Process and system for integrating information from disparate databases for purposes of predicting consumer behavior

• US 7,490,052 B2
• Filed: 02/23/2006
• Issued: 02/10/2009
• Est. Priority Date: 12/30/1998
• Status: Expired due to Fees

• Current Assignee: GfK US MRI, LLC
• Original Assignee: Experian Marketing Solutions Incorporated (Experian plc)
• Inventors: Kilger, Max F., Engel, William E.
• Primary Examiner(s): Sterrett; Jonathan G
• Application Number:
  US11/361,751
• Publication Number:
  US 20060143073A1
• Time in Patent Office: 1,083 Days
• Field of Search: 705/10
• US Class Current: 705/7.31
• CPC Class Codes: G06F 16/27 Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G06Q 30/0203 Market surveys; Market polls
• CPC Class Codes: G06Q 30/0204 Market segmentation
• CPC Class Codes: Y10S 707/99932 Access augmentation or optimizing
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

A process and system for integrating information stored in at least two disparate databases. The stored information includes consumer transactional information. According to the process and system, at least one qualitative variable which is common to each database is identified, and then transformed into one or more quantitative variables. The consumer transactional information in each database is then converted into converted information in terms of the quantitative variables. Thereafter, an integrated database is formed for predicting consumer behavior by combining the converted information from the disparate databases.

Claims

1. A computerized method for predicting individual behavior from information stored in a plurality of separate and disparate databases, the method comprising:

accessing a first database of the plurality of separate and disparate databases, the first database comprising a first plurality of records respectively associated with a first plurality of consumers, wherein the first plurality of records each comprise a first plurality of data variables associated with survey data provided from the respective of the first plurality of consumers;

accessing a second database of the plurality of separate and disparate databases, the second database comprising a second plurality of records associated with a second plurality of consumers, wherein the second plurality of records each comprise a second plurality of data variables associated with transactions of respective of the second plurality of consumers;

wherein the first plurality of data variables includes data variables that are not included in the second plurality of data variables, and wherein at least some consumers of the second plurality of consumers are not in the first plurality of consumers;

identifying a data variable which is common to the first data variables and the second data variables, wherein identifying the data variable comprises identifying at least one qualitative variable which is common to each database of the plurality of separate and disparate databases and transforming the at least one qualitative variable into one or more quantitative variables;

using a processing device, performing a cluster analysis across all separate databases, of the plurality of separate and disparate databases, using the common data variable from each database of the plurality of separate and disparate databases to form a plurality of clusters, at least one cluster of the plurality of clusters containing at least some of the first plurality of consumers and at least some of the second plurality of consumers that are not in both the first and second databases;

using a processing device, forming an integrated database by linking the plurality of clusters through the plurality of separate and disparate databases;

using a processing device, creating a behavioral model, using corresponding data of the plurality of clusters of the integrated database, for predicting individual behavior of respective of the second plurality of consumers.

2. The method of claim 1, further comprising:

selecting at least one discriminating subset of the one or more quantitative variables to create one or more statistical drivers; and

evaluating a plurality of individuals represented in the plurality of separate and disparate databases using the one or more statistical drivers.

3. The method of claim 2, wherein the selecting step further comprises:

identifying one or more industries which have discriminating consumers and grouping selected merchants into the at least one discriminating subset.

4. The method of claim 1, further comprising:

converting one or more clusters of the plurality of clusters into at least one super cluster; and

assigning the plurality of individuals to a corresponding cluster or super cluster using data from each database of the plurality of separate and disparate databases.

5. The method of claim 1, wherein the at least one qualitative variable is a merchant and the one or more quantitative variable comprises one or more of the following:

mean number of transactions per person for the merchant,

mean amount per transaction for the merchant,

mean household income of shoppers shopping at the merchant, or

mean proportion of the shoppers for a particular area of the merchant.

6. The method of claim 1, further comprising:

prior to forming the integrated database, weighting data of the plurality of separate and disparate databases to adjust for differences in size and in time encompassed.

7. The method of claim 1, wherein the information stored in at least one database of the plurality of separate and disparate databases further comprises consumer transactional information and has instances of purchasing behavior by consumers.

8. The method of claim 7, wherein at least one of the separate and disparate databases includes joint account information for at least two consumers, and wherein the method further comprises:

determining a consumer of the at least two consumers who generated at least a portion of the consumer transactional information.

34. The computerized method of claim 1, wherein the data variables associated with transactions of respective of the second plurality of consumers comprise data variables associated with purchasing transactions.

35. The computerized method of claim 1, wherein the data variables associated with transactions of respective of the second plurality of consumers comprise data variables associated with media viewing transactions.

9. A system for predicting individual behavior from information stored in a plurality of separate and disparate databases, the system comprising:

a data interface configured to access a first database of the plurality of separate and disparate databases having at least a first plurality of records respectively associated with a first plurality of individuals, wherein the first plurality of records each comprise a first plurality of data variables associated with survey data provided from the respective of the first plurality of individuals, the data interface being further configured to access a second database of the plurality of separate and disparate databases having at least a second plurality of records respectively associated with a second plurality of individuals, wherein the second plurality of records each comprise a second plurality of data variables associated with transactions of respective of the second plurality of individuals, and wherein at least some of the first plurality of data variables are not contained in the second plurality of data variables and the first database comprises information for at least a plurality of individuals who are not contained in the second database, the system comprising;

a storage device adapted to store an integrated database; and

a processing device coupled to the storage device, the processing device adapted to perform a cluster analysis across all separate databases, of the plurality of separate and disparate databases, using at least one variable which is common to each database of the plurality of separate and disparate databases to form a plurality of clusters, at least one cluster of the plurality of clusters containing at least some individuals of the first and second databases who are not in both the first and second databases, the processing device further adapted to form the integrated database by linking through the plurality of clusters the plurality of separate and disparate databases, and to generate a behavioral model, using corresponding data of the plurality of clusters of the integrated database, for predicting individual behavior, wherein the processing device is further adapted to transform at least one qualitative variable, which is common to each database of the plurality of separate and disparate databases, into one or more quantitative variables.

10. The system of claim 9, wherein the processing device is further adapted to determine at least one discriminating subset of the one or more quantitative variables to form one or more statistical drivers, and to evaluate a plurality of individuals represented in the plurality of separate and disparate databases using the one or more statistical drivers.

11. The system of claim 10, wherein the processing device is further adapted to select the at least one discriminating subset by identifying one or more industries which have discriminating consumers, and further adapted to group selected merchants into the at least one discriminating subset.

12. The system of claim 9 wherein the at least one qualitative variable is a merchant and the one or more quantitative variable comprises one or more of the following:

mean number of transactions per person for the merchant,

mean amount per transaction for the merchant,

mean household income of shoppers shopping at the merchant, or

mean proportion of the shoppers for a particular area of the merchant.

13. The system of claim 9, wherein the processing device is further adapted, prior to the formation of the integrated database, to weight data of the plurality of separate and disparate databases to adjust for differences in size and in time encompassed.

14. The system of claim 9, wherein the processing device is further adapted to convert one or more clusters of the plurality of clusters into at least one super cluster, and to assign the plurality of individuals to a corresponding cluster or super cluster using data from each database of the plurality of separate and disparate databases.

15. The system of claim 9, wherein the information stored in the plurality of separate and disparate databases further comprises one or more of the following types of information:

consumer transactional information;

instances of purchasing behavior by consumers;

survey information;

self-reported behavioral information;

or media behavior.

16. The system of claim 15, wherein at least one of said separate and disparate databases includes joint account information for at least two consumers, and wherein the processing device is further adapted to determine a consumer of the at least two consumers who generated at least a portion of the consumer transactional information.

17. A method for creating a behavioral model from information stored in a plurality of separate and disparate databases, a first database of the plurality of separate and disparate databases having at least a plurality of data variables that are not contained in a second database of the plurality of separate and disparate databases, and the first database containing information for at least a plurality of individuals who are not contained in the second database, the method comprising:

determining a plurality of variables from each separate database, at least a portion of the plurality of variables common to each separate database of the plurality of separate and disparate databases, wherein the determining comprises selecting at least one qualitative variable which is common to each separate database of the plurality of separate and disparate databases and transforming the at least one qualitative variable from each separate database to a plurality of quantitative variables;

using a processing device, performing a first cluster analysis across all separate databases, of the plurality of separate and disparate databases, using corresponding data of the plurality of variables common to each separate database of the plurality of separate and disparate databases to create a plurality of cluster solutions, at least a first cluster solution of the plurality of cluster solutions containing at least some individuals of the first and second databases who are not in both the first and second databases;

using a processing device, linking information stored in the plurality of separate and disparate databases according to the first cluster solution, of the plurality of cluster solutions, the information stored in at least the first and the second separate and disparate databases; and

validating the first cluster solution of the plurality of cluster solutions as a discriminatory behavioral model for predicting individual behavior.

18. The method of claim 17, further comprising:

performing a principal components analysis on the plurality of quantitative variables using the information stored in each separate database of the plurality of separate and disparate databases to create a plurality of statistical drivers.

19. The method of claim 18, further comprising:

converting the information stored in the plurality of separate and disparate databases according to the plurality of statistical drivers to create the corresponding data of the plurality of variables.

20. The method of claim 18, wherein the validation step further comprises:

determining whether the cluster solution provides corresponding discrimination on a plurality of other variables which are not statistical drivers in the plurality of separate and disparate databases.

21. The method of claim 18, further comprising:

standardizing the plurality of quantitative variables;

transforming the standardized plurality of quantitative variables to be substantially orthogonal; and

differentially weighting the orthogonal, standardized plurality of quantitative variables to form the plurality of statistical drivers.

22. The method of claim 17, further comprising:

evaluating corresponding discrimination power of the plurality of statistical drivers using a second cluster analysis.

23. The method of claim 17, wherein at least one separate database of the plurality of separate and disparate databases stores one or more of the following types of information:

behavioral and attitudinal information;

consumer transactional information; and

media consumption information.

24. The method of claim 17, further comprising:

describing each cluster of a plurality of clusters of the validated simultaneous cluster solution using information stored in at least one separate database of the plurality of separate and disparate databases.

25. The method of claim 17, wherein the validation step further comprises:

determining whether the cluster solution provides corresponding discrimination separately within each separate database of the plurality of separate and disparate databases.

26. A system for creating a behavioral model from information stored in a plurality of separate and disparate databases, a first database of the plurality of separate and disparate databases having at least a plurality of data variables associated with survey data provided from the respective of the first plurality of consumers, the first plurality of data variables having at least some data variables that are not contained in a second database of the plurality of separate and disparate databases having a second plurality of data variables associated with transactions of respective of the second plurality of consumers, and the first database containing information for at least a plurality of individuals who are not contained in the second database, the system comprising:

a storage device storing data from one or more of the plurality of separate and disparate databases; and

a processing device coupled to the storage device, the processing device adapted

to determine a plurality of variables from each separate database, at least a portion of the plurality of variables common to each separate database of the plurality of separate and disparate databases, wherein the plurality of variables are determined by selecting at least one qualitative variable which is common to each separate database of the plurality of separate and disparate databases, and transforming the at least one qualitative variable from each separate database to a plurality of quantitative variables;

to perform a first cluster analysis across all separate databases, of the plurality of separate and disparate databases, using corresponding data of the plurality of variables common to each separate database of the plurality of separate and disparate databases to create a plurality of cluster solutions, at least one cluster solution of the plurality of cluster solutions containing at least some individuals of the first and second databases who are not in both the first and second databases;

to link through a cluster solution, of the plurality of cluster solutions, the information stored in the plurality of separate and disparate databases; and

to validate the cluster solution of the plurality of cluster solutions as a discriminatory behavioral model for predicting individual behavior.

27. The system of claim 26, wherein the processing device is further adapted to perform a principal components analysis on the plurality of quantitative variables using the information stored in each separate database of the plurality of separate and disparate databases to create a plurality of statistical drivers.

28. The system of claim 27, wherein the processing device is further adapted to convert the information stored in the plurality of separate and disparate databases according to the plurality of statistical drivers to create the corresponding data of the plurality of variables.

29. The system of claim 27, wherein the processing device is further adapted to validate the cluster solution by determining whether the cluster solution provides corresponding discrimination on a plurality of other variables which are not statistical drivers in the plurality of separate and disparate databases.

30. The system of claim 27, wherein the processing device is further adapted to standardize the plurality of quantitative variables;

transform the standardized plurality of quantitative variables to be substantially orthogonal; and

to differentially weight the orthogonal, standardized plurality of quantitative variables to form the plurality of statistical drivers.

31. The system of claim 27, wherein the processing device is further adapted to evaluating a corresponding discrimination power of the plurality of statistical drivers using a second cluster analysis.

32. The system of claim 26, wherein at least one separate database of the plurality of separate and disparate databases stores one or more of the following types of information;

behavioral and attitudinal information;

consumer transactional information; and

media consumption information.

33. The system of claim 26, wherein the processing device is further adapted to determine whether the cluster solution provides corresponding discrimination separately within each separate database of the plurality of separate and disparate databases.