Inspect instances in Graph Explorer

Inspect the iot_machine, iot_sensor, and iot_reads_from data loaded in the previous lessons. The Data tabs in Modeling show table rows; Graph Explorer shows whether those rows became nodes connected by a directed relation.

Open Graph Explorer

1. Select Ontology → Graph Explorer in the left sidebar.
2. In the left Graph Structure panel, select the course collection.
3. Confirm the expected metadata.

The 2D / 3D control changes only the visualization of the same result. Start with 2D because nodes and edges are easier to select.

Query entities and relations

• Select iot_machine under Entities to query nodes with that label.
• Select iot_reads_from under Relations to query the relation with its Source and Target nodes.
• Select * in either section to query all entities or all relations in the selected collection.

With the single practice row loaded, the result summary should report two nodes and one relation. Select a node or edge to inspect its properties. The iot_machine caption uses the id field selected as its Display Column in Lesson 2.

Open Query Console

Hover over the code icon in the graph toolbar to confirm its Query Console tooltip, then select it. The Advanced Query panel opens below the graph with:

• A Cypher Query editor
• A Run button and execution time
• Result count
• A Table tab for rows
• A Text tab for structured return values

Use the visible Run button to execute the query.

Run basic Cypher

Neo4j labels and relationship types use the ontology's actual system names. Wrap them in backticks as shown below.

Query machine nodes

MATCH (m:`iot_machine`)
RETURN m
LIMIT 100

Query the sensor-to-machine relation

MATCH (s:`iot_sensor`)-[r:`iot_reads_from`]->(m:`iot_machine`)
RETURN s, r, m
LIMIT 50

Query sensors attached to one machine

MATCH (s:`iot_sensor`)-[r:`iot_reads_from`]->(m:`iot_machine`)
WHERE m.id = 'machine_001'
RETURN s.id, r.installed_at, m.id
LIMIT 20

If no result appears:

1. Confirm that each entity and relation has rows under Data in Modeling.
2. Confirm that Cypher labels and relationship types match the lowercase system names.
3. Confirm that the arrow points from iot_sensor to iot_machine.
4. If data was just loaded, wait for graph synchronization and run again.

Compare 2D and 3D

Switch the same result between 2D / 3D. 3D can help with spatial exploration of complex connections, but rendering many nodes and relations may show a performance warning. Use 2D first for structural and property checks.

Share a reproducible exploration

The collection selected in the UI and the latest query are not continuously persisted to the address bar. Share enough information for another user to reproduce the result:

• Collection name
• Complete Cypher query
• Expected node and relation counts
• Properties to inspect
• A result screenshot when useful

Self-check

• iot_machine, iot_sensor, and iot_reads_from appear for the collection.
• Querying the relation shows two nodes and one relation.
• Query Console opens and the query runs with the visible Run button.
• Both Table and Text results are inspected.
• Relation direction and installed_at are correct.

Next lesson

Finally, inspect which model changes are supported, assess their impact, and hand analysts reproducible queries and engineers an exact pipeline output contract.