Abstract: A method to perform query operations on nodes of large graphs distributed across multiple machines by applying a graph-query language that implements lazy evaluation techniques are disclosed. A method includes receiving a graph query expression from a client, wherein a graph comprises a plurality of edges linking a plurality of vertices, receiving a first request for evaluating the graph query expression, evaluating a partial result set for the graph query expression, and sending the partial result to the client. The partial result including at least one of a successor query and a predecessor query, wherein the successor query and the predecessor query enable evaluation of the graph query expression at a point in the graph query expression where the partial result evaluation terminated. A system and non-transitory computer readable medium are also disclosed.

Abstract: A system and method for efficient storage and searching of object state data and relationships at a given point of time is disclosed. A server system stores one or more data objects in a database associated with a server system. The server system receives an update to one or more data objects stored in the database. For a respective object, the server system determines one or more differences between the respective original data object and the respective updated data object, creates a change entry for the respective data object, and stores the created change entry in a change database at the server system. The server system receives a data query and searches the one or more change entries stored in the change database to identify the information associated with the received search query. The server system sends the identified information to the client system.