CPC G06Q 40/123 (2013.12) [G06N 5/04 (2013.01)] | 20 Claims |
1. A system for probabilistically predicting a tax refund range, the system located in a service provider computing environment and comprising:
one or more processors; and
at least one memory coupled to the one or more processors and storing instructions that, when executed by the one or more processors, causes the system to perform operations including:
receiving, via a user interface module in connection with a user device located in a user computing environment, user data associated with a system user;
transforming the user data into a prediction of the system user's tax refund range based on:
receiving prior tax return data over a communication network from a system database, the prior tax return data indicating characteristics of prior system users;
identifying ones of the prior system users sharing at least one characteristic with the system user based on the prior tax return data received from the system database;
generating, from the user data, tax return data for the system user;
generating at least one statistical inference about characteristics of the system user based on the identified ones of the prior system users sharing at least one characteristic with the system user and the prior tax return data received from the system database;
determining, for each respective inference of the at least one statistical inference, a probability that the respective inference is correct based on a probabilistic analysis of the tax return data; and
in response to at least one of the determined probabilities being greater than a value, generating the prediction of the system user's tax refund range;
providing, via the user interface module, the system user with the generated prediction;
continuously refining the generated prediction based on:
receiving, via the user interface module, additional user data associated with the system user;
generating additional inferences about characteristics of the system user based on additional prior tax return data;
determining additional probabilities that the additional inferences are correct based on additional probabilistic analyses; and
refining the generated prediction based on the additional probabilities; and
providing, via the user interface module, the system user with the refined prediction.
|