Household level segmentation method and system

Abstract

Methods and apparatus for household level segmentation are disclosed. An example method to classify consumers in clusters includes receiving population data indicative of a population of consumers and receiving a plurality of profiles, at least one profile to evaluate partitioning of the population of consumers. The example method also includes selecting at least one of the plurality of profiles based on a count limit value in accordance with a classification tree dimension split.

1 . (canceled) 
     
     
         2 . A method to classify consumers in population partitions, comprising:
 building a first classification tree comprising a population of consumers;   applying a count limit value for decision splits of the first classification tree to derive nodes or terminal nodes, the nodes and the terminal nodes representing respective first and second partitions of the population of consumers; and   identifying the terminal nodes when one of the respective partitions of the population of consumers is homogeneous with respect to both behavior and demographics.   
     
     
         3 . A method as defined in  claim 2 , further comprising applying a first sequence of decisions of the decision splits to derive the nodes or the terminal nodes of the first classification tree. 
     
     
         4 . A method as defined in  claim 2 , further comprising calculating a first measure of behavior similarity for the terminal nodes associated with the first classification tree. 
     
     
         5 . A method as defined in  claim 4 , further comprising applying a second sequence of decisions to build a second classification tree having second nodes or second terminal nodes. 
     
     
         6 . A method as defined in  claim 5 , further comprising calculating a second measure of behavior similarity for the second terminal nodes associated with the second classification tree. 
     
     
         7 . A method as defined in  claim 6 , further comprising comparing the first classification tree to the second classification tree to select a first sequence of decisions associated with the first classification tree or the second sequence of decisions associated with the second classification tree. 
     
     
         8 . A method as defined in  claim 2 , further comprising removing from consideration a third partition of the population of consumers when the count limit is not exceeded. 
     
     
         9 . A method as defined in  claim 2 , further comprising retaining at least one of the nodes or the terminal nodes based on a Gini impurity measure. 
     
     
         10 . A tangible computer-readable storage medium comprising instructions stored thereon that, when executed, cause a machine to, at least:
 build a first classification tree comprising a population of consumers;   apply a count limit value for decision splits of the first classification tree to derive nodes or terminal nodes, the nodes and the terminal nodes to represent respective first and second partitions of the population of consumers; and   identify the terminal nodes when one of the respective partitions of the population of consumers is homogeneous with respect to both behavior and demographics.   
     
     
         11 . A tangible computer-readable storage medium as defined in  claim 10 , further comprising instructions to, when executed, apply a first sequence of decisions of the decision splits to derive the nodes or the terminal nodes of the first classification tree. 
     
     
         12 . A tangible computer-readable storage medium as defined in  claim 10 , further comprising instructions to, when executed, calculate a first measure of behavior similarity for the terminal nodes associated with the first classification tree. 
     
     
         13 . A tangible computer-readable storage medium as defined in  claim 12 , further comprising instructions to, when executed, apply a second sequence of decisions to build a second classification tree having second nodes or second terminal nodes. 
     
     
         14 . A tangible computer-readable storage medium as defined in  claim 12 , further comprising instructions to, when executed, calculate a second measure of behavior similarity for the second terminal nodes associated with the second classification tree. 
     
     
         15 . A tangible computer-readable storage medium as defined in  claim 14 , further comprising instructions to, when executed, compare the first classification tree to the second classification tree to select a first sequence of decisions associated with the first classification tree or the second sequence of decisions associated with the second classification tree. 
     
     
         16 . A tangible computer-readable storage medium as defined in  claim 10 , further comprising instructions to, when executed, remove from consideration a third partition of the population of consumers when the count limit is not exceeded. 
     
     
         17 . A tangible computer-readable storage medium as defined in  claim 10 , further comprising instructions to, when executed, retain at least one of the nodes or the terminal nodes based on a Gini impurity measure. 
     
     
         18 . A segmentation system to classify consumers in population partitions, comprising:
 a partitioning module to build a first classification tree comprising a population of consumers;   a profile definitions module to apply a count limit value for decision splits of the first classification tree to derive nodes or terminal nodes, the nodes and the terminal nodes to represent respective first and second partitions of the population of consumers; and   a segment definitions module to identify the terminal nodes when one of the respective partitions of the population of consumers is homogeneous with respect to both behavior and demographics.   
     
     
         19 . A segmentation system as defined in  claim 18 , wherein the profile definitions module is to apply a first sequence of decisions of the decision splits to derive the nodes or the terminal nodes of the first classification tree. 
     
     
         20 . A segmentation system as defined in  claim 18 , wherein the segment definitions module is to calculate a first measure of behavior similarity for the terminal nodes associated with the first classification tree. 
     
     
         21 . A segmentation system as defined in  claim 20 , wherein the profile definitions module is to apply a second sequence of decisions to build a second classification tree having second nodes or second terminal nodes.