System and method for automated retail product accounting

Abstract

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.

What is claimed is: 
     
         1 . A method for tracking purchased physical products, the method comprising:
 a. collecting receipt image data of a first physical receipt associated with a first user;   b. generating a receipt model including a first line item based on the receipt image data, the first line item corresponding to a physical product;   c. mapping the first line item to a product profile based on a product identifier associated with the product profile, the product identifier identifying the physical product;   d. repeating a) to c) for a second physical receipt associated with a second user, thereby mapping a second line item associated with the second physical receipt to the product profile;   e. generating data for the product profile based on the first and the second line items; and   f. transmitting the data to a device associated with a third user.   
     
     
         2 . The method of  claim 1 ,
 wherein generating the receipt model comprises identifying a retailer associated with the first physical receipt based on the receipt image data, and   wherein mapping the first line item to the product profile comprises matching the first line item to the product identifier based on a retailer-specific pattern associated with the retailer, wherein the retailer-specific pattern improves accuracy of mapping the first line item.   
     
     
         3 . The method of  claim 1 , further comprising:
 determining a plurality of alternative product identifiers based on the product identifier, the alternative product identifiers identifying the physical product,   wherein mapping the first line item to the product profile comprises mapping the first line item based on an exact match between the first line item and an alternative product identifier of the plurality of alternative product identifiers.   
     
     
         4 . The method of  claim 3 , wherein the product identifier comprises a string identifier, wherein determining the plurality of alternative product identifiers comprises:
 generating the alternative product identifier based on substitution of a character of the string identifier with a substitute character; and   associating the alternative product identifier with the product profile.   
     
     
         5 . The method of  claim 1 , wherein mapping the first line item comprises:
 detecting an error in the first line item;   in response to detecting the error, determining a character modification to a character of the first line item; and   mapping the first line item to the product profile based on the product identifier and the modification.   
     
     
         6 . The method of  claim 5 , wherein detecting the error comprises:
 generating comparisons between the first line item and each of a plurality of product identifiers identifying a plurality of physical products; and   detecting the error in the first line item when inexact matches for each of the comparisons are detected.   
     
     
         7 . The method of  claim 5 , wherein mapping the first line item comprises:
 determining a set of character modifications for transforming the first line item into the product identifier, wherein the set of character modifications comprises the character modification;   calculating a number of character modifications in the set of character modifications;   wherein mapping the first line item comprises mapping the first line item to the product profile when the number of character modifications satisfies a threshold condition.   
     
     
         8 . The method of  claim 7 , wherein calculating the number of character modifications comprises:
 identifying a modification type of the character modification;   assigning a weight to the character modification based on the modification type; and   determining the number of character modifications based on the weight.   
     
     
         9 . The method of  claim 1 , further comprising:
 automatically transforming the receipt image data into machine readable text;   determining a confidence level associated with the machine readable text;   in response to the confidence level being below a threshold level, collecting a user-entered translation of the receipt image data; and   generating the first line item of the receipt model based on the user-entered translation.   
     
     
         10 . A method for tracking purchased products, the method comprising:
 a. generating a first receipt model for a first receipt associated with a first user account and a product;   b. mapping the first receipt model to a product profile based on a product identifier associated with the product profile, the product identifier identifying the product;   c. updating a first purchase history associated with the first user account based on the product profile;   d. repeating a) to c) for a second receipt model and a second receipt, thereby updating a second purchase history associated with a second user account based on the product profile;   e. generating data for the product profile based on the first and the second purchase histories; and   f. transmitting the data to a device associated with a third user account.   
     
     
         11 . The method of  claim 10 , wherein updating the first purchase history and updating the second purchase history are performed at a remote system, the method further comprising, at the remote system:
 determining purchase feedback tailored to the second user account based on the first and the second purchase histories; and   transmitting the purchase feedback to a user device associated with the second user account.   
     
     
         12 . The method of  claim 10 , further comprising:
 prior to generating the first receipt model, promoting an electronic reward for purchase of the product; and   in response to verifying the purchase based on the first purchase history, associating the electronic reward with the first user account,   wherein generating the data for the product profile comprises generating the data based on the electronic reward, the first purchase history, and the second purchase history.   
     
     
         13 . The method of  claim 12 , wherein verifying the purchase comprises:
 identifying a retailer associated with the purchase based on receipt image data of a retailer identifier physically printed on the first receipt; and   verifying the retailer based on a match with a supported retailer associated with the electronic reward.   
     
     
         14 . The method of  claim 10 , further comprising:
 collecting image data of a code associated with a set of line items physically printed on the first receipt, wherein generating the first receipt model comprises generating the first receipt model based on the image data of the code;   mapping the first receipt model to a set of product profiles comprising the product profile, the set of product profiles identifying a set of products corresponding to the set of line items; and   updating the first purchase history based on the set of product profiles.   
     
     
         15 . The method of  claim 14 , wherein generating the first receipt model based on the receipt image data of the code comprises:
 generating an API request based on the receipt image data of the code;   in response to transmitting the API request to an external product database, receiving a set of item descriptions for the set of physical products; and   generating the first receipt model based on the set of item descriptions.   
     
     
         10 . The method of  claim 10 , further comprising:
 for each alternative product identifier of a set of alternative product identifiers, determining a number of character modifications for transforming the alternative product identifier into the product identifier;   in response to the number of character modifications satisfying a threshold condition, clustering the corresponding alternative product identifier into a subset of the set of alternative product identifiers; and   associating the subset of alternative product identifiers with the product profile.   
     
     
         17 . The method of claim  16 , wherein the product identifier and the set of alternative product identifiers are associated with a shared retailer. 
     
     
         18 . The method of  claim 10 , further comprising:
 collecting a user input associated with the first user account;   accessing an outside account associated with an external product database based on the user input; and   querying the outside product database with the outside account for an item description corresponding to the product,   wherein generating the first receipt model comprises generating the first receipt model based on the item description.   
     
     
         19 . The method of  claim 18 , wherein accessing the outside account and querying the outside product database are performed at a user device associated with the first user account and the outside account, and wherein generating the first receipt model is performed at a remote server. 
     
     
         20 . The method of  claim 10 , wherein the first receipt comprises a web receipt, the method further comprising:
 collecting machine readable data corresponding to the web receipt from an external database;   generating a query to a search engine based on the machine readable data; and   automatically scraping an item description for the product based on a response to the query, wherein generating the first receipt model comprises generating the first receipt model based on the item description.