Associating digital activities with positions in physical retail stores

1. A non-transitory computer-readable medium including instructions that when executed by a processor cause the processor to perform a method for associating online digital activities with conditions in a retail store, the method comprising:

detecting a plurality of online digital activities including subject matter related to a retail store, wherein the plurality of online digital activities originate from customers in the retail store;

receiving position information indicative of positions in the retail store from which each of the detected plurality of online digital activities originated;

constructing a mathematical graph based on the received position information, where each online digital activity of the detected plurality of online digital activities is associated with a node of the mathematical graph, a particular edge of the mathematical graph connects a first node associated with a first online digital activity with a second node associated with a second online digital activity, and the particular edge is associated with a weight based on a similarity between the first online digital activity and the second online digital activity and on a distance between a position associated with the first online digital activity and a position associated with the second online digital activity;

analyzing the mathematical graph to cluster the detected plurality of online digital activities to provide one or more clusters of detected online digital activities, wherein each of the one or more clusters is associated with a different region of the retail store;

using an image acquisition device to capture one or more images representative of a particular region of the retail store associated with a particular cluster of detected online digital activities;

analyzing the one or more captured images using a machine learning model to detect a condition of the retail store associated with the particular cluster of detected online digital activities, the machine learning model having been trained using a set of training images with corresponding labels indicating conditions represented in the training images; and

automatically initiating at least one remedial action to address the detected condition.

2. The non-transitory computer-readable medium of claim 1 , wherein the position information further indicates at least a portion of a path traversed by at least one customer in the retail store.

3. The non-transitory computer-readable medium of claim 1 , wherein the position information is generated by one or more mobile devices.

4. The non-transitory computer-readable medium of claim 3 , wherein the position information further indicates an orientation of the one or more mobile devices during generation of one or more of the plurality of online digital activities.

5. The non-transitory computer-readable medium of claim 1 , wherein the particular region of the retail store includes a display area for a particular product in the retail store.

6. The non-transitory computer-readable medium of claim 5 , wherein the online digital activities are online digital activities for checking a price of the particular product.

7. The non-transitory computer-readable medium of claim 5 , wherein the detected condition is associated with the particular product.

8. The non-transitory computer-readable medium of claim 7 , wherein the detected condition is associated with a price of the particular product.

9. The non-transitory computer-readable medium of claim 7 , wherein the detected condition is associated with an availability of the particular product.

10. The non-transitory computer-readable medium of claim 9 , wherein the online digital activities are online digital activities for ordering the particular product.

11. The non-transitory computer-readable medium of claim 7 , wherein the at least one remedial action is selected based on sales data associated with the particular product and the retail store.

12. The non-transitory computer-readable medium of claim 1 , wherein the detected condition is associated with a queue of customers in the retail store.

13. The non-transitory computer-readable medium of claim 1 , wherein the detected condition is associated with an obstruction in the retail store.

14. The non-transitory computer-readable medium of claim 1 , wherein the detected condition is associated with a promotion for a product in the retail store.

15. The non-transitory computer-readable medium of claim 1 , wherein the at least one remedial action includes a prompt to increase a number of available checkout locations.

16. The non-transitory computer-readable medium of claim 1 , wherein the at least one remedial action includes a prompt to clean a spill in the retail store.

17. The non-transitory computer-readable medium of claim 1 , wherein the at least one remedial action includes a prompt to restock a product display shelf in the retail store.

18. The non-transitory computer-readable medium of claim 1 , wherein the at least one remedial action includes a prompt to adjust a price associated with at least one product in the retail store.

19. A system for associating online digital activities with positions in retail stores, the system comprising:

at least one processing unit; and

a memory storing instructions executable by the at least one processing unit to cause the at least one processing unit configured to:

detect a plurality of online digital activities including subject matter related to a retail store, wherein the plurality of online digital activities originate from customers in the retail store;

receive position information indicative of positions in the retail store from which each of the detected plurality of online digital activities originated;

construct a mathematical graph based on the received position information, where each online digital activity of the detected plurality of online digital activities is associated with a node of the mathematical graph, a particular edge of the mathematical graph connects a first node associated with a first online digital activity with a second node associated with a second online digital activity, and the particular edge is associated with a weight based on a similarity between the first online digital activity and the second online digital activity and on a distance between a position associated with the first online digital activity and a position associated with the second online digital activity;

analyze the mathematical graph to cluster the detected plurality of online digital activities to provide one or more clusters of detected online digital activities, wherein each of the one or more clusters is associated with a different region of the retail store;

use an image acquisition device to capture one or more images representative of a particular region of the retail store associated with a particular cluster of detected online digital activities;

analyze the one or more captured images using a machine learning model to detect a condition of the retail store associated with the particular cluster of detected online digital activities, the machine learning model having been trained using a set of training images with corresponding labels indicating conditions represented in the training images; and

automatically initiate at least one remedial action to address the detected condition.

20. A method for associating digital activities with positions in retail stores, the method comprising:

detecting a plurality of online digital activities including subject matter related to a retail store, wherein the plurality of online digital activities originate from customers in the retail store;

receiving position information indicative of positions in the retail store from which each of the detected plurality of online digital activities originated;

constructing a mathematical graph based on the received position information, where each online digital activity of the detected plurality of online digital activities is associated with a node of the mathematical graph, a particular edge of the mathematical graph connects a first node associated with a first online digital activity with a second node associated with a second online digital activity, and the particular edge is associated with a weight based on a similarity between the first online digital activity and the second online digital activity and on a distance between a position associated with the first online digital activity and a position associated with the second online digital activity;

analyzing the mathematical graph to cluster the detected plurality of online digital activities to provide one or more clusters of detected online digital activities, wherein each of the one or more clusters is associated with a different region of the retail store;

using an image acquisition device to capture one or more images representative of a particular region of the retail store associated with a particular cluster of detected online digital activities;

analyzing the one or more captured images using a machine learning model to detect a condition of the retail store associated with the particular cluster of detected online digital activities, the machine learning model having been trained using a set of training images with corresponding labels indicating conditions represented in the training images; and

automatically initiating at least one remedial action to address the detected condition.