Add event based data model

modules/ROOT/attachments/transactions-base-model.txt → modules/ROOT/attachments/llm-transaction-base-model.txt

@@ -354,7 +354,25 @@ CALL db.index.vector.createNodeIndex(
 CREATE FULLTEXT INDEX customer_name_idx IF NOT EXISTS FOR (c:Customer) ON EACH [c.firstName, c.lastName, c.middleName];
 ```
 
-## 4. Minimal Demo Code
+## 4. Customer enhancements
+
+**Location Data Source Enhancement:**
+
+While reviewing this data model, we identified an opportunity to improve location accuracy for fraud detection scenarios. Currently, the model derives location data from IP addresses, but mobile device GPS coordinates provide significantly higher precision when available.
+
+**Proposed Enhancement:**
+
+- Establish a direct `LOCATED_IN` relationship from `Session` nodes to `Location` nodes
+- Add a `source` property to the relationship indicating data origin (e.g., "IP", "Device_GPS", "WiFi_Network")
+- This would allow the system to prioritize more accurate location data while maintaining fallback options
+
+**Benefits:**
+
+- Enhanced fraud detection through precise geolocation
+- Flexible data source hierarchy (GPS > WiFi > IP)
+- Maintains backward compatibility with IP-based location data
+
+## 5. Minimal Demo Code
 
 The following Cypher code creates a complete example dataset demonstrating all the main entities and relationships in the model. This code is suitable for testing and demonstration purposes.
 

modules/ROOT/content-nav.adoc

@@ -27,4 +27,6 @@
 ** xref:agnostic/it-service-graph/it-service-graph.adoc[]
 
 * xref:data-models/index.adoc[]
-** xref:data-models/transactions/transactions-base-model.adoc[]
+** xref:data-models/transaction-graph/index.adoc[]
+*** xref:data-models/transaction-graph/transaction/transaction-base-model.adoc[Base Model]
+*** xref:data-models/transaction-graph/fraud-event-sequence/fraud-event-sequence-model.adoc[Ext. Fraud Event Sequence]