PulseAugur
EN
LIVE 14:49:27

New TDPM framework enhances generative recommendation with time-aware diffusion

By PulseAugur Editorial · [3 sources] ·

Researchers have developed a new framework called TDPM for generative recommendation systems that utilizes time-aware diffusion models. This approach addresses the limitation of existing models by accounting for the temporal evolution of user preferences, which can be influenced by long-term trends and recent events. Experiments on real-world datasets show TDPM significantly outperforms current methods, achieving up to a 29.21% improvement in HR@20 and 25.45% in NDCG@20. AI

IMPACT Enhances generative recommendation systems by incorporating temporal user preferences, potentially leading to more accurate and personalized suggestions.

RANK_REASON The cluster contains a research paper detailing a new framework for generative recommendation systems.

Read on Hugging Face Daily Papers →

AI-generated summary · Google Gemini · from 3 sources. How we write summaries →

COVERAGE [3]

  1. arXiv cs.AI TIER_1 English(EN) · Bangguo Zhu, Peng Huo, Yuanbo Zhao, Zhicheng Du, Jun Yin, Senzhang Wang ·

    Time-Aware Diffusion based on Preference Disentanglement for Generative Recommendation

    arXiv:2606.01670v1 Announce Type: cross Abstract: Recently, Generative Recommenders (GRs) have emerged as a transformative recommendation paradigm by replacing traditional item IDs with semantic indices (SIDs). Owing to the exceptional generative capabilities of diffusion models,…

  2. Hugging Face Daily Papers TIER_1 English(EN) ·

    Time-Aware Diffusion based on Preference Disentanglement for Generative Recommendation

    Recently, Generative Recommenders (GRs) have emerged as a transformative recommendation paradigm by replacing traditional item IDs with semantic indices (SIDs). Owing to the exceptional generative capabilities of diffusion models, a few pioneering works explore developing GRs wit…

  3. arXiv cs.IR (Information Retrieval) TIER_1 English(EN) · Senzhang Wang ·

    Time-Aware Diffusion based on Preference Disentanglement for Generative Recommendation

    Recently, Generative Recommenders (GRs) have emerged as a transformative recommendation paradigm by replacing traditional item IDs with semantic indices (SIDs). Owing to the exceptional generative capabilities of diffusion models, a few pioneering works explore developing GRs wit…

RELATED ENTITIES

RELATED TOPICS