Back to results
US 12,749,059 B2Granted
System and process for electronic payments
AssigneeBank of Montreal
InventorsBrian Chan, Kashif Arshad, Peter Hing-Cheong Poon, Vikram Pal
Life of this patent
Estimating the term…
Abstract
A platform and process for electronic payment processing using electronic communications from different communication channels or bands. The system and process can generate alerts using fraud detection and verify payment requests using historical data and pattern recognition. The system and process can categorize images and extract payment data.
Claims20
A method for improving model accuracy using curated training datasets, the method comprising:
- executing, by one or more processors, a predictive computer model using a first image to determine if the first image has a status corresponding to a request and predicting a confidence score associated with the status;
- when the status of the first image corresponds to the request, extracting, by the one or more processors, data values from the first image;
- generating, by the one or more processors, a training dataset comprising only the first image and the data values from the first image;
- revising, by the one or more processors, the training dataset by:
- altering, by the one or more processors, at least one image factor of the first image by adjusting at least one of a contrast, brightness, pixel orientation, greyscale, black and white, blur, or bit color attribute of the first image;
- generating, by the one or more processors, a set of second images based on at least one altered first image
- adding, by the one or more processors, the set of second images to the training dataset;
- training, by the one or more processors, the predictive computer model using the revised training dataset;
- generating, by the one or more processors, a payload using the extracted data values based on a vendor format, wherein the extracted data values indicate a vendor identifier linked to the vendor format;
- upon transmitting a confirmation request to an electronic address associated with the first image, receiving, by the one or more processors, an approval notification in response to the confirmation request, wherein the approval notification is further used for training the predictive computer model;
- and
- in response to receiving the approval notification, transmitting, by the one or more processors to a server, an instruction to execute the payload responsive to the request.
9 more elements
A system for improving model accuracy using curated training datasets, the system comprising:
- a server having one or more processors configured to:
- execute a predictive computer model using a first image to determine if the first image has a status corresponding to a request and predicting a confidence score associated with the status;
- when the status of the first image corresponds to the request, extract data values from the first image;
- generate a training dataset comprising only the first image and the data values from the first image;
- revise the training dataset by:
- altering at least one image factor of the first image by adjusting at least one of a contrast, brightness, pixel orientation, greyscale, blur, or bit color attribute of the first image;
- generating a set of second images based on at least one altered first image and
- adding the set of second images to the training dataset;
- train the predictive computer model using the revised training dataset;
- generate a payload using the extracted data values based on a vendor format, wherein the extracted data values indicate a vendor identifier linked to the vendor format;
- upon transmitting a confirmation request to an electronic address associated with the first image, receive an approval notification in response to the confirmation request, wherein the approval notification is further used for training the predictive computer model;
- and
- in response to receiving the approval notification, transmit an instruction to execute the payload responsive to the request.
10 more elements
A computer system for improving model accuracy using curated training datasets, the computer system comprising:
- a computer-readable medium configured to store non-transitory instructions, that when executed, cause one or more processors to:
- execute a predictive computer model using a first image to determine if the first image has a status corresponding to a request and predicting a confidence score associated with the status;
- when the status of the first image corresponds to the request, extract data values from the first image;
- generate a training dataset comprising only the first image and the data values from the first image;
- revise the training dataset by:
- altering at least one image factor of the first image by adjusting at least one of a contrast, brightness, pixel orientation, greyscale, blur, or bit color attribute of the first image;
- generating a set of second images based on at least one altered first image;
- and
- adding the set of second images to the training dataset;
- train the predictive computer model using the revised training dataset;
- generate a payload using the extracted data values based on a vendor format, wherein the extracted data values indicate a vendor identifier linked to the vendor format;
- upon transmitting a confirmation request to an electronic address associated with the first image, receive an approval notification in response to the confirmation request, wherein the approval notification is further used for training the predictive computer model;
- and
- in response to receiving the approval notification, transmit an instruction to execute the payload responsive to the request.
11 more elements
Dependencies are inferred from the claim text.