| CPC G06Q 30/018 (2013.01) [G06Q 20/023 (2013.01); G06Q 20/3825 (2013.01); H04L 63/0884 (2013.01)] | 12 Claims |
|
1. A method for performing a distributed online transaction in a clearinghouse device, the method comprising:
establishing a secure connection over a communication network between a network interface in a clearinghouse device and (i) a network interface in a user device and (ii) a network interface in a signature authorizing agent device based on a request from the user device, the secure connection used for signature validation of an electronic file representing a document as part of a signature validated transaction associated with a user;
generating, by a processor in the clearinghouse device, a first random set of user identification elements to be provided from a plurality of user identification elements available, the first random set of user identification elements including an identification image and at least one additional identification attribute, the identification image including at least one jurisdiction attribute;
receiving, from the user device over the secure connection, information for the first random set of user identification elements at the network interface in the clearinghouse device;
applying, using the processor in the clearinghouse device, at least one machine learning technique to generate a first identity attribute verification score comparing the information related to the first random set of user identification elements to available data stored in a database, the at least one machine learning technique including generating an identification credential metric, including a jurisdiction metric using a location associated with the requested signature validation transaction and the at least one jurisdiction attribute, as part of the first identity attribute verification score, the jurisdiction metric determined by applying at least one machine learning technique to generate a jurisdiction verification metric using a training model associated with a jurisdiction attribute identified from the identification image;
applying, using the processor in the clearinghouse device, the at least one machine learning technique to generate a second random subset of user identification elements and generating a second identity attribute verification score if the identification credential metric is below a first predetermined threshold in a credentials database;
providing, through the network interface in the clearinghouse device, one of the first identity attribute verification score and the second identity attribute verification score to the signature authorizing agent device;
applying, using the processor in the clearinghouse device, at least one machine learning technique to generate a legal compliance score for the signature validated transaction using at least one of a signature authorization certificate and a notarial seal which is associated with a jurisdiction of the signature authorizing agent and providing the legal compliance score to the signature authorizing agent device;
applying, using the processor in the clearinghouse device, at least one machine learning technique to generate a venue validity score by applying at least one machine learning technique to generate a venue verification metric using a training model associated with the location of the signature authorized transaction and a database of venues and providing the venue validity score to the signature authorizing agent device over the secure connection; and
amending the electronic file representing the document by attaching an electronic signature validation to the electronic file using the processor in the clearinghouse device if (i) the identification credential metric is above the first predetermined threshold, (ii) the legal compliance score is above a second predetermined threshold, and (iii), the venue validity score is above a third predetermined threshold, the electronic signature validation associated with the signature authorizing agent for the signature validated transaction;
the training model for the at least one machine learning technique updated based on an evaluation of at least one of the first random set of user identification elements and the second random set of user identification elements.
|