Back to results
US 2020/0349573 A1Published application
Detecting fraud using machine-learning
AssigneePaypal, Inc.
InventorYanfei Dong
Life of this patent
Estimating the term…
Abstract
A fraud detection model is used by a computer system to evaluate whether to grant a request to access a secure electronic resource. The fraud detection model is generated by the computer using a plurality of previously received requests. Each request is associated with a sender account of the computer system and a recipient account to which the computer system previously sent a message containing a link to the secure electronic resource.
Claims20
A method comprising:
- sending, from a computer system to a first recipient account, a first message containing a first link to a first electronic resource of a plurality of electronic resources, wherein the first electronic resource is associated with a first sender account of the computer system;
- receiving, at a computer system, a request to access the first electronic resource via the first link;
- and
- before granting the request to access the first electronic resource, evaluating the request to access the first electronic resource using a fraud detection model generated by:
- receiving a plurality of previous requests, wherein each of the plurality of previous requests (a) is a request to access one of the plurality of electronic resources and (b) is associated with a respective sender account of the computer system and a respective recipient account,
- for each of plurality of previous requests, sequentially generating, for the fraud detection model, embedding values corresponding to both the sender account and the recipient account associated with that previous request, wherein each embedding value represents a particular sender account or a particular recipient account using a reduced number of dimensions relative to the number of dimensions in which the corresponding previous request was captured, and
- training the fraud detection model using indications that ones of the plurality of requests were fraudulent.
4 more elements
A method for generating a fraud detection model, the method comprising:
- receiving, at a computer system, a plurality of requests to access respective electronic resources, wherein each of the plurality of requests is associated with a respective sender account of the computer system and a respective recipient account;
- initializing, with the computer system, embedding values for the respective sender accounts and respective recipient accounts within the fraud detection model;
- and
- incorporating, with the computer system, each of the plurality of requests into the fraud detection model by:
- generating an updated embedding value for the sender account associated with the request based on (a) the request and (b) a previous embedding value for the sender account associated with the request;
- and
- generating an updated embedding value for the recipient account associated with the request based on (a) the request, (b) the updated embedding value of the sender account associated with the request, and (c) a previous embedding value of the recipient account associated with the request.
4 more elements
A method comprising:
- modeling, with a computer system in a fraud detection model, a plurality of sender accounts, a plurality of recipient accounts, and a plurality of requests to access a plurality of secure electronic resources, wherein:
- the modeling includes calculating an embedding value for each of the plurality of sender accounts and an embedding value for each of the plurality of recipient accounts, and
- each of the plurality of requests is associated with a given sender account and a given recipient account;
- receiving, at the computer system, a first additional request to access a first secure electronic resource associated with a first sender account and a first recipient account;
- and
- adding, with the computer system, the first additional request to the fraud detection model including calculating an updated embedding value for the first sender account within the model and calculating an updated embedding value of the first recipient account within the model.
3 more elements
Dependencies are inferred from the claim text.