## Chat Transcript
Exploring the SAWGraph Spatial Knowledge Graph
👤 **User**
@spatialkg
---
🧠 **Assistant**
I'll help you explore the spatialkg knowledge graph. Let me first get information about this knowledge graph to understand what data it contains.
*[Called spatialkg-dev:get_description]*
Great! Now let me get the schema to show you what kinds of data and relationships are available in this knowledge graph.
*[Called spatialkg-dev:get_schema with compact=true]*
# SAWGraph Spatial KG
Welcome! This is the **SAWGraph Spatial Knowledge Graph**, a large-scale geospatial knowledge graph covering the 48 contiguous United States. It contains 756.9 million triples describing 16.8 million spatial entities.
## What's in this knowledge graph?
The knowledge graph contains:
- **7.4 million S2 cells** at Level 13 (~1.27 km² each) - a hierarchical grid system covering the entire area
- **Administrative regions** at three levels:
- Level 1: 102 states
- Level 2: 6,228 counties
- Level 3: 35,458 county subdivisions (towns, townships, etc.)
All these spatial entities are connected through various spatial relationships, making it possible to query geographic relationships, find nearby areas, understand administrative hierarchies, and more.
## What can you do with it?
Here are some example queries you could run:
- Find all counties in a particular state
- Get S2 cells connected to a specific administrative region
- Explore spatial relationships (contains, touches, intersects, etc.)
- Look up regions by FIPS codes
- Find administrative hierarchies
Would you like me to:
1. Run a sample query to show you how the data looks?
2. Help you find specific geographic information?
3. Visualize the schema in a diagram?
4. Something else?
What would you like to explore?
---
👤 **User**
Visualize the schema
---
🧠 **Assistant**
Let me create a visual schema diagram for you.
*[Called bash_tool to create initial Mermaid diagram draft]*
*[Called spatialkg-dev:clean_mermaid_diagram to clean the diagram]*
*[Created file at /mnt/user-data/outputs/spatialkg-schema.mermaid]*
*[Called present_files to share the diagram]*
Here's the schema visualization for the SAWGraph Spatial Knowledge Graph:
```mermaid
classDiagram
direction TB
class SpatialObject {
+spatialRelation
}
class Feature {
+hasGeometry
+defaultGeometry
}
class Geometry {
+asWKT
+hasMetricArea
}
class Region
class AdministrativeRegion {
+hasFIPS
+administrativePartOf
}
class AdministrativeRegion_0
class AdministrativeRegion_1
class AdministrativeRegion_2
class AdministrativeRegion_3
class Cell {
+cellID
}
class S2Cell
class S2Cell_Level13
class Place
class PointOfInterest
class StatisticalArea
class ZipCodeArea
SpatialObject <|-- Feature
Feature <|-- Region
Region <|-- AdministrativeRegion
Region <|-- Cell
Region <|-- Place
Region <|-- StatisticalArea
AdministrativeRegion <|-- AdministrativeRegion_0
AdministrativeRegion <|-- AdministrativeRegion_1
AdministrativeRegion <|-- AdministrativeRegion_2
AdministrativeRegion <|-- AdministrativeRegion_3
Cell <|-- S2Cell
S2Cell <|-- S2Cell_Level13
Place <|-- PointOfInterest
StatisticalArea <|-- ZipCodeArea
Feature --> Geometry : hasGeometry
Feature --> Geometry : defaultGeometry
Feature --> Feature : sfContains
Feature --> Feature : sfWithin
Feature --> Feature : sfIntersects
Feature --> Feature : sfTouches
Feature --> Feature : sfOverlaps
Feature --> Feature : sfCrosses
Feature --> Feature : sfEquals
Feature --> Feature : connectedTo
AdministrativeRegion --> AdministrativeRegion : administrativePartOf
```
The diagram shows the hierarchical structure of the knowledge graph, with key spatial relationships between features including containment, intersection, touching, and administrative hierarchies.
---
*Created by [mcp-proto-okn](https://github.com/sbl-sdsc/mcp-proto-okn) 0.4.0 for spatialkg on 2025-12-24*
claude-sonnet-4-20250514
