FastRAG: Retrieval Augmented Generation for Semi-structured Data

2024-11-21

arXiv.org

AI Summary

To address the challenges of semantic ambiguity, high computational overhead, and poor timeliness when applying retrieval-augmented generation (RAG) to semi-structured technical data (e.g., network management logs), this paper proposes a lightweight structured understanding framework. Our method innovatively integrates schema learning and script learning to enable efficient, large-model-free structured extraction without requiring full input sequences. Furthermore, we design a hybrid retrieval mechanism that synergistically combines textual search with knowledge graph querying, enabling context-aware and precise retrieval. Experimental results demonstrate substantial improvements: our approach achieves significantly higher question-answering accuracy compared to baselines; inference latency is reduced by up to 90% relative to GraphRAG; and computational cost decreases by up to 85%. The framework thus delivers a balanced trade-off among performance, efficiency, and practical deployability.

Technology Category

• Machine Learning: Structured Learning
• Data Mining & Knowledge Management: Conversational Systems for Recommendation & Retrieval
• Search and Optimization: Learning to Search

Application Category

• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Graph Algorithms and Modeling for the Web: Querying, indexing, and retrieval in Web-related graphs

Abstract

Efficiently processing and interpreting network data is critical for the operation of increasingly complex networks. Recent advances in Large Language Models (LLM) and Retrieval-Augmented Generation (RAG) techniques have improved data processing in network management. However, existing RAG methods like VectorRAG and GraphRAG struggle with the complexity and implicit nature of semi-structured technical data, leading to inefficiencies in time, cost, and retrieval. This paper introduces FastRAG, a novel RAG approach designed for semi-structured data. FastRAG employs schema learning and script learning to extract and structure data without needing to submit entire data sources to an LLM. It integrates text search with knowledge graph (KG) querying to improve accuracy in retrieving context-rich information. Evaluation results demonstrate that FastRAG provides accurate question answering, while improving up to 90% in time and 85% in cost compared to GraphRAG.

Problem

• Efficiently process complex semi-structured network data
• Overcome inefficiencies in existing RAG methods
• Improve accuracy and reduce time-cost in retrieval

Innovation

• Schema learning for semi-structured data extraction
• Combines text search with knowledge graph querying
• Reduces time and cost by avoiding full LLM submission