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

• US 10,474,705 B2
• Filed: 10/16/2018
• Issued: 11/12/2019
• Est. Priority Date: 08/15/2014
• Status: Active Grant

• Current Assignee: Ask Sydney, LLC
• Original Assignee: Paul Lawrence Epstein, Sydney Nicole Epstein
• Inventors: Epstein, Sydney Nicole, Epstein, Paul Lawrence
• Primary Examiner(s): Pond, Robert M
• Application Number:
  US16/162,024
• Publication Number:
  US 20190050477A1
• Time in Patent Office: 392 Days
• Field of Search: 705 26, 705 27
• CPC Class Codes: G06F 16/334 Query execution G06F16/335 takes precedence
• CPC Class Codes: G06F 16/51 Indexing; Data structures therefor; Storage structures
• CPC Class Codes: G06F 16/58 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/5866 using information manually generated, e.g. tags, keywords, comments, manually generated location and time information
• CPC Class Codes: G06F 3/04842 Selection of displayed objects or displayed text elements G06F3/0482 takes precedence
• CPC Class Codes: G06Q 30/0631 Item recommendations
• CPC Class Codes: H04L 67/306 User profiles
• CPC Class Codes: H04W 4/021 Services related to particular areas, e.g. point of interest [POI] services, venue services or geofences

Abstract

System and computer-implemented method of analyzing tags associated with a sequence of images presented to a user to present a current interest of the user is disclosed. An image from among a plurality of images is presented on an electronic display. The image is associated with a set of tags. An input is received indicating a user'"'"'s preference for the image. A plurality of tags is processed based on the preference and the set of tags to determine a next set of tags from the plurality of tags. A next image is determined from the plurality of images based on the next set of tags. The next image represents a physical object, different from a physical object represented by the previous image. A sequence of images is generated by repeating the above process with the next image in place of the previous image for present a user'"'"'s current interest.

Claims

1. A computer-implemented image search method filtered by multiple human-machine inputs on images presented to a user of the image search method, the method comprising the steps of:

determining a plurality of different digital images to present on the video display device to generate a sequence of digital images, each of the digital images being associated with a plurality of tags indicating one or more attributes of an image featured in the corresponding digital image;

receiving, via a user input device, an unfavorable indication of a disinclination for the image features in the one of the digital images, the unfavorable indication being one of at least two input options, the at least two input options including, among the unfavorable indication, a favorable indication of a preference for the image featured in one of the digital images;

analyzing at least some of the tags to determine a next set of tags associated with a subsequent digital image;

adjusting weights of at least some of the tags based on an association relative to tags within the same category of tags to which the subsequent digital image belongs;

transitioning the one of the digital images with the subsequent digital image on the video display device to replace the one of the digital images with the subsequent digital image on the video display device; and

receiving via the user input device a further input corresponding to one of the at least two input options including a favorable indication for the subsequent digital image and an unfavorable indication for the subsequent digital image.

2. The method of claim 1, wherein the corresponding image featured on respective ones of the at least some of the digital images are consumer goods, food dishes, clothing, automobiles, physical locations, vacation spots, museums, or sports venues.

3. The method of claim 1, further comprising:

displaying on the video display device graphical user interface elements corresponding to a like virtual button and a dislike virtual button, wherein the like virtual button when selected indicates the favorable indication and wherein the dislike virtual button when selected indicates the unfavorable indication.

4. The method of claim 1, in response to receiving the favorable indication for the image featured in the one of the digital images, increasing weights of at least some of the tags associated with the one of the digital images based on an association relative to tags within the same category of tags to which the subsequent digital image belongs.

5. The method of claim 4, wherein a subsequent digital image to be presented in response to the favorable indication is required to have at least some of the tags associated with the one of the digital images.

6. The method of claim 5, wherein the subsequent digital image in response to the favorable indication is required to have all of the tags associated with the one of the digital images.

7. The method of claim 1, further comprising processing at least some of the tags differently in response to the favorable indication versus the unfavorable indication.

8. The method of claim 7, wherein the processing is based on a number of times a user indicates a preference for the one of the digital images.

9. A computer-implemented image search method, comprising the steps of:

determining a plurality of digital images to present on a video display device of a computer, each of the digital images being associated with a plurality of tags;

repeating, as long as human-machine inputs are received from a user input device indicating a like or dislike preference of a user for a selected digital image of the digital images, the steps of;

processing the tags associated with the selected digital image to determine a next set of tags associated with a subsequent digital image to be presented on the video display;

in response to the dislike preference, the processing including adjusting weights of at least some of those of the processed tags in the same category as the tags associated with the selected digital image so that the subsequent digital image has a different set of associated tags from those associated with the selected digital image, and replacing the selected digital image with the subsequent digital image on the video display device;

in response to the like preference, the processing including adjusting weights of at least some of the processed tags so that the subsequent digital image has at least some of the same tags as those associated with the selected digital image, and presenting the subsequent digital image on the video display device;

wherein the repeating causes the image search method to iteratively resolve toward at least one digital image presented on the video display device, which satisfies a subjective need of the user via the multiple human-machine inputs, and the repeating includes at least one like preference and at least one dislike preference.

10. The method of claim 9, wherein in response to the like preference, the subsequent digital image is required to have all of the tags associated with the selected digital image.

11. A computer-implemented image search method, comprising the steps of:

determining a plurality of digital images to present on a video display device of a computer, each of the digital images being associated with a plurality of tags;

repeating, as long as human-machine inputs are received from a user input device indicating one of at least two input options for a selected image of the digital images, the at least two input options including a favorable indication of a preference for the image featured in one of the digital images and an unfavorable indication of a disinclination for the image featured in the one of the digital images, the steps of;

processing the tags associated with the selected digital image to determine a next set of tags associated with a subsequent digital image to be presented on the video display;

in response to the unfavorable indication, the processing including adjusting weights of at least some of those of the processed tags in the same category as the tags associated with the selected digital image so that the subsequent digital image has a different set of associated tags from those associated with the selected digital image, and replacing the selected digital image with the subsequent digital image on the video display device;

in response to the favorable indication, the processing including adjusting weights of at least some of the processed tags so that the subsequent digital image has at least some of the same tags as those associated with the selected digital image, and presenting the subsequent digital image on the video display device;

wherein the repeating causes the image search method to iterate toward at least one digital image presented on the video display device, which satisfies a subjective need of the user via the multiple human-machine inputs, and the repeating includes at least one favorable indication and at least one unfavorable indication.