US 11,756,140 B2
Method and system for routing path selection
Anthony Alexis Ghislain Barrier, Antibes (FR); Simon Nanty, Antibes (FR); Youri Thibault Marie Le Toquin, Nice (FR); Chakib Belgani, Valbonne (FR); and Erwan Viollet, Mougins (FR)
Assigned to Amadeus S.A.S., Biot (FR)
Filed by AMADEUS S.A.S., Biot (FR)
Filed on Nov. 16, 2020, as Appl. No. 17/98,893.
Prior Publication US 2022/0156859 A1, May 19, 2022
Int. Cl. G06Q 50/14 (2012.01); G06Q 30/0283 (2023.01); G06Q 30/0601 (2023.01); G06N 20/00 (2019.01)
CPC G06Q 50/14 (2013.01) [G06N 20/00 (2019.01); G06Q 30/0284 (2013.01); G06Q 30/0627 (2013.01)] 26 Claims
OG exemplary drawing
 
1. A method comprising:
at a first device comprising one or more processors:
obtaining historical travel data of selected fare paths, wherein the historical travel data comprises at least one historical characteristic of a plurality of historical characteristics for each selected fare path;
storing the historical travel data of selected fare paths in a data storage, wherein a machine learning model is trained at a second device based on the historical travel data of selected fare paths received from the data storage by iteratively:
evaluating the plurality of historical characteristics associated with the selected fare paths,
determining a selection prediction for each selected fare path based on the plurality of historical characteristics,
determining one or more features of a plurality of features that have an impact on the selected fare paths, and
updating the machine learning model using the determined one or more features as additional inputs; and
in response to a travel search request from a client device:
determining a set of flight paths fulfilling the travel search request and based on at least one characteristic of a set of characteristics associated with the travel search request;
determining a set of fare paths by combining the set of flight paths with pricing units from a set of pricing units;
applying the machine learning model at the second device on the set of fare paths to determine, based on the travel search request and the at least one characteristic associated with the travel search request, a subset of fare paths for which the selection prediction is higher than a threshold;
querying details of the subset of fare paths, wherein querying details of the subset of fare paths comprises querying at least one carrier providing a fare path in the subset of fare paths and a third-party system providing fare path selection rules;
determining, based on the at least one characteristic associated with the travel search request, valid fare paths; and
returning the details of at least a subset of the valid fare paths to the client device.