grafendatabases

A graph database is a database designed to store and query data whose relationships are a first-class concern. It represents data as a graph comprising nodes (entities), edges (relationships), and properties attached to both. In German-speaking contexts, graph databases are often called Grafendatenbanken.

In the property graph model, nodes and edges can have labels and key–value properties, and edges have types. Graph databases excel at traversing relationships, enabling queries that follow paths of arbitrary length. Some systems also support RDF and SPARQL for semantic graphs.

Querying is typically done with graph-specific languages such as Cypher, Gremlin, or SPARQL. The focus on traversal makes graph databases suitable for deeply connected data, such as social graphs or knowledge graphs.

Architectures range from single-node to distributed clusters. Many graph databases provide ACID transactions and strong consistency guarantees, with optimizations for traversals and index-free adjacency. Scaling strategies include clustering, sharding, and replication.

Common use cases include social networks, recommendation engines, fraud detection, network topology, and knowledge graphs. They complement traditional relational databases by handling complex relationships more efficiently, though they may require specialized data modeling and tooling.

History and ecosystem: Graph databases emerged from research in graph theory and object-graph mapping, with Neo4j among early popular systems. Today several open-source and commercial products exist, and graph analytics libraries complement transactional graph databases.