Graph Explorer

Graph Explorer is a tool to visually inspect and analyze completed ontology data. Data accumulated in the graph database is searched using Cypher queries and drawn on the screen. The page icon has a shape that emphasizes nodes and edges (relationships).

URL deep link

Graph Explorer pre-fills the startup status with the value appended to the web address (URL). Therefore, you can share the navigation screen starting from a specific node or query with a single link. It is convenient when exchanging data checks or research requests.

Screen composition

1. Metadata panel

The left panel shows graph structure information for the currently selected collection.

• Labels: List of defined entity labels (e.g. Person, Company).
  • The number of nodes is displayed next to each label.
• Relationship Types: List of defined relationship types (e.g. BOUGHT, MANAGED).
  • The number of relationships is displayed next to each type.
• Property Keys: List of property keys currently in use.
• When you click on an item, a query that searches all data of that type is automatically executed.

2. Visualization area

The middle screen shows the searched data as a graph consisting of nodes and links.

• Screen manipulation: You can drag nodes to move them or zoom in/out.
• Detailed Information: When you click on a node or link, the property information of that element appears.
• Continue navigation: Double-clicking a node brings up additional neighboring nodes connected to that node.
• Placement method: Nodes are automatically positioned as if they are pushing and pulling each other (automatic placement, force-directed), and can be moved manually if necessary.

3. Query editor

Write and run Cypher queries directly from the bottom panel.

• Query input: Supports Cypher, a query language widely used in graph databases.
• Run: When you run the query, the results are reflected in the visualization area.
• View results: You can check the results in graph, table, or text format.

Cypher Query Guide

Basic pattern matching

-- Query all Person nodes
MATCH (p:Person)
RETURN p
LIMIT 100

-- Query nodes connected by a specific relationship
MATCH (p:Person)-[:BOUGHT]->(product:Product)
RETURN p, product
LIMIT 50

-- Filter by condition
MATCH (p:Person)-[:BOUGHT]->(product)
WHERE p.age > 30 AND product.price > 10000
RETURN p, product

Aggregate queries

-- Count nodes by label
MATCH (n)
RETURN labels(n) as label, count(*) as count
ORDER BY count DESC

-- Statistics by relationship
MATCH ()-[r]->()
RETURN type(r) as relationship, count(*) as count
ORDER BY count DESC

-- Customer with the most purchases
MATCH (p:Person)-[:BOUGHT]->(product)
RETURN p.name, count(product) as purchase_count
ORDER BY purchase_count DESC
LIMIT 10

Path navigation

-- Shortest path between two nodes
MATCH path = shortestPath(
  (a:Person {name: "Alice"})-[*]-(b:Person {name: "Bob"})
)
RETURN path

-- Query N-hop neighbors
MATCH (p:Person {name: "Alice"})-[*1..3]-(neighbor)
RETURN DISTINCT neighbor
LIMIT 50

Useful Cypher functions

function	Description	Example
count()	Counting	count(n)
collect()	Collect by list	collect(n.name)
labels()	node label lookup	labels(n)
type()	Relationship type query	type(r)
properties()	View all properties	properties(n)
size()	list size	size(collect(n))
coalesce()	NULL replacement value	coalesce(n.name, "Unknown")

Result View Mode

Graph View

• Nodes and relationships are shown graphically.
• Node color is automatically determined according to the label.
• Node size can be adjusted according to attribute values.

table view

• Shows query results in table format.
• You can sort or filter columns.

text view

• Results are shown in unprocessed JSON format.
• Useful for investigating problems or checking data in detail.

Data Exploration Tips

• If you only want to see nodes with a specific label, click on that label in the metadata panel.
• To find complex conditions, write the MATCH statement directly in the query editor.
• If there are too many nodes and it is complicated, reduce the number of queries using the LIMIT clause.
• To speed up the search, filter the search by indexed properties as much as possible.