Graph databases do well with high volumes of complex, interconnected low-structure data which change rapidly and require frequent querying.
Neo4j represents data with nodes, similar to an object instance and with a unique ID; relationships or edges, which links two nodes and has a direction and RelationshipType; and properties, a string key/object value pair which can exist on nodes and relationships. And can navigate nodes and edges with equal speed, independently of the amount of data constituting the graph. Access to the graph structure is managed by an ACID transaction system, with graph traversal by a Traverser API. Neo4j is also integrated with Lucene for indexing support, and an integration with Solr is in process.