Knowledge Conceptualization Impacts RAG Efficacy

Abstract

Interpretability and adaptability are cornerstones of frontier and next-generation artificial intelligence (AI) systems. This is especially true in recent systems, such as large language models (LLMs), and more broadly, generative AI. As such, we are interested in how we can merge these efforts, that is, investigate the design of explainable and adaptable neurosymbolic AI systems. Specifically, we focus on a class of systems referred to as “Agentic Retrieval-Augmented Generation” systems, which actively select, interpret, and query knowledge sources in response to natural language prompts. In this paper, we systematically evaluate how different conceptualizations and representations of knowledge, particularly the structure and complexity, impact an LLM agent in effectively querying a triplestore. Our results show an impact from both approaches, and we discuss them and their implications.