neo4j
  1. neo4j-data-modelling
PreviousNext

Data Modelling in GraphDB

Data modelling in GraphDB is the process of designing the structure of a graph database. It involves identifying the entities, relationships, properties, constraints, and hierarchies of the data to be stored in the graph database. This process is critical to achieving high performance, scalability, and accuracy in querying and manipulating data.

Syntax

The syntax for data modelling in GraphDB depends on the specific schema language being used. GraphDB supports several schema languages including RDF Schema (RDFS), Web Ontology Language (OWL), OWL 2, and SHACL. Here is an example of a basic RDFS schema:

@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .

<http://example.com/person>      a rdfs:Class .
<http://example.com/hasName>     a rdfs:Property ;
            rdfs:domain         <http://example.com/person> ;
            rdfs:range          rdfs:Literal .
<http://example.com/hasEmail>    a rdfs:Property ;
            rdfs:domain         <http://example.com/person> ;
            rdfs:range          rdfs:Literal .

This schema defines a person entity that has two properties: hasName and hasEmail.

Example

Consider the following scenario where we want to model a social network in a graph database using GraphDB.

We have users who have friends, and users can post statuses and comments. Users can also follow other users, and receive notifications from their friends and the people they follow.

To model this scenario, we could define the following entities:

  • user: represents a user in the social network with properties such as username, email address, and password.
  • friendship: represents a friendship relationship between two users in the social network.
  • status: represents a user's status update, with properties such as date posted, content, and number of likes.
  • comment: represents a comment made on a user's status update, with properties such as date posted, content, and number of likes.
  • follow: represents a follower relationship between two users in the social network.
  • notification: represents a notification sent to a user about an event, with properties such as date sent, content, and type.

Here is an example of a data model for the social network using OWL:

@prefix : <http://example.com/social-network#> .
@prefix owl: <http://www.w3.org/2002/07/owl#> .

:User a owl:Class .
:Friendship a owl:ObjectProperty ;
            owl:domain :User ;
            owl:range :User .
:Status a owl:Class .
:Comment a owl:Class .
:Post a owl:Class ;
       owl:unionOf (:Status :Comment) .
:Author a owl:ObjectProperty ;
        owl:domain :Post ;
        owl:range :User .
:Follow a owl:ObjectProperty ;
         owl:domain :User ;
         owl:range :User ;
         owl:propertyChainAxiom (:Friendship owl:inverseOf).
:Notification a owl:Class .
:Notifies a owl:ObjectProperty ;
           owl:domain :Notification ;
           owl:range :User .

Explanation

In the example above, we define a User class that represents a user in the social network. We also define an ObjectProperty named Friendship that represents the friendship relationship between two users.

We then define a Class named Status that represents a user's status update, and a Class named Comment that represents a comment made on a user's status update. We also define a Class named Post that represents both statuses and comments as a union.

We define an ObjectProperty named Author that relates a Post to a User.

We define another ObjectProperty named Follow that relates two User entities and it is a composition of the Friendship relationship follows that holds between the people who are friends (this is supported by the propertyChainAxiom).

Finally, we define a Class named Notification that represents a notification sent to a user about an event. We also define an ObjectProperty named Notifies that relates a notification to a user.

Use

Data modelling in GraphDB is critical to achieving high performance, scalability, and accuracy in querying and manipulating data. It is used to identify and design the structure, relationships, properties, and constraints of the data to be stored in the graph database.

Important Points

  • Data modelling in GraphDB involves identifying and designing the structure, relationships, properties, and constraints of the data to be stored in the graph database.
  • GraphDB supports several schema languages including RDF Schema (RDFS), Web Ontology Language (OWL), OWL 2, and SHACL.
  • Data modelling is critical to achieving high performance, scalability, and accuracy in querying and manipulating data.

Summary

Data modelling in GraphDB is a critical process that involves identifying and designing the structure, relationships, properties, and constraints of the data to be stored in the graph database. GraphDB supports several schema languages including RDF Schema (RDFS), Web Ontology Language (OWL), OWL 2, and SHACL. Effective data modelling is critical to achieving high performance, scalability, and accuracy in querying and manipulating data.

GraphDB vs NoSQLWhat is Neo4j CQL
Published on: