IamBishalIamBishal

Research Paper

Peer-ReviewedOpen Access

A Hybrid Approach for Intelligent Document Processing Using LLMs and Knowledge Graphs

A framework that combines Large Language Models (LLMs) with knowledge graphs for intelligent document processing, improving information extraction accuracy and enabling contextual understanding across domains.

Bishal Saha1, Anirban Das2, R. K. Sharma3

  1. 1 Independent Researcher, India
  2. 2 Department of Computer Science, XYZ University, India
  3. 3 School of Computing, ABC Institute, India

International Journal of Advanced Computer Science and Applications (IJACSA)

Vol. 15, No. 6, pp. 112–125, 2024

Download PDF (2.4 MB)Cite

Abstract

This paper presents a hybrid framework that combines Large Language Models (LLMs) with knowledge graphs for intelligent document processing. Our approach improves information extraction accuracy, enables contextual understanding across domains, and demonstrates significant performance gains over traditional methods.

We evaluate the framework on real-world datasets and show its applicability in enterprise document processing scenarios such as contracts, reports, and research articles.

Key Highlights

  • Hybrid Architecture

    Combines LLMs with knowledge graphs

  • Improved Accuracy

    Higher information extraction performance

  • Multi-Domain Applicability

    Works across various document types

  • Real-World Evaluation

    Tested on enterprise-grade datasets

Figures

Fig. 1: Overall system architecture
Fig. 2: Knowledge graph representation
Fig. 3: Performance comparison

Cite This Paper

Saha, B., Das, A., & Sharma, R. K. (2024). A Hybrid Approach for Intelligent Document Processing Using LLMs and Knowledge Graphs. International Journal of Advanced Computer Science and Applications (IJACSA), 15(6), 112–125. https://doi.org/10.14569/IJACSA.2024.0150612
Research is not just about new knowledge, but about creating real impact.
— Bishal Saha