Generative Recommender System
An emerging recommendation paradigm that directly predicts discrete item code sequences through generative models to produce recommendations, avoiding the overhead of conventional large-scale item ID embedding tables.
Timeline (last 90 days)
Generative recommender systems are gradually becoming the mainstream paradigm for sequential recommendation, in which semantic IDs represent items as discrete code sequences and recommendations are generated by directly predicting these codes.
In offline evaluations of an industrial-grade generative recommendation system, EWM substantially improved recall and enhanced multiple behavioral proxy metrics.
All Facts (2)
Generative recommender systems are gradually becoming the mainstream paradigm for sequential recommendation, in which semantic IDs represent items as discrete code sequences and recommendations are generated by directly predicting these codes.
50%UnverifiedIn offline evaluations of an industrial-grade generative recommendation system, EWM substantially improved recall and enhanced multiple behavioral proxy metrics.
50%