Graph processing using a document store

Patent metadata

Field	Value
Publication number	US-12406003-B2
Application number	US-202318531298-A
Country	US
Kind code	B2
Filing date	Dec 6, 2023
Priority date	Dec 6, 2023
Publication date	Sep 2, 2025
Grant date	Sep 2, 2025

Abstract

The present invention relates to an improved database system architecture designed for generating and managing heterogeneous graphs. The system utilizes a document store in conjunction with a persistent adjacency index to enhance flexibility, adaptability, and transactional visibility during the generation and modification of complex graph structures. Data stored in the document store utilizes a flexible schema for storing data which improves the efficiency in both storing and retrieving data from the document store during graph-related processing tasks.

First claim

What is claimed is:

1. A database system for generating a heterogeneous graph, the database system comprising:
a memory configured to store a document store, wherein the document store comprises a JSON storage and an adjacency index, wherein the JSON storage is configured to store graph-related data in one or more documents, wherein each document of the one or more documents represents a node within the heterogeneous graph, wherein the adjacency index is implemented as a persistent cache, and wherein the adjacency index is configured to store mapping information between nodes within the heterogeneous graph, wherein the adjacency index comprises a hash set with a hash map and a row index, and wherein the hash map is configured to store the mapping information between nodes within the heterogeneous graph, the row index is configured to store a sequential ordering of the nodes within the heterogeneous graph; and
a processor in communication with the memory and configured to cause a graph engine configured to:
store a graph script procedure for generating a heterogeneous table, wherein the graph script procedure identifies the one or more documents in the JSON storage; and
initiate the graph script procedure in a graph workspace, wherein the initiating provides one or more graph-related operations for processing the graph-related data;
wherein the processor is further configured to cause an adapter to:
receive the one or more graph-related operations from the graph engine;
translate the one or more graph-related operations into a task to be performed by the document store; and
transmit the task to the document store;
wherein the document store is further configured to:
perform execution of the task to process the graph-related data based on the graph-related data of the one or more documents; and
return an output of the execution of the task to the adapter.

2. The database system of claim 1 , wherein the execution of the task comprises:
querying the one or more documents in the JSON storage;
receiving the graph-related data of the one or more documents;
identifying at least one of a node attribute or an edge relationship from the graph-related data; and
updating a node or an edge of the heterogeneous graph based on the at least one of the node attribute or the edge relationship.

3. The database system of claim 1 , wherein the graph-related data in the one or more documents is in a JSON format and the one or more documents are implemented as JSON documents.

4. The database system of claim 1 , wherein the one or more documents include attributes associated with one or more properties of nodes and edges of the heterogeneous graph.

5. The database system of claim 1 , wherein the document store is further configured to update the mapping information in the adjacency index responsive to a modification of any node of the nodes of the heterogeneous graph without requiring recomputation of the mapping information.

6. The database system of claim 5 , wherein the one or more graph-related operations further include the modification of any node of the nodes of the heterogeneous graph.

7. A method for generating a heterogeneous graph, the method comprising:
storing, in a document store comprising JSON storage and an adjacency index, graph-related data in one or more documents, wherein each document of the one or more documents represents a node within the heterogeneous graph, wherein the adjacency index is implemented as a persistent cache, and wherein the one or more documents are stored in the JSON storage, wherein the adjacency index comprises a hash set with a hash map and a row index, and wherein the hash map is configured to store mapping information between nodes within the heterogeneous graph, the row index is configured to store a sequential ordering of the nodes within the heterogeneous graph;
storing, in the adjacency index, the mapping information between nodes within the heterogeneous graph;
storing, by a graph engine, a graph script procedure for generating a heterogeneous table, wherein the graph script procedure identifies the one or more documents in the JSON storage;
initiating, by the graph engine, the graph script procedure in a graph workspace, wherein the initiating provides one or more graph-related operations for processing the graph-related data;
receiving, by an adapter communicatively coupled to the document store and the graph engine, the one or more graph-related operations from the graph engine;
translating, by the adapter, the one or more graph-related operations into a task to be performed by the document store;
transmitting, by the adapter, the task to the document store;
performing, by the document store, execution of the task to process the graph-related data based on the graph-related data of the one or more documents; and
returning an output of the execution of the task to the adapter.

8. The method of claim 7 , wherein the execution of the task comprises:
querying the one or more documents in the JSON storage;
receiving the graph-related data of the one or more documents;
identifying at least one of a node attribute or an edge relationship from the graph-related data; and
updating a node or an edge of the heterogeneous graph based on the at least one of the node attribute or the edge relationship.

9. The method of claim 7 , wherein the graph-related data in the one or more documents is in a JSON format and the one or more documents are implemented as JSON documents.

10. The method of claim 7 , wherein the one or more documents include attributes associated with one or more properties of nodes and edges of the heterogeneous graph.

11. The method of claim 7 , wherein the document store is further configured to update the mapping information in the adjacency index responsive to a modification of any node of the nodes of the heterogeneous graph without requiring recomputation of the mapping information.

12. The method of claim 11 , wherein the one or more graph-related operations further include the modification of any node of the nodes of the heterogeneous graph.

13. A non-transitory computer-readable device having instructions stored thereon that, when executed by at least one computing device, cause the at least one computing device to generate a heterogeneous graph by performing operations comprising:
storing, in a document store comprising JSON storage and an adjacency index, graph-related data in one or more documents, wherein each document of the one or more documents represents a node within the heterogeneous graph, wherein the adjacency index is implemented as a persistent cache, and wherein the one or more documents are stored in the JSON storage, wherein the adjacency index comprises a hash set with a hash map and a row index, and wherein the hash map is configured to store mapping information between nodes within the heterogeneous graph, the row index is configured to store a sequential ordering of the nodes within the heterogeneous graph;
storing, in the adjacency index, the mapping information between nodes within the heterogeneous graph, wherein the mapping information is persistently stored in the adjacency index;
storing, in a graph engine, a graph script procedure for generating a heterogeneous table, wherein the graph script procedure identifies the one or more documents in the JSON storage;
initiating the graph script procedure in a graph workspace, wherein the initiating provides one or more graph-related operations for processing the graph-related data;
receiving the one or more graph-related operations from the graph engine;
translating the one or more graph-related operations int

Assignees

• Sap Se

Inventors

• Bensberg Christian
• Belloni Stefano
• Hauck Matthias
• Kemeter Mathias
• Rawat Umang
• Sedler Roland
• Braenzel Alan
• Fath Markus
• Franz Gerald
• Ritter Daniel
• Tetzel Frank
• Jacob Marc
• Jakschitsch Hannes

Classifications

• G06F16/93

  Document management systems · CPC title

• G06F16/9024 Primary

  Graphs; Linked lists (G06F16/9027 takes precedence) · CPC title