, BEIR, Natural Questions, domain corpora, public recommendation sets.
  • 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.
Quantitative results should be verified against the original PDF.

Takeaways

1. Architecture: cascaded retrieve–rerank–generate remains mainstream, but agentic paradigms make retrieval strategy itself learnable. 2. Data: high-quality instruction data and click/session logs are critical; synthetic data risks leakage and distribution shift. 3. Evaluation: the gap between offline metrics and online satisfaction is widening; LLM-as-judge needs cross-validation with human assessment. 4. Production: latency, cost, explainability, and safety are hard constraints for deployment—do not optimize academic benchmarks alone.

Open challenges

Insufficient unified benchmarks, non-reproducible private data, LLM evaluation bias, and safety/cost constraints of agentic systems. Future directions include finer-grained process supervision, joint retrieval-reasoning training, multimodal and cross-lingual consistency, and causal/fairness constraints for recommenders.

> Original abstract (excerpt): "Personalized recommendation serves as a ubiquitous channel for users to discover information tailored to their interests... This survey seeks to provide a comprehensive exploration of the latest advancements and future trajectories in multimodal pretraining, adaptation, and generation techniques, as well as their applications in enhancing recommender systems."

Tags

#recommender-systems#multimodal-learning#survey#llm#retrieval-augmented-generation#generative-retrieval#information-retrieval#pretraining

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/178208902