US 11,893,627 B2
Adaptive risk-based verification and authentication platform
Farhang Kassaei, Saratoga, CA (US); Amanda A. Earhart, Santa Cruz, CA (US); Snezana Sahter, San Jose, CA (US); Srinivasu Gottipati, Fremont, CA (US); Lars Wright, Campbell, CA (US); Craig Rowley, San Jose, CA (US); Lakshman Shyam Sundar Maddali, Sunnyvale, CA (US); and Nainesh Nayudu, Santa Clara, CA (US)
Assigned to EBAY INC., San Jose, CA (US)
Filed by eBay Inc., San Jose, CA (US)
Filed on Sep. 5, 2019, as Appl. No. 16/561,866.
Application 16/561,866 is a continuation of application No. 15/818,126, filed on Nov. 20, 2017, granted, now 10,489,853.
Application 15/818,126 is a continuation of application No. 12/483,506, filed on Jun. 12, 2009, granted, now 9,830,643.
Claims priority of provisional application 61/164,847, filed on Mar. 30, 2009.
Prior Publication US 2019/0392514 A1, Dec. 26, 2019
Int. Cl. G06Q 40/00 (2023.01); G06Q 10/00 (2023.01); G06Q 30/0601 (2023.01)
CPC G06Q 40/00 (2013.01) [G06Q 10/00 (2013.01); G06Q 30/0601 (2013.01)] 19 Claims
OG exemplary drawing
 
1. A method to perform a risk-based assessment of a user, the method comprising:
receiving, by a server, a request from a user to perform an action to sell an item on an e-commerce site, the request comprising search data, the item being identified in response to:
extracting one or more words from the search data; and
storing the extracted words in a database in association with the item;
retrieving, by the server, a first risk assessment factor associated with the user;
retrieving, by the server, a second risk assessment factor associated with the action on the item;
training a neural network to implement a risk assessment model by performing training operations comprising:
processing training data by the neural network to estimate a risk assessment level corresponding to the risk assessment model that is used to determine that the action is not permitted;
incrementally changing internal weights of the neural network used to generate the estimate of the risk assessment level; and
updating a final outcome produced by the neural network in response to incrementally changing the internal weights of the neural network;
automatically determining, by the server, the action is not permitted based on a new risk assessment level generated by the risk assessment model, the risk assessment model comprising a plurality of factors including at least the first risk assessment factor and the second risk assessment factor;
identifying a risk mitigation process from a plurality of risk mitigation processes based on the new risk assessment level generated by the risk assessment model, a first of the risk mitigation processes being associated with a first type of risk determined by the risk assessment model, a second of the risk mitigation processes being associated with a second type of risk determined by the risk assessment model, the first of the risk mitigation processes including analysis of seller feedback information, and the second of the risk mitigation processes including obtaining identity verification information from the user;
communicating the identified risk mitigation process to the user;
in response to communicating the identified risk mitigation process to the user, receiving, by the server, risk mitigation data from the user corresponding to the identified risk mitigation process; and
in response to receiving the risk mitigation data from the user, automatically modifying, by the server, the risk new assessment level generated by the risk assessment model.