Enabling return path data on a non-hybrid set top box for a television

An Artificial Intelligence (AI) based intelligent return path data (iRPD) system comprising:

• at least one processor;
• a non-transitory processor readable medium storing machine-readable instructions that cause the at least one processor to:
• receive via a communication network, remote control data including keypress data manipulating a television via a non-hybrid set top box, wherein the keypress data is indicative of one or more keys pressed by a viewer on a remote control device for operating the television and the remote control data includes date-time stamps for the keys pressed and location information of the television;
• generate a first set of clusters by clustering the keypress data received in the remote control data using a first clustering methodology based on the date-time stamps;
• identify the keys pressed by the viewer;
• determine operations executed by the viewer on the television based at least on the first set of clusters, wherein the operations include programming operations and non programming control operations;
• further cluster the first set of clusters using a trained clustering behavior model that implements a second clustering methodology different from the first clustering methodology, the trained clustering behavior model analyzes viewer behavioral patterns identified from the operations, wherein the further clustering is based at least on a plurality of factors;
• identify the viewer at a location of the television based on an output from the trained clustering behavior model;
• and
• identify viewer behavioral patterns of the viewer based on patterns of clusters formed by the further clustering of the first set of clusters, wherein the identification of the viewer behavioral patterns includes:
• determining whether the viewer arrived at a selected television channel via channel scrubbing or if the viewer directly accessed the selected television channel based on the patterns of the clusters generated by the trained clustering behavior model,
• identifying numeric clusters from the further clusters using a classifier, wherein the numeric clusters represent the programming operations, and
• determining affinity of the viewer to television channels accessed by the viewer based at least on a distance between the numeric clusters generated by the trained clustering behavior model, wherein greater distance between the numeric clusters generated by the trained clustering behavior model indicates the viewer accessing the television channels using numeric keys of the remote control device as determined from the keypress data and the higher affinity of the viewer to the television channels represented by the numeric clusters separated by the greater distance.