All versions of this manual
X
 

Search: Pattern search

Most graph database support a query language that can be used to express pattern queries in the graph. Neo4j supports the Cypher language, JanusGraph and DataStax Enterprise Graph support the Gremlin language and AllegroGraph and Stardog support the SPARQL language.

In Linkurious Enterprise, you can directly use these query language from the "Find" > "Patterns" menu. Once a query is previewed (using the Preview button), you can click the "Add all" button to add all matching results to the current visualization.

When clicking on the button highlighted in the following picture, you can access the list of queries saved by you or shared by other users.

In the query list you can select a query to run and create, edit or delete existing queries.

With Neo4j: the Cypher query language

The Cypher query language is similar to SQL and can be learned from Neo4j's online documentation.

Here is an example Cypher query that is using the Crunchbase dataset (see our online demo):

MATCH (city:CITY)<-[hasCity:HAS_CITY]-(company:COMPANY)
WITH count(company) as score, city, company, hasCity
RETURN city, company, hasCity, score
ORDER BY score DESC
LIMIT 25

This query will match all companies that are connected to a city in the database, count the number of companies in each city, sort the sub-graphs by city and return the 25 cities that have the most companies, with their companies.

Note the Cypher query has to contain a RETURN statement and that all information that need to be displayed in Linkurious Enterprise must be included in the RETURN statement.