US 11,748,435 B2
Content-free system and method to recommend news and articles
Alka Dagar, New Delhi (IN); and Ratheen Chaturvedi, Gurgaon (IN)
Assigned to S&P Global Inc., New York, NY (US)
Filed by S&P Global Inc., New York, NY (US)
Filed on Apr. 23, 2021, as Appl. No. 17/302,090.
Prior Publication US 2022/0342945 A1, Oct. 27, 2022
Int. Cl. G06F 16/9535 (2019.01); G06F 16/957 (2019.01); G06N 20/00 (2019.01); G06F 16/951 (2019.01)
CPC G06F 16/957 (2019.01) [G06F 16/951 (2019.01); G06F 16/9535 (2019.01); G06N 20/00 (2019.01)] 27 Claims
OG exemplary drawing
 
1. A computer-implemented method for recommending information sources, the method comprising:
using a number of processors to perform the steps of:
collecting raw user log data of a browser user;
transforming the raw user log data to remove data that is specific to the user including logins, sign-ups, and personal dashboards, wherein the transformed raw user log data comprises only general information regarding pages visited, wherein the general information comprises at least one of structure or indices of the pages visited, and wherein the general information excludes content of the pages visited and does not use explicit content information viewed by the user;
modifying the transformed raw user log data by appending a number of key page mappings and user engagement metrics to the transformed raw user log data, wherein the number of key page mappings comprises a generic identifier referring to a type of each of the pages visited;
determining, by a machine learning model, a user interaction function according to only the modified transformed raw user log data;
generating, by the machine learning model, a number of web page recommendations, wherein each web page recommendation has a respective probability of engagement based on the user interaction function;
receiving user input of a specified type of web page visited;
selecting a subset of the web page recommendations that were generated from the modified transformed user log data having the specified type of web page visited appended; and
displaying the subset of web page recommendations to the user on an interface.