Method and system for credit card holder identification

Abstract

The present invention is associated with a method that takes a subsequence of a credit card number, in particular the suffix (or last n digits), and transaction data associated with the cardholder. For each subsequence of the credit card number, a set of data may be used in order to uniquely identify the cardholder. The identification of the cardholder may be performed in a particular search space that is composed of all possible credit card numbers and includes the transaction data associated with the cardholder.

1 .- 19 . (canceled) 
     
     
         20 . An apparatus comprising at least one processor and at least one memory storing instructions that, when executed by the at least one processor, cause the apparatus to:
 receive a credit card number suffix, wherein the credit card number suffix is associated with a cardholder;   receive available transaction data for the cardholder;   access a data structure of aggregated transaction data received from a plurality of different data collection mechanisms, wherein the credit card number suffix is associated with multiple transactions of the aggregated transaction data;   regenerate, based on the credit card number suffix, a set of possible valid credit card numbers for the cardholder, wherein each possible valid credit card number contains the credit card number suffix;   generate concatenated credit card numbers based at least in part on the regenerated set of possible valid credit card numbers and the received available transaction data;   map the concatenated credit card numbers to one or more indices of a hash table;   determine if the received available transaction data is already in the data structure at the locations corresponding to the one or more indices;   if the received available transaction data is already in the data structure at the locations corresponding to one of the indices, identify an indices-based cardholder identifier associated with such locations and associate the indices-based cardholder identifier with the received credit card number suffix; and   if the received available transaction data is not already in the data structure at the locations corresponding to one of the indices, modify the data structure to create an updated data structure, wherein modifying the data structure comprises:
 applying a pattern recognition algorithm to the data structure of aggregated transaction data to identify a most similar data record; 
 identifying a similarity-based cardholder identifier associated with the most similar data record; 
 updating the data structure to add the received available transaction data that was not already in the data structure; 
 associating the similarity-based cardholder identifier with the credit card number suffix; and 
 updating the pattern recognition algorithm using the updated data structure of aggregated transaction data as a training set. 
   
     
     
         21 . The apparatus of  claim 20 , wherein modifying the data structure further comprises updating a weight value associated with the received available transaction data added to the data structure. 
     
     
         22 . The apparatus of  claim 20 , wherein if the received available transaction data is already in the data structure at the locations corresponding to one of the indices, the instructions further cause the apparatus to update a weight value associated with the stored received available transaction data. 
     
     
         23 . The apparatus of  claim 20 , wherein the pattern recognition algorithm includes at least one of unsupervised learning, supervised learning, semi-supervised learning, reinforcement learning, association rules learning, Bayesian learning, or solving for probabilistic graphical models. 
     
     
         24 . The apparatus of  claim 20 , wherein the data structure of aggregated transaction data defines cardholder profiles comprising a hashing representation of a credit card number and one or more of cardholder email, start date, end date, registration data, birthdate, name on card, first activity date, last activity date, card type, address, city, state, zip code, and country code. 
     
     
         25 . The apparatus of  claim 20 , wherein the data structure of aggregated transaction data defines cardholder profiles, each cardholder profile corresponding to a cardholder buying fingerprint. 
     
     
         26 . The apparatus of  claim 20 , wherein the instructions further cause the apparatus to:
 apply data clustering strategies and the pattern recognition algorithm to the data structure of aggregated transaction data to classify a plurality of cardholder identifiers into clusters based on the aggregated transaction data;   create a buying fingerprint for each of the plurality of cardholder identifiers based on the aggregated transaction data and on the clusters; and   store the buying fingerprint in the data structure of aggregated transaction data associated with each of the plurality of cardholder identifiers.   
     
     
         27 . The apparatus of  claim 20 , wherein the available transaction data relates to a commercial transaction by the cardholder with a merchant. 
     
     
         28 . The apparatus of  claim 20 , wherein the available transaction data comprises one or more of cardholder name, expiration date, or service code. 
     
     
         29 . A computer-implemented method comprising:
 receiving a credit card number suffix, wherein the credit card number suffix is associated with a cardholder;   receiving available transaction data for the cardholder;   accessing a data structure of aggregated transaction data received from a plurality of different data collection mechanisms, wherein the credit card number suffix is associated with multiple transactions of the aggregated transaction data;   regenerating, based on the credit card number suffix, a set of possible valid credit card numbers for the cardholder, wherein each possible valid credit card number contains the credit card number suffix;   generating concatenated credit card numbers based at least in part on the regenerated set of possible valid credit card numbers and the received available transaction data;   mapping the concatenated credit card numbers to one or more indices of a hash table;   determining if the received available transaction data is already in the data structure at the locations corresponding to the one or more indices;   if the received available transaction data is already in the data structure at the locations corresponding to one of the indices, identifying an indices-based cardholder identifier associated with such locations and associating the indices-based cardholder identifier with the received credit card number suffix.   
     
     
         30 . The method of  claim 29 , further comprising:
 if the received available transaction data is not already in the data structure at the locations corresponding to one of the indices, modifying the data structure to create an updated data structure, wherein modifying the data structure comprises:
 applying a pattern recognition algorithm to the data structure of aggregated transaction data to identify a most similar data record; 
 identifying a similarity-based cardholder identifier associated with the most similar data record; 
 updating the data structure to add the received available transaction data that was not already in the data structure; 
 associating the similarity-based cardholder identifier with the credit card number suffix; and 
 updating the pattern recognition algorithm using the updated data structure of aggregated transaction data as a training set. 
   
     
     
         31 . The method of  claim 30 , wherein modifying the data structure further comprises updating a weight value associated with the received available transaction data added to the data structure. 
     
     
         32 . The method of  claim 29 , further comprising updating a weight value associated with the stored received available transaction data. 
     
     
         33 . The method of  claim 29 , wherein the pattern recognition algorithm includes at least one of unsupervised learning, supervised learning, semi-supervised learning, reinforcement learning, association rules learning, Bayesian learning, or solving for probabilistic graphical models. 
     
     
         34 . The method of  claim 29 , wherein the data structure of aggregated transaction data defines cardholder profiles comprising a hashing representation of a credit card number and one or more of cardholder email, start date, end date, registration data, birthdate, name on card, first activity date, last activity date, card type, address, city, state, zip code, and country code. 
     
     
         35 . The method of  claim 29 , wherein the data structure of aggregated transaction data defines cardholder profiles, each cardholder profile corresponding to a cardholder buying fingerprint. 
     
     
         36 . The method of  claim 29 , further comprising:
 applying data clustering strategies and the pattern recognition algorithm to the data structure of aggregated transaction data to classify a plurality of cardholder identifiers into clusters based on the aggregated transaction data;   creating a buying fingerprint for each of the plurality of cardholder identifiers based on the aggregated transaction data and on the clusters; and   storing the buying fingerprint in the data structure of aggregated transaction data associated with each of the plurality of cardholder identifiers.   
     
     
         37 . The method of  claim 29 , wherein the available transaction data relates to a commercial transaction by the cardholder with a merchant. 
     
     
         38 . The method of  claim 29 , wherein the available transaction data comprises one or more of cardholder name, expiration date, or service code.