[103] Why Knowledge Graphs Are Hot Right Now (Amy E. Hodler)

Video Description

Join our Meetup group: https://www.meetup.com/data-umbrella Why Knowledge Graphs Are Hot Right Now (Amy E. Hodler) ## About the Event Our talk will delve into the various facets of knowledge graphs, including their traditional uses for data integration, search, and discovery, as well as their modern applications in decision-making, analytics, and predictions. We will explore the sudden surge of interest in knowledge graphs and the interplay with AI trends. Additionally, we'll look beyond LLMs and RAG to gain a broader perspective on the role of knowledge graphs as we go forward in the data landscape. This session promises to provide a comprehensive understanding of why knowledge graphs are currently a hot topic in the tech industry. This presentation covers: What commonly defines a knowledge graph (and the surrounding disagreements) - Traditional uses for data integration, search, and discovery - Modern uses for decision-making, analytics, and predictions - The sudden surge of knowledge graphs - What Knowledge Graphs bring to AI and vice versa - Beyond LLMs and RAG ## Resources - Slides: https://github.com/data-umbrella/event-transcripts/blob/main/resources/knowledge-graphs.pdf - Graph Geeks: https://www.graphgeeks.org/ ## Timestamps 00:00 Data Umbrella introduction 04:00 Amy begins talk 04:50 Basics of Knowledge Graphs (KG) 06:45 Property graphs, Resource Description Graphs (RDF),(LPG=Label Property Graph) 10:40 Traditional Knowledge Graph Uses 16:29 Metrics and Markets 18:59 Graphy startups and investments surge 20:06 Market size for knowledge graphs 23:45 Definition of knowledge graph 25:02 Q: What do you think explains the graph database curve to go down after 2022 in the database engine chart? 26:15 Q: Do you have any further details on repurposed drugs using graph networks? 27:20 Q: Are graph networks used on biology modeling with AI? 28:46 Q: Are there some visual exploration tools for knowledge graphs? 29:40 Q: Using machine learning in graphs 29:54 What is driving the interest in knowledge graphs? 30:02 -- GenAI 33:16 -- Knowledge graph maturity in comparison to GenAI 34:31 -- RAG: Retrieval-Augmented Generation 36:54 Standard RAG vs Graph RAG 39:15 Graph RAG improves results but… 41:19 Gartner’s list of emerging technologies & trends reports 43:10 Q: How do you get started with graph analytics? 44:25 Q: Do knowledge graphs help overcome LLM limitations? 45:06 Q: How do you use graphs for RAG without similarity scoring? 46:16 What’s next? 46:37 Knowledge graph challenges we still have to address 51:40 Knowledge graph themes: Predictions for the next 6 months 55:30 Undercurrents to watch (graph frameworks, de-evolution of query languages, AI & graphs, graph BI, advanced graphs) 57:26 Vendor neutral graph community: GraphGeeks 58:07 Q: Can you talk about long term maintenance of graph database and pitfalls? 59:50 Q: Do you have suggestions for textbooks on knowledge graphs? 01:00:44 Q: What percentage of companies use knowledge graphs or hire vendors for KG? 01:01:50 Q: What is the role of knowledge graphs in the use of LLMs, ethically? 01:03:20 Q: Are there benchmarks for graph databases? ## About the Speaker Amy E. Hodler is graph advisor and consultant. With decades of commercial experience, including Neo4j and RelationalAI, she promotes graph technology in various industries and has founded the GraphGeeks community to foster cross-discipline interaction. Amy is highlighted as a distinguished speaker by G-Research and has authored/contributed to several books including Graph Algorithms (O’Reilly). - LinkedIn: https://www.linkedin.com/in/amyhodler/ #OpenSource #GraphDatabases #KnowledgeGraphs