Method for constructing an updateable database of subject behavior patterns

• US 7,373,311 B2
• Filed: 03/04/2005
• Issued: 05/13/2008
• Est. Priority Date: 06/01/1999
• Status: Expired due to Fees

• Current Assignee: Lucent Technologies, Inc. (Nokia Corporation)
• Original Assignee: Lucent Technologies, Inc. (Nokia Corporation)
• Inventors: Pinheiro, Jose′ C, Lambert, Diane, Sun, Don X
• Primary Examiner(s): MEINECKE DIAZ, SUSANNA M
• Application Number:
  US11/072,176
• Publication Number:
  US 20050154631A1
• Time in Patent Office: 1,166 Days
• Field of Search: 705/10
• US Class Current: 705/7.33
• CPC Class Codes: G06Q 10/063 Operations research, analysis or management
• CPC Class Codes: G06Q 10/0631 Resource planning, allocation, distributing or scheduling for enterprises or organisations
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0204 Market segmentation
• CPC Class Codes: G06Q 30/0255 based on user history
• CPC Class Codes: G06Q 40/00 Finance; Insurance; Tax strategies; Processing of corporate or income taxes

Abstract

Disclosed is a method for maintaining compact descriptions of the behavior of customers in a database of information relating to customer transactions. Of all possible variables that might be used to characterize customer transactions, a limited number are selected on the basis of discriminating power to define customer profiles. Customers are classified according to distinguishing features of their profiles. A new customer can be initially assigned to a particular class based on data characterizing his first few transactions.

Claims

1. A method to be used by or for an entity which transacts with customers to provide the customers with goods or services and which records transactional variables, that is, characteristics of customer transactions of the kind that have two or more possible values, lend themselves to treatment as statistical variables, and are useful for classifying customers according to their transactional behavior, comprising:

(a) obtaining initiation data of a new customer, that is, the values of transactional variables that characterize the first transaction or first few transactions with the new customer;

(b) storing the initiation data in a computer-accessible medium;

(c) accessing a data structure to obtain histograms, that is, stored records of frequency distributions of values of specified transactional variables based on historical customer data;

(d) selecting one or more histograms from the data structure as a representative profile that is to be imputed to the new customer, based on the initiation data, and subject to updating with data from further transactions with the new customer; and

(e) modifying the data structure such that the selected one or more histograms are associated with the new customer as a representative profile,

wherein the selecting step is carried out by computer and comprises;

(f) evaluating one or more indices, each index being a value or combination of values of specified transactional variables from the initiation data, wherein each index has predictive power with respect to at least one of the transactional variables that are represented by histograms stored in the data structure, predictive power being the property that the histograms of customers who share the same index tend to be separated by a smaller statistical distance than the histograms of randomly selected customers; and

(g) using the one or more indices to identify the one or more histograms to be selected for the representative profile.

2. The method of claim 1, wherein the selecting step includes adding individual candidate variables to form the representative profile until the incremental benefit of additional individual candidate indices is insignificant based on a predetermined threshold level.