A Framework for Dynamic Memory Management and Personalized Agent Decision-Making in Retrieval- Augmented Generation
Abstract
Abstract This paper proposes DMAP-RAG, a framework that integrates Dynamic Memory Management and Personalized Agent Decision-Making into Retrieval-Augmented Generation (RAG) to address challenges in memory retention and personalized adaptation within multi-turn dialogues. The framework includes a tree-structured dual-storage memory architecture, a user profile module for incremental attribute extraction, and an agentic decision-making core for adaptive retrieval and generation. Experimental results on standard benchmarks and a self-constructed personalized tutoring dataset suggest that DMAP-RAG performs competitively relative to several baseline approaches, achieving F1 scores of 78.9% on QASPER and 66.1% on HotPotQA, along with human evaluation scores of 4.4/5 for dialogue coherence on RiSAWOZ and 4.3/5 for user satisfaction. These findings indicate that integrating structured memory management with agent-based reasoning could offer a promising direction for developing context-aware and user-adaptive dialogue systems.
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.