Abstraction for Human Understanding
summary
Claims that human working memory is ~7±2 guide a workflow: store richly (120-column US flights table with diversion fields that occupy ~40% and admit up to five diversions), transform with code or a visual graph calculus (map, aggregate, link, extract, merge, shortcut) into concise views, and load insights into Neo4j knowledge graph via LLM (uploader notes). Concrete builds include a Neo4j pipeline from documents, a visualization toolkit that compresses transaction graphs and flight records, and applied examples that reveal no political "shifts" in Twitter data and clear money-flow pathways.
AI-written summary · about
chapters (7)
- 0:00Introduction and background
- 3:12Graphs for insight discovery
- 6:23From data storage to communication
- 10:46Challenges in data management
- 15:06Graph calculus for abstraction
- 20:01Simplifying complex data
- 23:01Practical implications and future
related talks
31mIs My Brain a Knowledge Graph? A Human Take on Connecting the Dots
25mKeynote: Think Faster: Graphs and AI for Real-Time Crisis Resolution
56mConnecting the Dots with Context Graphs
28mGrounded Intelligence Boosting LLMs with Real-Time Graph Context
30mAgent Reasoning with Graph World Models
30mNeurosymbolic Architecture in Production with Neo4j, GDS, Node2Vec & LLM Grounding
23mConnecting the Dots With Context Graphs
31mLive Q&A