Method and system for automated product localization through mobile data analysis
An approach for product localization is described that includes detecting one or more movement activities of a user in an indoor environment, and determining a respective position for each of the one or more movement activities using a sensor associated with the user. The approach further includes identifying one or more stoppage locations from the detected one or more movement activities, and receiving a list of one or more purchased items purchased by the user in the indoor environment. Finally, the approach further includes correlating at least one of the more purchased items with at least one of the one or more stoppage locations, and predicting the location of the one or more purchased items based on the correlating.
1. A computer-implemented method of product localization comprising:
detecting, using a real-time location system (RTLS), one or more stoppage locations each user of a plurality of users makes in an indoor environment, wherein detecting the one or more stoppage locations includes using the RTLS that is an electromagnetic technology-based system;
determining, by the RTLS, whether each user stays within a small area relative to the indoor environment for a predetermined amount of time;
receiving a list of one or more purchased items purchased by each user of the plurality of users in the indoor environment, the list being provided by a purchasing system associated with the indoor environment;
correlating at least one of the one or more purchased items with at least one of the one or more stoppage locations, wherein the correlating includes populating a two-dimensional item-stop array for each user of the plurality of users, wherein one dimension of the two-dimensional item-stop array represents purchased items and the other dimension of the two-dimensional item-stop array represents the one or more stoppage locations; and
predicting a location of the one or more purchased items based on the correlating, wherein the predicting further includes accumulating respective two-dimensional item-stop arrays for respective multiple users to form an accumulated two-dimensional item-stop array, wherein the accumulating includes matrix addition of the respective two-dimensional item-stop arrays, and the predicting is based on the location having a highest score in the accumulated two-dimensional item-stop array for the respective one or more purchased items.
2. The computer-implemented method of claim 1 , wherein the detecting the one or more stoppage locations includes using the RTLS that is also an ultrasound-based system.
3. The computer-implemented method of claim 1 , wherein the detecting the one or more stoppage locations includes using the RTLS that is also an infrared-based system.
4. The computer-implemented method of claim 1 , wherein the small area relative to the indoor environment is an area of one meter diameter and the predetermined amount of time is at least three seconds.
5. The computer-implemented method of claim 1 , wherein the detecting the one or more stoppage locations includes using a human activity classifying algorithm in conjunction with the RTLS.
6. The computer-implemented method of claim 1 , wherein the receiving the list of one or more purchased items includes receiving the list via a user application on a smartphone.
7. The computer-implemented method of claim 1 , wherein a non-zero entry of the two-dimensional item-stop array is indicative that the user purchased the respective purchased item and the user stopped at the respective stoppage location.
8. The computer-implemented method of claim 1 , wherein the predicting the location further includes:
using a leaky customer score accumulation algorithm.
9. The computer-implemented method of claim 1 , wherein the predicting the location further includes:
using a softmax-inference based algorithm.
10. A product localization system comprising at least one processor, the at least one processor configured to:
identify, using a real-time location system (RTLS), one or more stoppage locations made by each user of a plurality of users in an indoor environment, wherein identifying the one or more stoppage locations includes using the RTLS that is an electromagnetic technology-based system;
determine, by the RTLS, whether each user stays within a small area relative to the indoor environment for a predetermined amount of time;
receive a list of one or more purchased items purchased by each user of the plurality of users in the indoor environment, the list being provided by a purchasing system associated with the indoor environment;
correlate at least one of the one or more purchased items with at least one of the one or more stoppage locations, wherein the correlating includes populating a two-dimensional item-stop array for each user of the plurality of users, wherein one dimension of the two-dimensional item-stop array represents purchased items and the other dimension of the two-dimensional item-stop array represents the one or more stoppage locations; and
predict a location of the one or more purchased items based on the correlating, wherein the predicting further includes accumulating respective two-dimensional item-stop arrays for respective multiple users to form an accumulated two-dimensional item-stop array, wherein the accumulating includes matrix addition of the respective two-dimensional item-stop arrays, and the predicting is based on the location having a highest score in the accumulated two-dimensional item-stop array for the respective one or more purchased items.
11. The system of claim 10 , wherein the at least one processor is further configured to identify the one or more stoppage locations by using the RTLS that is also an ultrasound-based system.
12. The system of claim 10 , wherein the at least one processor is further configured to identify the one or more stoppage locations by using the RTLS that is also an infrared-based system.
13. The system of claim 10 , wherein the small area relative to the indoor environment is an area of one meter diameter and the predetermined amount of time is at least three seconds.
14. The system of claim 10 , wherein the at least one processor is further configured to identify the one or more stoppage locations by using a human activity classifying algorithm in conjunction with the RTLS.
15. The system of claim 10 , wherein a non-zero entry of the two-dimensional item-stop array is indicative that the user purchased the respective purchased item and the user stopped at the respective stoppage location.
16. The system of claim 10 , wherein the at least one processor is further configured to predict the location by:
using a leaky customer score accumulation algorithm.
17. The system of claim 10 , wherein the at least one processor is further configured to predict the location by:
using a dynamic leaky customer score accumulation algorithm.
18. The system of claim 10 , wherein the at least one processor is further configured to predict the location by:
using a softmax-inference based algorithm.