System and method for matching purchase records to mobile devices

Abstract

A mobile device and sales transaction matching system matches sales records to a buyer operating a mobile device. The sales transactions are generated by point-of-sales system, The matching system defines a mobile data sphere for each mobile data point and a transaction sphere for each purchase record before determining whether they overlap or not. For each mobile data point, there is a list of corresponding overlapping transactions. Similarity function is created and applied to the two dimensional maps of overlapping transactions corresponding to a list of mobile data points of the mobile device. The application of the similarity function determines the transactions that are made by the buyer. Both the mobile data points and purchase records are public data. Each sphere indicates a physical distance and a time radius.

What is claimed is: 
     
         1 - 6  (canceled) 
     
     
         7 . A mobile device and sales transaction matching system comprising:
 i) a database;   ii) a server computer having a memory, a processing unit accessing said memory, a network interface coupled to said processing unit and operatively coupled to said database, and a specialized server software application operating on said processing unit, wherein said server computer is adapted to communicate with a sales transaction server system and a mobile device location data server system;   iii) said specialized server software application adapted to:
 1) retrieve a set of purchase records from said sales transaction server system; 
 2) retrieve a set of mobile data points from said mobile device location data server system, wherein said set of mobile data points corresponds to a first mobile device; 
 3) for each mobile data point within said set of mobile data point, define a mobile data sphere; 
 4) for each purchase record within said set of purchase records, define a transaction sphere; 
 5) compare each said mobile data sphere with each said transaction sphere to determine a list of over lapping purchase records for each mobile data point, thereby forming a plurality of lists of over lapping purchase records for said first mobile device, wherein said mobile data sphere of said each mobile data point overlaps with said transaction sphere of each purchase record within said list of over lapping purchase records corresponding to said each mobile data point; 
 6) create a similarity function for determining similarity between transactions; and 
 7) apply said similarity function to each pair of purchase records within said plurality of lists of over lapping purchase records to determine purchase records of a user of said first mobile device. 
   
     
     
         8 . The mobile device and sales transaction matching system of  claim 7  wherein:
 1) each said mobile data sphere indicates a physical distance from a location of corresponding said mobile data point and a radius time from a timestamp of corresponding said mobile data point; and 
 2) each said transaction sphere indicates a physical distance from a store location of corresponding said purchase record and radius of time from a timestamp of corresponding said purchase record. 
 
     
     
         9 . The mobile device and sales transaction matching system of  claim 8  wherein a mobile data sphere is said to be overlapping with a transaction sphere based on a Euclidean distance or a Manhattan metric. 
     
     
         10 . The mobile device and sales transaction matching system of  claim 8  wherein said similarity function returns a value between zero and one when applied to a pair of purchase records. 
     
     
         11 . The mobile device and sales transaction matching system of  claim 10  wherein said similarity function is a Jaccard Index, a modified Jaccard Index, a Tanimoto similarity, or a Sorensen-Dice Index. 
     
     
         12 . The mobile device and sales transaction matching system of  claim 10  wherein said similarity function is created based on a set of training purchase records. 
     
     
         13 . The mobile device and sales transaction matching system of  claim 7  wherein said set of purchase records is public data and said set of mobile data points is public data. 
     
     
         14 . The mobile device and sales transaction matching system of  claim 7  wherein each purchase record within said set of purchase records includes a store location, a sales timestamp, and a product identifier, and wherein each data point within said set of mobile data points includes a mobile device identifier, a mobile device location, and a mobile device timestamp.