HighlightCurated by Aramai Editorial

Structured reflective reasoning for precise medical knowledge graph retrieval augmented generation

A research paper on using structured reflective reasoning to improve medical knowledge graph retrieval and augmented generation.

This paper proposes a method called Structured Reflective Reasoning (SRR) to enhance the precision of medical knowledge graph retrieval and augmentation. SRR combines a knowledge graph with a reflection mechanism to refine the retrieval results. The authors evaluate their approach on several medical datasets, demonstrating its effectiveness in improving retrieval accuracy.

Based on: Structured reflective reasoning for precise medical knowledge graph retrieval augmented generation · Health Information Science and Systems

HighlightCurated by Aramai Editorial

Hybrid Multi-Agent GraphRAG for E-Government: Towards a Trustworthy AI Assistant

A modular framework integrating standard RAG, embedding-based retrieval, and LLM-generated structured graphs for e-government question answering.

This paper introduces a hybrid multi-agent graph retrieval-augmented generation (GraphRAG) framework designed to enhance policy-focused question answering in e-government settings. The framework integrates standard RAG, embedding-based retrieval, real-time web search, and LLM-generated structured Graphs to optimize knowledge discovery from public e-government data. This approach aims to provide an overview of a hybrid architecture for operational deployment in e-government settings.

Based on: Hybrid Multi-Agent GraphRAG for E-Government: Towards a Trustworthy AI Assistant · Applied Sciences

HighlightCurated by Aramai Editorial

Automating construction contract review using knowledge graph-enhanced large language models

A paper that explores the use of knowledge graphs and large language models to automate construction contract review.

The authors propose a method for automating construction contract review using knowledge graphs and large language models. They enhance a pre-trained LLM with a knowledge graph, which is used to extract relevant information from contracts. The enhanced model is then evaluated on its ability to accurately identify issues in construction contracts.

Based on: Automating construction contract review using knowledge graph-enhanced large language models · Automation in Construction

HighlightCurated by Aramai Editorial

DO-RAG: A Domain-Specific QA Framework Using Knowledge Graph-Enhanced Retrieval-Augmented Generation

Proposes a hybrid QA framework integrating knowledge graph construction and semantic vector retrieval.

DO-RAG is a scalable and customizable QA framework that integrates multi-level knowledge graph construction with semantic vector retrieval. It employs an agentic chain-of-thought architecture to extract structured relationships from unstructured documents, constructing dynamic knowledge graphs. The system fuses graph and vector retrieval results to generate context-aware responses.

Based on: DO-RAG: A Domain-Specific QA Framework Using Knowledge Graph-Enhanced Retrieval-Augmented Generation

HighlightCurated by Aramai Editorial

LLaVA Needs More Knowledge: Retrieval Augmented Natural Language Generation with Knowledge Graph for Explaining Thoracic Pathologies

A paper proposing a novel Vision-Language framework for generating natural language explanations with knowledge graph augmentation.

The authors propose a framework that integrates a pre-trained LLaVA model with a knowledge graph-based datastore to generate accurate and informative natural language explanations for thoracic pathologies. The framework is designed as a plug-and-play module, allowing seamless integration with various model architectures. Three distinct frameworks are introduced and evaluated on the MIMIC-NLE dataset.

Based on: LLaVA Needs More Knowledge: Retrieval Augmented Natural Language Generation with Knowledge Graph for Explaining Thoracic Pathologies · Proceedings of the AAAI Conference on Artificial Intelligence

HighlightCurated by Aramai Editorial

Pythia-RAG: Retrieval-augmented generation over a unified multimodal knowledge graph for enhanced QA

A paper proposing Pythia-RAG, a retrieval-augmented generation model for question answering.

The authors introduce Pythia-RAG, a unified multimodal knowledge graph that combines retrieval and generation capabilities. This approach aims to enhance question-answering performance by leveraging the strengths of both methods. The paper presents experiments demonstrating the effectiveness of Pythia-RAG in various QA tasks.

Based on: Pythia-RAG: Retrieval-augmented generation over a unified multimodal knowledge graph for enhanced QA · Knowledge-Based Systems

HighlightCurated by Aramai Editorial

TagRAG: Tag-guided Hierarchical Knowledge Graph Retrieval-Augmented Generation

A framework for efficient global reasoning and scalable graph maintenance in knowledge graphs.

The authors propose a tag-guided hierarchical knowledge graph retrieval-augmented generation (RAG) framework, called TagRAG. It introduces two key components: Tag Knowledge Graph Construction and Tag-Guided Retrieval-Augmented Generation. This design aims to improve efficiency and adaptability in global reasoning and graph maintenance.

Based on: TagRAG: Tag-guided Hierarchical Knowledge Graph Retrieval-Augmented Generation · arXiv (Cornell University)

HighlightCurated by Aramai Editorial

TagRAG: Tag-guided Hierarchical Knowledge Graph Retrieval-Augmented Generation

A framework for efficient global reasoning and scalable graph maintenance in knowledge graphs.

This paper proposes a tag-guided hierarchical knowledge graph retrieval-augmented generation (RAG) framework called TagRAG.,TagRAG introduces two key components: Tag Knowledge Graph Construction and Tag-Guided Retrieval-Augmented Generation.,The framework is designed to address limitations of traditional RAG methods, such as inefficiencies in information extraction and costly resource consumption.

Based on: TagRAG: Tag-guided Hierarchical Knowledge Graph Retrieval-Augmented Generation · arXiv (Cornell University)

HighlightCurated by Aramai Editorial

Empowering LLMs by hybrid retrieval-augmented generation for domain-centric Q&A in smart manufacturing

A study on a hybrid framework for retrieval-augmented generation in smart manufacturing.

The paper proposes a hybrid knowledge graph-vector retrieval framework to enhance the performance of large language models in question-answering tasks. The approach combines structured knowledge graph metadata with unstructured vector retrieval and achieves high accuracy and contextual relevance. Evaluated on design for additive manufacturing tasks, the proposed method demonstrates its effectiveness.

Based on: Empowering LLMs by hybrid retrieval-augmented generation for domain-centric Q&A in smart manufacturing · Advanced Engineering Informatics

HighlightCurated by Aramai Editorial

Translating and Formalizing the MIRAGE Guidelines to a Prototype MIRAGE Ontology and DCAT3 Extension Vocabulary for Glycomics Data Management

A paper on formalizing MIRAGE guidelines into a prototype ontology and DCAT3 extension vocabulary for glycomics data management.

The authors present a comprehensive semantic formalization of MIRAGE guidelines using an integrated RDF ontology framework. The framework models glycan structures, biological specimens, analytical instruments, and experimental processes with formal OWL semantics and SHACL validation constraints. It enables automated quality assessment, federated data querying, and enhanced reproducibility in glycomics research.

Based on: Translating and Formalizing the MIRAGE Guidelines to a Prototype MIRAGE Ontology and DCAT3 Extension Vocabulary for Glycomics Data Management

HighlightCurated by Aramai Editorial

KA-RAG: Integrating Knowledge Graphs and Agentic Retrieval-Augmented Generation for an Intelligent Educational Question-Answering Model

A course-oriented question answering framework that integrates a structured Knowledge Graph with an Agentic Retrieval-Augmented Generation workflow.

The paper introduces KA-RAG, a QA framework combining symbolic graph reasoning with dense semantic retrieval. It achieves high retrieval accuracy and semantic consistency on a graduate-level Pattern Recognition course. User surveys show improvements in learning efficiency and satisfaction.

Based on: KA-RAG: Integrating Knowledge Graphs and Agentic Retrieval-Augmented Generation for an Intelligent Educational Question-Answering Model · Applied Sciences

HighlightCurated by Aramai Editorial

MedPix 2.0: A Comprehensive Multimodal Biomedical Data Set for Advanced AI Applications with Retrieval Augmented Generation and Knowledge Graphs

A comprehensive multimodal biomedical data set for advanced AI applications.

MedPix 2.0 is a large-scale biomedical dataset that combines medical scans with clinical reports and findings. It was developed to address the lack of high-quality data sets in the medical domain, which hinders the development of AI applications. The dataset includes a semi-automatic pipeline for extracting visual and textual data, followed by manual curation.

Based on: MedPix 2.0: A Comprehensive Multimodal Biomedical Data Set for Advanced AI Applications with Retrieval Augmented Generation and Knowledge Graphs · Data Science and Engineering