Summary
Keywords
Full Transcript
Every investigator faces the same challenge—connecting scattered facts and finding hidden patterns in complex data. Traditional analysis tools often fall short when dealing with the relational nature of investigative work, where context and connections are more important than isolated data points. In this session, Paweł will demonstrate Networks Notebook, an innovative system that transforms multimodal investigative data into knowledge graphs for automated analysis. You'll discover how this murder board-inspired solution processes documents and images into interconnected graph structures, then applies sophisticated analytical methods to uncover hidden insights. The presentation will showcase a two-stage analysis approach that combines precise graph queries with the interpretation of generative AI. Unlike traditional vector-based RAG architectures that flatten relationships and lose informational context, this graph-powered system preserves the relational nature of reality. You'll learn how ""cold"" graph analysis using Cypher queries identifies key structures and anomalies, which language models then interpret to generate actionable hypotheses that neither technology could produce alone. Attendees will explore practical applications through a live case study analyzing a crime novel, demonstrating how the system detects probable event sequences, information flow paths, knowledge gaps, and conflicting witness statements. You'll walk away understanding how to build automated knowledge graph extraction from text and images, implement graph-based RAG systems, and apply drill-down analysis techniques that follow each thread until the whole story unravels. The session covers practical implementation strategies for forensics, security, audit, legal, and medical applications, providing sufficient technical depth for immediate application in your projects. Speaker: Paweł Gołąb View Presentation: https://drive.google.com/file/d/1jWDJwMwDzKPehothKTrUnNYlJHp1BQns/view?usp=drive_link Resources: Get Started with Aura - https://bit.ly/3LOLrjh Deployment Center - https://bit.ly/4jOelM3 Ground AI Systems and Agents with Neo4j - https://bit.ly/4oVsnyb #nodes2025 #neo4j #graphdatabase #graphrag #knowledgegraph
