Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

PersonaRAG: Enhancing Retrieval-Augmented Generation Systems with User-Centric Agents

Saber Zerhoudi, Michael Granitzer

TL;DR

PersonaRAG introduces a user-centric, multi-agent extension to retrieval-augmented generation that continuously adapts both retrieval and generation using live user data. By deploying specialized agents (profile, contextual retrieval, live session, document ranking, and feedback) and a Global Message Pool, the framework integrates user context and feedback into each step of the QA process. Experimental results show PersonaRAG improves accuracy across multiple QA datasets and generalizes to other LLM architectures, while maintaining personalization advantages over traditional RAG. The work highlights practical benefits for personalized information retrieval and identifies trade-offs in latency and cost, pointing to future efficiency and expansion of agent capabilities.

Abstract

Large Language Models (LLMs) struggle with generating reliable outputs due to outdated knowledge and hallucinations. Retrieval-Augmented Generation (RAG) models address this by enhancing LLMs with external knowledge, but often fail to personalize the retrieval process. This paper introduces PersonaRAG, a novel framework incorporating user-centric agents to adapt retrieval and generation based on real-time user data and interactions. Evaluated across various question answering datasets, PersonaRAG demonstrates superiority over baseline models, providing tailored answers to user needs. The results suggest promising directions for user-adapted information retrieval systems.

PersonaRAG: Enhancing Retrieval-Augmented Generation Systems with User-Centric Agents

TL;DR

PersonaRAG introduces a user-centric, multi-agent extension to retrieval-augmented generation that continuously adapts both retrieval and generation using live user data. By deploying specialized agents (profile, contextual retrieval, live session, document ranking, and feedback) and a Global Message Pool, the framework integrates user context and feedback into each step of the QA process. Experimental results show PersonaRAG improves accuracy across multiple QA datasets and generalizes to other LLM architectures, while maintaining personalization advantages over traditional RAG. The work highlights practical benefits for personalized information retrieval and identifies trade-offs in latency and cost, pointing to future efficiency and expansion of agent capabilities.

Abstract

Large Language Models (LLMs) struggle with generating reliable outputs due to outdated knowledge and hallucinations. Retrieval-Augmented Generation (RAG) models address this by enhancing LLMs with external knowledge, but often fail to personalize the retrieval process. This paper introduces PersonaRAG, a novel framework incorporating user-centric agents to adapt retrieval and generation based on real-time user data and interactions. Evaluated across various question answering datasets, PersonaRAG demonstrates superiority over baseline models, providing tailored answers to user needs. The results suggest promising directions for user-adapted information retrieval systems.
Paper Structure (35 sections, 3 figures, 4 tables)

This paper contains 35 sections, 3 figures, 4 tables.

Table of Contents

  1. Introduction
  2. Related Work
  3. Methodology
  4. Fundamentals of Retrieval-Augmented Generation (RAG) Models
  5. PersonaRAG: RAG with User-Centric Agents
  6. User Interaction Analysis
  7. Cognitive Dynamic Adaptation
  8. User Profile Agent
  9. Contextual Retrieval Agent
  10. Live Session Agent
  11. Document Ranking Agent
  12. Feedback Agent
  13. PersonaRAG Operational Workflow
  14. Experimental Setups
  15. Datasets
  16. ...and 20 more sections

Figures (3)

  • Figure 1: Illustrations of Various RAG Models. Vanilla RAG and Chain-of-Thought YuZPMWY23 use passive learning, while PersonaRAG involves user-centric knowledge acquisition.
  • Figure 2: Overview of Our PersonaRAG Model showcasing the dynamic interaction among specialized agents within the system, facilitated by a global message pool for structured communication. The diagram illustrates the flow from user query input through various agents, including User Profile, Context Retrieval, Session Analysis, Document Ranking, and Feedback Agents, highlighting their contributions to real-time adaptation and personalized content generation by integrating live user data and feedback for continuous improvement and contextually relevant search experiences.
  • Figure 3: Text Similarity between Chain-of-Note (CoN) and Other Methods Using BLEU-2 Score for Evaluation, with Normalized Average Sentence Length and Average Syllable Count.