Advances in Conversational Recommendation and Reinforcement Learning

The field of conversational recommendation and reinforcement learning is moving towards more sophisticated and effective methods for capturing user preferences and optimizing recommendation systems. Recent developments have focused on improving the accuracy and robustness of conversational recommendation models, as well as enhancing the efficiency of reinforcement learning algorithms. Notably, researchers are exploring the use of large language models, uncertainty-aware methods, and hybrid approaches to address challenges such as reward hacking and data noise. These innovations have the potential to significantly improve the performance of recommendation systems and enable more effective human-computer interaction. Noteworthy papers include: Empowering Retrieval-based Conversational Recommendation with Contrasting User Preferences, which proposes a novel conversational recommender model that leverages contrasting user preferences to improve recommendation accuracy. Sharpe Ratio-Guided Active Learning for Preference Optimization in RLHF, which introduces an active learning approach to efficiently select prompt and preference pairs using a risk assessment strategy based on the Sharpe Ratio. LaViC: Adapting Large Vision-Language Models to Visually-Aware Conversational Recommendation, which integrates compact image representations into dialogue-based recommendation systems to capture visual attributes of products.

Sources

Empowering Retrieval-based Conversational Recommendation with Contrasting User Preferences

Sharpe Ratio-Guided Active Learning for Preference Optimization in RLHF

Exploring Data Scaling Trends and Effects in Reinforcement Learning from Human Feedback

Probabilistic Uncertain Reward Model: A Natural Generalization of Bradley-Terry Reward Model

Zero-Shot LLMs in Human-in-the-Loop RL: Replacing Human Feedback for Reward Shaping

LaViC: Adapting Large Vision-Language Models to Visually-Aware Conversational Recommendation

SalesRLAgent: A Reinforcement Learning Approach for Real-Time Sales Conversion Prediction and Optimization

RuleAgent: Discovering Rules for Recommendation Denoising with Autonomous Language Agents

Get the Agents Drunk: Memory Perturbations in Autonomous Agent-based Recommender Systems

Learning a Canonical Basis of Human Preferences from Binary Ratings

Rec-R1: Bridging Generative Large Language Models and User-Centric Recommendation Systems via Reinforcement Learning

LLM-Augmented Graph Neural Recommenders: Integrating User Reviews

Retrieval-Augmented Purifier for Robust LLM-Empowered Recommendation

Prompt Optimization with Logged Bandit Data

Built with on top of