Reproduce “Recommendation with Generative Models” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Yashar Deldjoo, Zhankui He, Julian McAuley, Anton Korikov, Scott Sanner, Arnau Ramisa, Rene Vidal, Maheswaran Sathiamoorthy and 3 more

This overview explores Generative Recommender Systems (Gen-RecSys), a paradigm shift from traditional prediction-based methods to systems leveraging deep generative models. It details how these models can generate new content, provide natural language explanations, and enable conversational interaction, transforming the recommendation experience.

Key insights

  • Deep Generative Models (DGMs) like VAEs, GANs, Diffusion Models, and LLMs are adaptable for various recommendation tasks, offering distinct strengths in content generation, data representation, and interaction.
  • Generative Recommender Systems (Gen-RecSys) enable new capabilities such as natural language interaction, on-demand content creation, and multimodal recommendations, moving beyond traditional item prediction.
  • The field faces challenges including computational cost, training stability, complex evaluation, and significant ethical concerns like privacy, bias, and potential for manipulation.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.