Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

PROPEX-RAG: Enhanced GraphRAG using Prompt-Driven Prompt Execution

Tejas Sarnaik, Manan Shah, Ravi Hegde

TL;DR

ProPEX-RAG tackles multi-hop question answering by fusing prompt-driven control with a symbolic knowledge graph and a Personalized PageRank-based retrieval scheme. It builds a persistent graph from LLM-extracted facts, selectively filters information with prompt-conditioned LLMs, and propagates seed entities via PPR to retrieve and re-rank passages for grounded answering. Answers are produced through a structured, evidence-focused prompt and deterministic decoding, reducing hallucinations and improving traceability. On HotpotQA and 2WikiMultiHopQA, ProPEX-RAG achieves state-of-the-art F1 and Recall@5, underscoring the importance of prompt design for accuracy and explainability in graph-augmented RAG systems.

Abstract

Retrieval-Augmented Generation (RAG) has become a robust framework for enhancing Large Language Models (LLMs) with external knowledge. Recent advances in RAG have investigated graph based retrieval for intricate reasoning; however, the influence of prompt design on enhancing the retrieval and reasoning process is still considerably under-examined. In this paper, we present a prompt-driven GraphRAG framework that underscores the significance of prompt formulation in facilitating entity extraction, fact selection, and passage reranking for multi-hop question answering. Our approach creates a symbolic knowledge graph from text data by encoding entities and factual relationships as structured facts triples. We use LLMs selectively during online retrieval to perform semantic filtering and answer generation. We also use entity-guided graph traversal through Personalized PageRank (PPR) to support efficient, scalable retrieval based on the knowledge graph we built. Our system gets state-of-the-art performance on HotpotQA and 2WikiMultiHopQA, with F1 scores of 80.7% and 78.9%, and Recall@5 scores of 97.1% and 98.1%, respectively. These results show that prompt design is an important part of improving retrieval accuracy and response quality. This research lays the groundwork for more efficient and comprehensible multi-hop question-answering systems, highlighting the importance of prompt-aware graph reasoning.

PROPEX-RAG: Enhanced GraphRAG using Prompt-Driven Prompt Execution

TL;DR

ProPEX-RAG tackles multi-hop question answering by fusing prompt-driven control with a symbolic knowledge graph and a Personalized PageRank-based retrieval scheme. It builds a persistent graph from LLM-extracted facts, selectively filters information with prompt-conditioned LLMs, and propagates seed entities via PPR to retrieve and re-rank passages for grounded answering. Answers are produced through a structured, evidence-focused prompt and deterministic decoding, reducing hallucinations and improving traceability. On HotpotQA and 2WikiMultiHopQA, ProPEX-RAG achieves state-of-the-art F1 and Recall@5, underscoring the importance of prompt design for accuracy and explainability in graph-augmented RAG systems.

Abstract

Retrieval-Augmented Generation (RAG) has become a robust framework for enhancing Large Language Models (LLMs) with external knowledge. Recent advances in RAG have investigated graph based retrieval for intricate reasoning; however, the influence of prompt design on enhancing the retrieval and reasoning process is still considerably under-examined. In this paper, we present a prompt-driven GraphRAG framework that underscores the significance of prompt formulation in facilitating entity extraction, fact selection, and passage reranking for multi-hop question answering. Our approach creates a symbolic knowledge graph from text data by encoding entities and factual relationships as structured facts triples. We use LLMs selectively during online retrieval to perform semantic filtering and answer generation. We also use entity-guided graph traversal through Personalized PageRank (PPR) to support efficient, scalable retrieval based on the knowledge graph we built. Our system gets state-of-the-art performance on HotpotQA and 2WikiMultiHopQA, with F1 scores of 80.7% and 78.9%, and Recall@5 scores of 97.1% and 98.1%, respectively. These results show that prompt design is an important part of improving retrieval accuracy and response quality. This research lays the groundwork for more efficient and comprehensible multi-hop question-answering systems, highlighting the importance of prompt-aware graph reasoning.

Paper Structure

This paper contains 13 sections, 4 equations, 1 figure, 3 tables.

Table of Contents

  1. Introduction
  2. Related Work:
  3. Proposed Methodology
  4. Symbolic Graph Construction (Offline Indexing)
  5. Entity-Guided Graph Traversal (Online Retrieval)
  6. Answer Generation:
  7. Results & discussion
  8. Baselines
  9. Implementation Details
  10. Dataset Description
  11. Evaluation Metrics:
  12. Results:
  13. Conclusion

Figures (1)

  • Figure 1: Architecture of our retrieval-augmented QA framework. Phase I constructs a symbolic knowledge graph from LLM-extracted entities and facts triples. Phase II performs PPR-based traversal using query-aligned seeds and filtered facts to retrieve and re-rank passages for grounded answer generation.