Recent advancements in recommendation systems are increasingly focused on enhancing personalization and scalability through innovative modeling techniques. Researchers are leveraging large language models to generate user and item profiles from reviews, capturing nuanced decision-making factors that traditional methods often overlook. This shift towards utilizing multi-factor attention mechanisms aims to improve recommendation robustness, particularly in domains like restaurant and industrial commodity recommendations. Additionally, the integration of orthogonal constrained projection methods is addressing challenges related to sparse scaling, enhancing representation quality and generalization across massive item sets. In the realm of generative recommendation, new frameworks are emerging that combine semantic ID learning with multimodal data, improving the coherence and effectiveness of recommendations. Furthermore, the development of interactive decision support systems is enabling more dynamic user engagement by managing ambiguity in preferences. Collectively, these trends signal a move towards more sophisticated, user-centric recommendation systems capable of driving significant commercial outcomes across diverse platforms.
In recommender systems, large language models (LLMs) have gained popularity for generating descriptive summarization to improve recommendation robustness, along with Graph Convolution Networks. Howeve...
LLMs are increasingly applied to recommendation, retrieval, and reasoning, yet deploying a single end-to-end model that can jointly support these behaviors over large, heterogeneous catalogs remains c...
Semantic ID learning is a key interface in Generative Recommendation (GR) models, mapping items to discrete identifiers grounded in side information, most commonly via a pretrained text encoder. Howev...
We present LLaTTE (LLM-Style Latent Transformers for Temporal Events), a scalable transformer architecture for production ads recommendation. Through systematic experiments, we demonstrate that sequen...
Accurately capturing feature interactions is essential in recommender systems, and recent trends show that scaling up model capacity could be a key driver for next-level predictive performance. While ...
In industrial commodity recommendation systems, the representation quality of Item-Id vocabularies directly impacts the scalability and generalization ability of recommendation models. A key challenge...
Users on e-commerce platforms can be uncertain about their preferences early in their search. Queries to recommendation systems are frequently ambiguous, incomplete, or weakly specified. Agentic syste...
Multimodal recommendation systems (MRS) jointly model user-item interaction graphs and rich item content, but this tight coupling makes user data difficult to remove once learned. Approximate machine ...
Multi-Behavior Recommendation (MBR) leverages multiple user interaction types (e.g., views, clicks, purchases) to enrich preference modeling and alleviate data sparsity issues in traditional single-be...
Generative Recommender Systems using semantic ids, such as TIGER (Rajput et al., 2023), have emerged as a widely adopted competitive paradigm in sequential recommendation. However, existing architectu...