Connecting decisions through customer transaction profiles

Abstract

An apparatus and method for developing financial risk decisions for a customer associated with a number of different financial services/channels are disclosed. A hierarchy of relationships among the financial services/channels is generated. Transactional behaviors of the customer related to each of the financial services/channels is summarized, using one or more analytical approaches executed on the hierarchy of relationships, to generate a customer level transactional behavior summary. A customer profile associated with the customer is generated which includes the transactional behavior summary and aggregated information on recent financial transactions associated with each of the financial services/channels. A score for a risk decision can be generated for one or more specific services/channels, based on the customer profile.

1 . A method for developing financial risk decisions for a customer associated with a plurality of financial accounts, the method being executed by a computer system having one or more processors, the method comprising:
 generating, by the one or more processors, a hierarchy of relationships of the customer with the plurality of financial accounts based on risk attributes and fraud attributes determined from each of the plurality of financial accounts, the hierarchy of relationships being based on a self-learning outlier analytics approach;   generating, by the one or more processors, a customer level transactional behavior summary according to the hierarchy of relationships;   receiving aggregated information on recent financial transactions associated with each of the plurality of financial accounts;   generating, by the one or more processors, a customer profile associated with the customer, the customer profile including the transactional behavior summary and the aggregated information.   
     
     
         2 . The method in accordance with  claim 1 , further comprising generating, by the one or more processors, a score for one of the financial risk decisions based on the customer profile. 
     
     
         3 . The method in accordance with  claim 1 , further comprising generating, by the one or more processors, a financial account summary based on the aggregated information on recent financial transactions associated with each of the plurality of financial accounts and their respective analytic monitors. 
     
     
         4 . The method in accordance with  claim 3 , further comprising storing, by the one or more processors, the transactional behavior summary and financial account transaction scores in the financial account summary in the customer profile. 
     
     
         5 . The method in accordance with  claim 1 , wherein generating the customer profile includes using a supervised customer model scoring approach based on a supervised customer model developed for the customer. 
     
     
         6 . The method in accordance with  claim 1 , wherein the self-learning outlier analytics approach further includes a rules based customer decision approach based on a plurality of scoring rules developed for the customer. 
     
     
         7 . The method in accordance with  claim 1 , wherein the self-learning outlier analytics customer approach is based on a fixed set of at least one of the (i) risk attributes and (ii) fraud attributes scaled based on customer groupings, each customer grouping corresponds to customers that have values associated with the at least one of risk attributes and fraud attributes within a corresponding predetermined range, and the self-learning outlier analytics customer approach is configured to determine outliers within at least one customer grouping of the customer groupings. 
     
     
         8 . An apparatus executing computer instructions for developing financial risk decisions for a customer associated with a plurality of financial accounts, the apparatus comprising:
 a computer system including:
 a processor; 
 a main memory coupled to the processor; and 
   persistent storage, associated with the computer system, the computer system executing:
 instructions for generating, by the one or more processors, a hierarchy of relationships of the customer with the plurality of financial accounts based on risk attributes and fraud attributes determined from each of the plurality of financial accounts, the hierarchy of relationships being based on a self-learning outlier analytics approach; 
 instructions for generating, by the one or more processors, a customer level transactional behavior summary according to the hierarchy of relationships; 
 instructions for receiving aggregated information on recent financial transactions associated with each of the plurality of financial accounts; 
 instructions for generating, by the one or more processors, a customer profile associated with the customer, the customer profile including the transactional behavior summary and the aggregated information. 
   
     
     
         9 . The apparatus in accordance with  claim 8 , wherein the self-learning outlier analytics customer approach is based on a fixed set of at least one of the (i) risk attributes and (ii) fraud attributes scaled based on customer groupings, each customer grouping corresponds to customers that have values associated with the at least one of risk attributes and fraud attributes within a corresponding predetermined range, and the self-learning outlier analytics customer approach is configured to determine outliers within at least one customer grouping of the customer groupings. 
     
     
         10 . The apparatus in accordance with  claim 9 , wherein the computer system further executes instructions for generating a score for one of the financial risk decisions based on the customer profile. 
     
     
         11 . The apparatus in accordance with  claim 9 , wherein the computer system further executes instructions for generating a financial account summary based on the aggregated information on recent financial transactions associated with each of the plurality of financial accounts and their respective analytic monitors. 
     
     
         12 . The apparatus in accordance with  claim 11 , wherein the computer system further executes instructions for storing the transactional behavior summary and associated financial account transaction scores in the financial account summary in the customer profile. 
     
     
         13 . The apparatus in accordance with  claim 9 , wherein the analytical approach includes a supervised customer model scoring approach based on a supervised customer model developed for the customer. 
     
     
         14 . The apparatus in accordance with  claim 9 , wherein the analytical approach includes a rules based customer decision approach based on a plurality of scoring rules developed for the customer. 
     
     
         15 . A method for developing financial risk decisions for a customer associated with a plurality of financial accounts, the method being executed by a computer system having one or more processors, the method comprising:
 generating, by the one or more processors, a hierarchy of relationships among the plurality of accounts, the hierarchy of relationships being based on at least one of risk attributes and fraud attributes determined from each of the plurality of financial accounts;   receiving, by the one or more processors, data representative of transactional behaviors of the customer related to each of the plurality of accounts and their respective analytic monitors;   summarizing, by the one or more processors, the transactional behaviors using an analytical approach executed on the hierarchy of relationships to generate a transactional behavior summary, the transactional behavior summary including associated account transaction scores;   receiving, by the one or more processors, data representative of aggregated information on recent financial transactions associated with each of the plurality of accounts; and   generating, by the one or more processors, a customer profile associated with the customer, the customer profile including the transactional behavior summary and the aggregated information;   
     
     
         16 . The method in accordance with  claim 15 , wherein the analytical approach includes a self-learning outlier analytics customer approach, and wherein the self-learning outlier analytics customer approach is based on a fixed set of at least one of (i) risk attributes and (ii) fraud attributes scaled based on customer groupings, each customer grouping corresponds to customers that have values associated with the at least one of risk attributes and fraud attributes within a corresponding predetermined range, and the self-learning outlier analytics customer approach is configured to determine outliers within at least one customer grouping of the customer groupings. 
     
     
         17 . The method in accordance with  claim 15 , further comprising generating, by the one or more processors, a score for one of the financial risk decisions based on the customer profile. 
     
     
         18 . The method in accordance with  claim 15 , further comprising aggregating, by the one or more processors, information on the recent financial transactions associated with each of the plurality of accounts to generate the data representative of the aggregated information. 
     
     
         19 . The method in accordance with  claim 15 , wherein the analytical approach includes a supervised customer model scoring approach based on a supervised customer model developed for the customer. 
     
     
         20 . The method in accordance with  claim 15 , wherein the analytical approach includes a rules based customer decision approach based on a plurality of scoring rules developed for the customer.