US 12,254,004 B2
Database query generation from natural language statements
Amit Aggarwal, Los Altos, CA (US)
Assigned to DataIris Platform, Inc., Palo Alto, CA (US)
Filed by DataIris Platform, Inc., Los Altos, CA (US)
Filed on Mar. 22, 2023, as Appl. No. 18/188,083.
Claims priority of provisional application 63/484,302, filed on Feb. 10, 2023.
Prior Publication US 2024/0273090 A1, Aug. 15, 2024
Int. Cl. G06F 16/242 (2019.01); G06F 16/22 (2019.01); G06F 16/2452 (2019.01); G06F 16/2453 (2019.01)
CPC G06F 16/243 (2019.01) [G06F 16/2246 (2019.01); G06F 16/24522 (2019.01); G06F 16/24539 (2019.01)] 18 Claims
OG exemplary drawing
 
1. A method for accessing one or more databases comprising:
receiving, by a computing device, a natural language statement;
translating, by the computing device, the natural language statement into an initial knowledge graph language (KGL) statement having a structure that is programmatically translatable into a query language that is executable with respect to one or more databases;
processing, by the computing device, the initial KGL statement to obtain a final KGL statement by replacing one or more first references in the initial KGL statement with one or more second references, each reference of the one or more second references referring to at least one of a table and an operation with respect to the one or more databases; and
generating, by the computing device, a query from the final KGL statement according to the query language;
wherein translating the natural language statement into the initial KGL statement comprises translating the natural language statement using a logic learning model and a knowledge graph mapping a plurality of tables of the one or more databases and a plurality of operations with respect to the plurality of tables to natural language labels, the plurality of tables including the table and the plurality of operations including the operation.