The field of conversational recommender systems and generative recommendation is rapidly evolving, with a focus on improving the accuracy and efficiency of recommendation models. Recent developments have highlighted the importance of incorporating external domain knowledge, contextual information, and semantic understanding into these models. The use of large language models, multimodal fusion, and dynamic alignment has shown promise in enhancing the performance of recommender systems. Furthermore, the integration of techniques such as masked history learning, entropy-guided masking, and curriculum learning has improved the ability of models to capture complex user behaviors and preferences. Noteworthy papers in this area include ReGeS, which proposes a reciprocal retrieval-generation synergy framework for conversational recommender systems, and SynerGen, which introduces a novel generative recommender model that bridges the gap between personalized search and recommendation. Additionally, papers such as GoalRank and FreeRet have made significant contributions to the development of more effective and efficient ranking and retrieval models.
Advancements in Conversational Recommender Systems and Generative Recommendation
Sources
ReGeS: Reciprocal Retrieval-Generation Synergy for Conversational Recommender Systems
SynerGen: Contextualized Generative Recommender for Unified Search and Recommendation
GoalRank: Group-Relative Optimization for a Large Ranking Model
Does Generative Retrieval Overcome the Limitations of Dense Retrieval?
From Past To Path: Masked History Learning for Next-Item Prediction in Generative Recommendation
PartnerMAS: An LLM Hierarchical Multi-Agent Framework for Business Partner Selection on High-Dimensional Features
Multi-Item-Query Attention for Stable Sequential Recommendation
FreeRet: MLLMs as Training-Free Retrievers
Understanding Generative Recommendation with Semantic IDs from a Model-scaling View
HiFIRec: Towards High-Frequency yet Low-Intention Behaviors for Multi-Behavior Recommendation
SETR: A Two-Stage Semantic-Enhanced Framework for Zero-Shot Composed Image Retrieval
Reinforced Strategy Optimization for Conversational Recommender Systems via Network-of-Experts
Items Proxy Bridging: Enabling Frictionless Critiquing in Knowledge Graph Recommendations
Leveraging Scene Context with Dual Networks for Sequential User Behavior Modeling
SQUARE: Semantic Query-Augmented Fusion and Efficient Batch Reranking for Training-free Zero-Shot Composed Image Retrieval
Bridging Collaborative Filtering and Large Language Models with Dynamic Alignment, Multimodal Fusion and Evidence-grounded Explanations
AgentRec: Next-Generation LLM-Powered Multi-Agent Collaborative Recommendation with Adaptive Intelligence
LLM4Rec: Large Language Models for Multimodal Generative Recommendation with Causal Debiasing
TalkPlay-Tools: Conversational Music Recommendation with LLM Tool Calling
Tree-based Dialogue Reinforced Policy Optimization for Red-Teaming Attacks