Persona-DB: Efficient Large Language Model Personalization for Response Prediction with Collaborative Data Refinement

2024-02-16

International Conference on Computational Linguistics

AI Summary

To address challenges in personalized response prediction for large language models—including difficulty modeling user preferences, low retrieval efficiency under resource constraints, and poor cold-start performance—this paper proposes a hierarchical collaborative data representation framework. Methodologically, it designs a hierarchical vector index and database structure to jointly encode general knowledge and individual preferences; introduces user collaborative representation learning and context-aware embedding refinement to mitigate data sparsity via cross-user knowledge complementarity; and integrates retrieval-augmented generation (RAG) for efficient low-rank retrieval. Its primary contribution is the first collaborative data refinement paradigm, significantly enhancing cross-task generalization and knowledge sharing. Experiments demonstrate stable response accuracy even with minimal retrieval size, over 10% improvement in cold-start scenarios, and substantially reduced context-length requirements—yielding exceptional practicality in long-history and limited-context settings.

Technology Category

• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Machine Learning: Learning Preferences or Rankings
• Knowledge Representation and Reasoning: Preferences

Application Category

• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

The increasing demand for personalized interactions with large language models (LLMs) calls for methodologies capable of accurately and efficiently identifying user opinions and preferences. Retrieval augmentation emerges as an effective strategy, as it can accommodate a vast number of users without the costs from fine-tuning. Existing research, however, has largely focused on enhancing the retrieval stage and devoted limited exploration toward optimizing the representation of the database, a crucial aspect for tasks such as personalization. In this work, we examine the problem from a novel angle, focusing on how data can be better represented for more data-efficient retrieval in the context of LLM customization. To tackle this challenge, we introduce Persona-DB, a simple yet effective framework consisting of a hierarchical construction process to improve generalization across task contexts and collaborative refinement to effectively bridge knowledge gaps among users. In the evaluation of response prediction, Persona-DB demonstrates superior context efficiency in maintaining accuracy with a significantly reduced retrieval size, a critical advantage in scenarios with extensive histories or limited context windows. Our experiments also indicate a marked improvement of over 10% under cold-start scenarios, when users have extremely sparse data. Furthermore, our analysis reveals the increasing importance of collaborative knowledge as the retrieval capacity expands.

Problem

• Personalization
• Efficiency
• User Preference Understanding

Innovation

• Persona-DB
• Hierarchical Data Organization
• Enhanced Retrieval Capability