×

Iterative image search algorithm informed by continuous human-machine input feedback

  • US 10,268,702 B2
  • Filed: 02/26/2016
  • Issued: 04/23/2019
  • Est. Priority Date: 08/15/2014
  • Status: Active Grant
First Claim
Patent Images

1. A computer-implemented method comprising:

  • determining, by one or more computer devices, a plurality of tags specific to a user, from among a pool of tags, based on each tag of the plurality of tags specific to the user being associated with a profile of the user;

    transmitting, from the one or more computer devices, one electronic image, from among a plurality of electronic images stored on the one or more computer devices, to an electronic device, the one image being associated with one set of tags from the plurality of tags specific to the user, each tag of the one set of tags describing or characterizing attributes of the one image;

    receiving, from the electronic device, an input from the user indicating a preference for the one image;

    processing, by the one or more computer devices, the plurality of tags specific to the user based on the preference and the one set of tags to determine a next set of tags from the plurality of tags and, in response to the preference for the one image being positive, the processing of the plurality of tags further comprising;

    determining tags, from the plurality of tags, having a threshold association with one or more tags of the one set of tags;

    determining a weighting of the tags having the threshold association based, at least in part, on a number of times each tag of the plurality of tags specific to the user is associated with a positive and/or a negative preference by the user; and

    determining the next set of tags based on the next set of tags including at least one tag of the tags having the threshold association,wherein the at least one tag is a highest weighted tag of the tags having the threshold association;

    determining, by the one or more computer devices, a next image from the plurality of images associated with the next set of tags, the next image being different from the one image, and the next set of tags describing or characterizing attributes of the next image; and

    generating a sequence of images by repeating the presenting, the receiving, the processing, and the determining with the next image in place of the one image during a session of presenting a current interest of the user,wherein the input from the user indicating the preference for the one image is positive at least once during the session of presenting the current interest of the user.

View all claims
  • 6 Assignments
Timeline View
Assignment View
    ×
    ×