Abstract:Traditional recommender systems (RS) typically use user-item rating histories as their main data source. However, deep generative models now have the capability to model and sample from complex data distributions, including user-item interactions, text, images, and videos, enabling novel recommendation tasks. This comprehensive, multidisciplinary survey connects key advancements in RS using Generative Models (Gen-RecSys), covering: interaction-driven generative models; the use of large language models (LLM) and textual data for natural language recommendation; and the integration of multimodal models for generating and processing images/videos in RS. Our work highlights necessary paradigms for evaluating the impact and harm of Gen-RecSys and identifies open challenges. This survey accompanies a tutorial presented at ACM KDD'24, with supporting materials provided at: <a class="link-external link-https" href="https://encr.pw/vDhLq" rel="external noopener nofollow">this https URL</a>.

What problem does this paper attempt to address?

The problems that this paper attempts to solve mainly focus on how recommender systems (RS) can utilize generative models to enhance the recommendation effect and user experience. Specifically, the paper explores the following aspects: 1. **Limitations of traditional recommender systems**: Traditional recommender systems usually rely on user - item rating history as the main data source, which limits their ability to handle complex data distributions (such as user - item interactions, text, images, and videos). 2. **Application of generative models**: The paper discusses how to use deep generative models (such as variational auto - encoders (VAEs), generative adversarial networks (GANs), diffusion models, etc.) to model and sample complex user - item interaction data, and how to achieve novel recommendation tasks through these models. 3. **Integration of multimodal data**: The paper also explores how to integrate multimodal data (such as text, images, videos) to enhance the performance of recommender systems. For example, by combining text data and image data, generative models can provide richer and more personalized recommendation results. 4. **Advances in natural language processing**: The paper emphasizes the application of large - language models (LLMs) in natural - language - based recommendation, including using pre - trained LLMs for zero - shot and few - shot learning, fine - tuning, retrieval - augmented generation (RAG) and other techniques to improve the accuracy and interpretability of recommendations. 5. **Ethical and social impacts**: The paper also pays attention to the ethical and social problems that may be brought about by the application of generative models in recommender systems, such as privacy protection, fairness, transparency, etc., and proposes some methods to evaluate and mitigate these problems. In summary, this paper aims to comprehensively review the application of generative models in modern recommender systems, covering various generative models from unimodal to multimodal, as well as the challenges and future research directions of these models in practical applications.

A Review of Modern Recommender Systems Using Generative Models (Gen-RecSys)

Analysis of Recommender System Using Generative Artificial Intelligence: A Systematic Literature Review

Recommender Systems Based on Generative Adversarial Networks: A Problem-Driven Perspective

Multi-modal Generative Models in Recommendation System

The 1st Workshop on Recommendation with Generative Models https://rgm-cikm23.github.io

GenRec: Large Language Model for Generative Recommendation

GEMRec: Towards Generative Model Recommendation

A Survey of Generative Search and Recommendation in the Era of Large Language Models

RsyGAN: Generative Adversarial Network for Recommender Systems.

Large Language Model for Generative Recommendation

Testing Deep Learning Recommender Systems Models on Synthetic GAN-Generated Datasets

A survey on Adversarial Recommender Systems: from Attack/Defense strategies to Generative Adversarial Networks

The 2nd Workshop on Recommendation with Generative Models

GenRec: Generative Sequential Recommendation with Large Language Models

The 1st Workshop on Recommendation with Generative Models

Adversarial Machine Learning in Recommender Systems (AML-RecSys)

On Generative Agents in Recommendation

Content-Based Collaborative Generation for Recommender Systems

Non-autoregressive Generative Models for Reranking Recommendation

Generative Recommendation: Towards Next-generation Recommender Paradigm

Advancements in Recommender Systems: A Comprehensive Analysis Based on Data, Algorithms, and Evaluation