US 12,443,631 B1
Machine learning model-based resolution of location information
Govind, Bengaluru (IN); Sayan Putatunda, Bangalore (IN); Saurabh Diwakar Sohoney, Ujjain (IN); and Amber Roy Chowdhury, Bellevue, WA (US)
Assigned to Amazon Technologies, Inc., Seattle, WA (US)
Filed by Amazon Technologies, Inc., Seattle, WA (US)
Filed on Sep. 16, 2024, as Appl. No. 18/886,672.
Int. Cl. G06F 16/00 (2019.01); G06F 16/23 (2019.01); G06F 16/29 (2019.01)
CPC G06F 16/29 (2019.01) [G06F 16/2365 (2019.01)] 20 Claims
OG exemplary drawing
 
1. A system comprising:
one or more processors; and
one or more memories storing a program code, wherein upon execution of the program code by the one or more processors, the system is configured to:
receive a natural language input that describes an address;
determine that the address is not mapped in a database to a geocode;
generate, using a machine learning (ML) model, an embedding vector of the address in an embedding space, the ML model trained to embed addresses and geocodes in the embedding space;
determine, by at least using the embedding vector in a query of the database, a set of known geocodes, the database storing known geocodes and embedding vectors, the embedding vectors belonging to the embedding space, generated by the ML model prior to the natural language input being received, and corresponding to at least one of the known geocodes or known addresses;
generate map data indicating an area encompassing the set of known geocodes; and cause the map data to be presented on a user interface, the presentation indicating that the address is within the area.