A property in a graph database is a key-value pair that stores data directly on nodes (entities) or relationships (connections between entities). Unlike relational databases, where data is stored in fixed table columns, graph databases allow properties to be attached flexibly to individual nodes or relationships. For example, a node representing a person might have properties like name: "Alice", age: 30, or email: "alice@example.com". Similarly, a relationship like FRIENDS_WITH between two person nodes could include a property such as since: 2020. Properties provide context and details about the entities and their connections, enabling richer queries and more precise data modeling.
Graph databases use properties to support schema flexibility. While relational databases require predefined columns, graph databases let nodes or relationships of the same type have different properties. For instance, a Product node might have price and category properties, while another Product node in the same database could include an additional discount property without requiring structural changes. This flexibility is useful for evolving applications or datasets with heterogeneous attributes. Properties can also be indexed for faster lookups. For example, indexing the email property on Person nodes allows quick searches for users by their email address, similar to a primary key in a relational table.
Properties on relationships add depth to how connections are interpreted. A PURCHASED relationship between a User and a Product node might include date: "2023-10-05" and quantity: 2, turning a simple connection into a meaningful event with metadata. This contrasts with relational databases, where such details would require a separate join table. Properties also enable efficient traversal and filtering. For example, querying for “users who bought a product in the last week” can directly filter relationships using their date property, avoiding complex joins. By combining properties on nodes and relationships, graph databases model complex real-world scenarios with clarity, making them well-suited for social networks, recommendation engines, or fraud detection systems.