US 12,112,653 B2
Systems and methods for providing tailored educational materials
Tushar Singh, Waterloo (CA)
Assigned to MINUTE SCHOOL INC., Waterloo (CA)
Filed by MINUTE SCHOOL INC., Waterloo (CA)
Filed on Apr. 15, 2021, as Appl. No. 17/231,103.
Application 17/231,103 is a division of application No. 15/787,193, filed on Oct. 18, 2017, granted, now 11,056,015.
Prior Publication US 2022/0114903 A1, Apr. 14, 2022
This patent is subject to a terminal disclaimer.
Int. Cl. G09B 7/08 (2006.01); G06F 16/951 (2019.01); G09B 5/06 (2006.01); G09B 5/12 (2006.01); G09B 7/07 (2006.01); G09B 19/18 (2006.01)
CPC G09B 7/08 (2013.01) [G06F 16/951 (2019.01); G09B 5/06 (2013.01); G09B 5/065 (2013.01); G09B 5/125 (2013.01); G09B 7/07 (2013.01); G09B 19/18 (2013.01)] 6 Claims
OG exemplary drawing
 
1. A computer-implemented method of presenting education content to at least one user on a computing device, the method comprising:
receiving, by a processor, data from a data store relating to the at least one user's performance in a course having a proficiency average, said course comprising a plurality of questions, said data including, for each of said plurality of questions, an amount of time elapsed between presentation of a respective question and an answer to said respective question from said at least one user, a number of times said respective question has been asked, and a number of times said respective question has been answered correctly;
training, by the processor, a recurrent neural network by:
creating a first training set comprising the data relating to the at least one user's performance in a first of said plurality of questions and said proficiency average;
creating a second training set comprising the data relating to the at least one user's performance in the first of said plurality of questions and at least a second of said plurality of questions and said proficiency average;
creating at least a third training set comprising the data relating to the at least one user's performance in each of said plurality of questions in said course and said proficiency average; and
training, at a single time instance, said recurrent neutral network using each of said first training set, said second training set, and said third training set;
selecting, by the processor, a plurality of content items including at least one question, wherein selecting the at least one question comprises:
for the at least one question, determining, by the processor, a probability that the at least one question will be answered correctly using the recurrent neural network; and
selecting said at least one question based on said at least one question having a probability of being answered correctly by said at least one user that is less likely than a threshold probability; and
presenting said plurality of selected content items including said at least one question to said at least one user on a client device associated with said at least one user.