US 12,217,210 B2
System and method for data selection and extraction based on historical user behavior
Michael James Mangino, Cincinnati, OH (US); Michael Davis Niessner, Jacksonville, FL (US); and Garrett Stephen Wolfe, Carrollton, TX (US)
Assigned to TBK BANK, SSB, Dallas, TX (US)
Filed by TBK Bank, SSB, Dallas, TX (US)
Filed on Nov. 27, 2023, as Appl. No. 18/519,799.
Application 18/519,799 is a continuation of application No. 18/067,817, filed on Dec. 19, 2022, granted, now 11,829,934.
Prior Publication US 2024/0202651 A1, Jun. 20, 2024
This patent is subject to a terminal disclaimer.
Int. Cl. G06Q 10/0834 (2023.01); G06Q 10/083 (2024.01)
CPC G06Q 10/0834 (2013.01) [G06Q 10/0838 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A system for data mapping and transformation, comprising:
a memory configured to store a trusted dataset associated with a first broker and a first carrier, wherein:
the trusted dataset associated with the first broker comprises at least one of a first name, a first identifying number, a first address, or a first phone number with respect to the first broker; and
the trusted dataset associated with the first carrier comprises at least one of a second name, a second identifying number, a second address, or a second phone number with respect to the first carrier; and
a processor operably coupled with the memory, and configured to:
receive load data from a second broker, wherein:
the load data indicates that a load is assigned to a second carrier to be transported;
the load is associated with a shipper; and
the second carrier is associated with a factor entity that facilitates keeping records of documents for the second carrier;
extract a first set of data elements from the load data, wherein the first set of data elements comprises indications of at least one of the second broker, the factor entity, the shipper, or the second carrier, wherein extracting the first set of data elements is in response to implementing a machine learning algorithm that is trained to map a given data element in the load data to a corresponding data field in the trusted dataset based at least in part upon a historical data mapping;
determine that a portion of the first set of data elements is anomalous indicating that the portion of the first set of data elements comprises incomplete or incorrect information about the at least one of the second broker, the factor entity, the shipper, or the second carrier, wherein determining that the portion of the first set of data elements is anomalous comprises determining that the portion of the first set of data element does not correspond to a respective data element in the trusted dataset as determined by the machine learning algorithm;
transform the portion of the first set of data elements that is determined to be anomalous to correct information retrieved from the trusted dataset;
generate a mapped dataset by mapping at least one of the first set of data elements to a respective data element from the trusted dataset; and
determine, based at least in part upon mapping of the at least one of the first set of data elements to the respective data element from the trusted dataset, an identity of at least one of the second broker, the factor entity, the shipper, the second carrier.