System and method for automated retail product accounting
View Patent ↗A system and method including collecting image data of a purchase receipt; converting image data to machine readable data; generating a receipt model including at least one line item from the machine readable text data; selecting a product profile to associate with a line item through one of a tiered set of receipt heuristics, that includes: through a direct identifier heuristic, searching for a product identifier from a line item that corresponds to a product profile, and through an approximate match heuristic, calculating a score for the line item compared to at least one product profile of a master product list; and accounting for the selected product profile of the line item.
1. A method for digitally tracking each of a set of physical products purchased by a user in a shopping trip in order to personalize digital purchase feedback served to the user, the method comprising:
a. capturing receipt image data of a physical purchase receipt from a camera of a user device, associated with a first user account, wherein the physical purchase receipt comprises a set of physically printed line items corresponding to the set of physical products purchased in the shopping trip;
b. converting the captured receipt image data to machine readable data;
c. generating a receipt model including a set of line items from the machine readable text data, the set of line items corresponding to the set of physically printed line items;
d. selecting a product profile to associate with a line item through one of a tiered set of receipt heuristics, comprising, through a direct identifier heuristic, searching for a product identifier from a line item that corresponds to a product profile, and through an approximate match heuristic, calculating a score for the line item compared to at least one product profile of a master product list, and selecting the product profile based on the score;
e. at a remote server, accounting for a set of selected product profiles associated with the set of physical products purchased in the shopping trip, comprising accounting for the selected product profile of the line item;
f. at the remote server, updating a purchase history of the first user account based on the set of selected product profiles of the line item;
g. at the remote server, generating purchase feedback tailored to the first user account based on the updated purchase history;
h. at the remote server, transmitting the purchase feedback to the user device;
i. repeating a) to h) for a second physical purchase receipt associated with a second user account; and
j. at the remote server, sending data for the selected product profile, generated based on the first and second physical purchase receipts, to a third user device associated with a third user account in response to receipt of a request for the data.
2. The method of claim 1 , wherein calculating a score for the line item comprises factoring in at least one context parameter of the receipt model.
3. The method of claim 1 , wherein generating a receipt model comprises identifying a merchant identifier from the machine readable data; and further comprising limiting product profiles compared in the master product list to products associated with the identified merchant identifier; and executing a receipt heuristic based on the merchant identifier.
4. The method of claim 3 , wherein generating a receipt model comprises identifying a category identifier of at least the one line item from the machine readable data; and further comprising limiting product profiles compared in the master product list to product profiles associated with the category identifier.
5. The method of claim 1 , wherein generating a receipt model further comprises identifying in the at least one line item a text description and price parameter.
6. The method of claim 5 , wherein calculating a score for the line item compared to at least one product profile of the master product list comprises scoring a comparison based on character matches of the text description of the line item and a text description of a product profile.
7. The method of claim 6 , wherein character matches are weighted based on a character weight rule.
8. The method of claim 6 , wherein scoring a comparison based on character matches comprises scoring based in part on character sequence order for the text description.
9. The method of claim 6 , wherein scoring a comparison based on character matches comprises weighting a word match greater than a subsequent word match and weighting a character match of a first word greater than a subsequent character match of the first word.
10. The method of claim 6 , wherein calculating a score for the line item compared to at least one product profile of the master product list comprises accounting for character set rules.
11. The method of claim 5 , wherein calculating a score for the line item compared to at least one product profile of the master product list comprises scoring a comparison in part on the comparison of price parameter of the line item to a price parameter of a product profile.
12. The method of claim 5 , wherein generating a receipt model further comprises identifying in the at least one line item a product metric parameter; and wherein calculating a score for the line item compared to at least one product profile of the master product list comprises scoring a comparison in part on the product metric of the line item matching a product metric of a product profile.
13. The method of claim 1 , wherein generating a receipt model further comprises identifying in the at least one line item a text description, product metric, and price parameter; wherein calculating a score for the line item compared to at least one product profile of the master product list comprises scoring a comparison based on character matches of the text description of the line item and a text description of a product profile; wherein character matches are weighted based on a character weight rule; wherein scoring a comparison based on character matches comprises scoring based in part on character sequence order for the text description.
14. The method of claim 1 , wherein selecting a product profile further comprises, retrieving a crowdsourced resolution of line item when a line item is determined to be unidentifiable.
15. The method of claim 1 , wherein the image data of the physical purchase receipt is collected from a first user, and, in g), image data of a second purchase receipt is received from a second user.
16. The method of claim 1 , wherein accounting for the selected product profile of the line item comprises updating a product profile of the master product list to reflect a purchase of the product.
17. The method of claim 16 , wherein accounting for the selected product profile of the line item further comprises updating a product profile of the master product list to include identified line item parameters of the receipt.
18. The method of claim 16 , wherein the selected product profile is a newly created product profile if no product profile is found in the product master list; and wherein updating a product profile of the master product list further comprises adding the newly created product profile to the master product list.
19. The method of claim 1 , wherein the image data is collected from an application of a first user; and wherein accounting for the selected product profile of the line item comprises transmitting information feedback to the application.
20. The method of claim 19 , wherein the transmitted information feedback is a monetary reward based on a number of image data collected from the first user account.
21. The method of claim 1 , further comprising providing a receipt processing API; wherein capturing image data of a purchase receipt is through the receipt processing API; and wherein accounting for the selected product profile of the line item comprises replying to the receipt processing API with data of at least the one selected product profile.
22. A method improving receipt processing and product matching performance of a computing system through digitally tracking each of a set of physical products purchased by a user in a shopping trip, the method comprising:
a. capturing, from a camera of a mobile device, receipt image data of a physical purchase receipt comprising a set of physically printed line items corresponding to the set of physical products purchased in the shopping trip, wherein the set of physically printed line items are associated with a retailer-specific pattern, wherein the set of physical products comprises a physical product in multiple quantities, and wherein the mobile device is associated with a first user account and comprises the camera and a wireless communications module configured to communicate with a remote master product system;
b. converting the captured receipt image data to machine readable data;
c. generating a digital receipt model including a set of line items from the machine readable text data, the set of line items corresponding to the set of physically printed line items;
d. selecting a product profile to associate with a line item through one of a tiered set of receipt heuristics, comprising, through a direct identifier heuristic, searching for a product identifier from a line item that corresponds to a product profile, and through an approximate match heuristic, calculating a score for the line item compared to at least one product profile of a master product list, wherein calculating the score is based on the retailer-specific pattern for improving the product matching performance associated with a remote master product system;
e. at the remote master product system, accounting for a set of selected product profiles associated with the set of physical products purchased in the shopping trip, comprising accounting for the selected product profile of the line item and accounting for the multiple quantities of the physical product;
f. at the remote master product system, updating a purchase history of the first user account based on the set of selected product profiles;
g. generating purchase feedback tailored to the first user account based on the updated purchase history;
h. transmitting the purchase feedback to the mobile device;
i. repeating a) to h) for a second physical purchase receipt and a second mobile device associated with a second user account, thereby aggregating physical purchase receipt data at the remote master product system for improving the receipt processing and the matching performance associated with the remote master product system; and
j. at the remote server, sending data for the selected product profile, generated based on the first and second physical purchase receipts, to a device associated with a third user account in response to receipt of a request for the data.