Method and system that identify market segments and that facilitate targeted information distribution
The current document is directed to automated market-segment-discovery methods and systems that may be incorporated within, or used in combination with, various types of analysis and optimization automated systems for automated discovery of market segments for subsequent use in targeted marketing and information distribution. In one implementation, a log of visitor records collected by an analysis and/or optimization system is processed to generate a segment-discovery tree. Construction of the segment-discovery tree produces a set of candidate market-segment-defining rules. Various different techniques and metrics can be applied to produce a set of market-segment-defining rules from these candidate rules. The market-segment-defining rules can then be exported to marketing systems or subsystems to facilitate targeted marketing and information distribution.
1. A system comprising:
one or more processors;
one or more memories; and
computer instructions, stored in the one or more memories and executed by the one or more processors that cause the one or more processors to perform operations including:
receiving a session dataset including, for each of one or user sessions, one or more interaction characteristics that characterize user interaction with a first version of a webpage during a session;
defining, from the session dataset a set of interaction characteristics;
receiving a definition of a target characteristic group, the definition identifying a first subset of the set of interaction characteristics;
identifying a second subset of the set of interaction characteristics, each characteristic in the second subset being absent from the first subset;
generating, using the second subset of the set of interaction characteristics, a segment-discovery tree by:
generating a plurality of level one nodes, each level one node of the plurality of level one nodes corresponding to an interaction characteristic of the second subset; and
generating a plurality of level two nodes, the plurality of level two nodes corresponding to a combined interaction characteristic, the combined interaction characteristic including a corresponding interaction characteristic of two or more level one nodes;
determining, for each level one node of the plurality of level one nodes in the segment-discovery tree, a probability that an interaction characteristic corresponding to the level one node correlates with the target characteristic group;
determining, for each level two node of the plurality of level two nodes in the segment-discovery tree, a probability that the combined interaction characteristic corresponding to the level two node correlate with the target characteristic group;
determining that a probability corresponding to each of a plurality of level one nodes and each of a plurality of level two nodes exceeds a threshold;
defining an interaction dataset that includes, for each level one node and each level two node with a corresponding probability that exceeds the threshold, the interaction characteristic corresponding to the level one node and the combined interaction characteristic corresponding to the level two node; generating an alternative version of the webpage based on the interaction dataset by modifying one or more nodes of a document object model corresponding to the webpage, the alternative version of the webpage including at least some content of a set of content that is different from the set of content rendered by the first version of the webpage; and
transmitting the alternative version of the webpage to one or more users that satisfy the interaction dataset within a particular geographical region, wherein the first version of the webpage is configured to be served to users that do not satisfy the interaction dataset.
2. The system of claim 1 , wherein each of the one or more interaction characteristics includes a set of attributes, and wherein the definition of the target characteristic group identifies a subset of the set of attributes for each interaction characteristic of the first subset of the set of interaction characteristics.
3. The system of claim 1 ,
wherein the combined interaction characteristic represents the corresponding interaction characteristic of two or more level one nodes as a Boolean expression.
4. The system of claim 1 , wherein generating the segment-discover tree further includes: generating a contingency table, the contingency table including a first row and second row;
determining, for each level one node of the plurality of level one nodes, a frequency value of an interaction characteristic corresponding to the level one node, the frequency value indicating a number of user sessions in which the interaction characteristic characterized a user session;
storing, in the first row, the frequency value corresponding to each level one node of the plurality of level one nodes;
determining, for each level one node of the plurality of level one nodes, an independence value of an interaction characteristic corresponding to the level one node, the independence value indicating a number of user sessions in which the interaction characteristic did not characterize a user session; and
storing, in the second row, the independence value corresponding to each level one node of the plurality of level one nodes.
5. The system of claim 4 , wherein generating the segment-discovery tree further includes:
for each level one node of the plurality of level one nodes:
determining, using the contingency table, a probability that an interaction characteristic corresponding to the level one node is independent from one or more other interaction characteristics, wherein an interaction characteristic is independent when a presence of the interaction characteristic in a user session is not dependent on a presence of another interaction characteristic in the user session;
determining whether the probability that the characteristic corresponding to the level one node is independent from one or more other interaction characteristics is below a threshold; and
removing the level one node of the plurality of level one nodes from the segment-discovery tree when the probability that the characteristic corresponding to the level one node is independent from one or more other interaction characteristics is below the threshold.
6. The system of claim 4 , wherein generating the segment-discovery tree further includes :
for each level two node of the plurality of level two nodes:
determining, using the contingency table, whether the level two node is to be expanded; and
generating, for each level two node that is to be expanded, one or more level three node, the one or more level three node corresponding to a combined interaction characteristic, the combined interaction characteristic including a corresponding interaction characteristic of a level two node and another level two node or a level one node.
7. A method comprising:
receiving a session dataset including, for each of one or user sessions, one or more interaction characteristics that characterize user interaction with a first version of a webpage during a session;
defining, from the session dataset a set of interaction characteristics;
receiving a definition of a target characteristic group, the definition identifying a first subset of the set of interaction characteristics;
identifying a second subset of the set of interaction characteristics, each characteristic in the second subset being absent from the first subset;
generating, using the second subset of the set of interaction characteristics, a segment-discovery tree by:
generating a plurality of level one nodes, each level one node of the plurality of level one nodes corresponding to an interaction characteristic of the second subset; and
generating a plurality of level two nodes, the plurality of level two nodes corresponding to a combined interaction characteristic, the combined interaction characteristic including a corresponding interaction characteristic of two or more level one nodes;
determining, for each level one node of the plurality of level one nodes in the segment-discovery tree, a probability that an interaction characteristic corresponding to the level one node correlates with the target characteristic group;
determining, for each level two node of the plurality of level two nodes in the segment-discovery tree, a probability that the combined interaction characteristic corresponding to the level two node correlate with the target characteristic group;
determining that a probability corresponding to each of a plurality of level one nodes and each of a plurality of level two nodes exceeds a threshold;
defining an interaction dataset that includes, for each level one node and each level two node with a corresponding probability that exceeds the threshold, the interaction characteristic corresponding to the level one node and the combined interaction characteristic corresponding to the level two node; generating an alternative version of the webpage based on the interaction dataset by modifying one or more nodes of a document object model corresponding to the webpage, the alternative version of the webpage including at least some content of a set of content that is different from the set of content rendered by the first version of the webpage; and
transmitting the alternative version of the webpage to one or more users that satisfy the interaction dataset within a particular geographical region, wherein the first version of the webpage is configured to be served to users that do not satisfy the interaction dataset.
8. The method of claim 7 , wherein each of the one or more interaction characteristics includes a set of attributes, and wherein the definition of the target characteristic group identifies a subset of the set of attributes for each interaction characteristic of the first subset of the set of interaction characteristics.
9. The method of claim 7 , wherein a the combined interaction characteristic represents the corresponding interaction characteristic of two or more level one nodes as a Boolean expression.
10. The method of claim 7 , wherein generating the segment-discover tree further includes:
generating a contingency table, the contingency table including a first row and second row;
determining, for each level one node of the plurality of level one nodes, a frequency value of an interaction characteristic corresponding to the level one node, the frequency value indicating a number of user sessions in which the interaction characteristic characterized a user session;
storing, in the first row, the frequency value corresponding to each level one node of the plurality of level one nodes;
determining, for each level one node of the plurality of level one nodes, an independence value of an interaction characteristic corresponding to the level one node, the independence value indicating a number of user sessions in which the interaction characteristic did not characterize a user session; and
storing, in the second row, the independence value corresponding to each level one node of the plurality of level one nodes.
11. The method of claim 10 , wherein generating the segment-discover tree further includes:
for each level one node of the plurality of level one nodes:
determining, using the contingency table, a probability that an interaction characteristic corresponding to the level one node is independent from one or more other interaction characteristics, wherein an interaction characteristic is independent when a presence of the interaction characteristic in a user session is not dependent on a presence of another interaction characteristic in the user session;
determining whether the probability that the characteristic corresponding to the level one node is independent from one or more other interaction characteristics is below a threshold; and
removing the level one node of the plurality of level one nodes from the segment-discovery tree when the probability that the characteristic corresponding to the level one node is independent from one or more other interaction characteristics is below the threshold.
12. The method of claim 10 , wherein generating the segment-discovery tree further includes :
for each level two node of the plurality of level two nodes:
determining, using the contingency table, whether the level two node is to be expanded; and
generating, for each level two node that is to be expanded, one or more level three node, the one or more level three node corresponding to a combined interaction characteristic, the combined interaction characteristic including a corresponding interaction characteristic of a level two node and another level two node or a level one node.
13. A non-transitory computer-readable storage media storing instructions that, when executed by one or more processors, causes the one or more processors to perform operations including:
receiving a session dataset including, for each of one or user sessions, one or more interaction characteristics that characterize user interaction with a first version of a webpage during a session;
defining, from the session dataset a set of interaction characteristics;
receiving a definition of a target characteristic group, the definition identifying a first subset of the set of interaction characteristics;
identifying a second subset of the set of interaction characteristics, each characteristic in the second subset being absent from the first subset;
generating, using the second subset of the set of interaction characteristics, a segment-discovery tree by:
generating a plurality of level one nodes, each level one node of the plurality of level one nodes corresponding to an interaction characteristic of the second subset; and
generating a plurality of level two nodes, the plurality of level two nodes corresponding to a combined interaction characteristic, the combined interaction characteristic including a corresponding interaction characteristic of two or more level one nodes;
determining, for each level one node of the plurality of level one nodes in the segment-discovery tree, a probability that an interaction characteristic corresponding to the level one node correlates with the target characteristic group;
determining, for each level two node of the plurality of level two nodes in the segment-discovery tree, a probability that the combined interaction characteristic corresponding to the level two node correlate with the target characteristic group;
determining that a probability corresponding to each of a plurality of level one nodes and each of a plurality of level two nodes exceeds a threshold;
defining an interaction dataset that includes, for each level one node and each level two node with a corresponding probability that exceeds the threshold, the interaction characteristic corresponding to the level one node and the combined interaction characteristic corresponding to the level two node;
generating an alternative version of the webpage based on the interaction dataset by modifying one or more nodes of a document object model corresponding to the webpage, the alternative version of the webpage including at least some content of a set of content that is different from the set of content rendered by the first version of the webpage; and
transmitting the alternative version of the webpage to one or more users that satisfy the interaction dataset within a particular geographical region, wherein the first version of the webpage is configured to be served to users that do not satisfy the interaction dataset.
14. The non-transitory computer-readable storage media of claim 13 , wherein each of the one or more interaction characteristics includes a set of attributes, and wherein the definition of the target characteristic group identifies a subset of the set of attributes for each interaction characteristic of the first subset of the set of interaction characteristics.
15. The non-transitory computer-readable storage media of claim 13 , wherein the combined interaction characteristic represents the corresponding interaction characteristic of two or more level one nodes as a Boolean expression.
16. The non-transitory computer-readable storage media of claim 13 , wherein generating the segment-discover tree further includes: generating a contingency table, the contingency table including a first row and second row;
determining, for each level one node of the plurality of level one nodes, a frequency value of an interaction characteristic corresponding to the level one node, the frequency value indicating a number of user sessions in which the interaction characteristic characterized a user session;
storing, in the first row, the frequency value corresponding to each level one node of the plurality of level one nodes;
determining, for each level one node of the plurality of level one nodes, an independence value of an interaction characteristic corresponding to the level one node, the independence value indicating a number of user sessions in which the interaction characteristic did not characterize a user session; and
storing, in the second row, the independence value corresponding to each level one node of the plurality of level one nodes.
17. The non-transitory computer-readable storage media of claim 16 , wherein generating the segment-discovery tree further includes:
for each level one node of the plurality of level one nodes:
determining, using the contingency table, a probability that an interaction characteristic corresponding to the level one node is independent from one or more other interaction characteristics, wherein an interaction characteristic is independent when a presence of the interaction characteristic in a user session is not dependent on a presence of another interaction characteristic in the user session;
determining whether the probability that the characteristic corresponding to the level one node is independent from one or more other interaction characteristics is below a threshold; and
removing the level one node of the plurality of level one nodes from the segment-discovery tree when the probability that the characteristic corresponding to the level one node is independent from one or more other interaction characteristics is below the threshold.
18. The non-transitory computer-readable storage media of claim 16 , wherein generating the segment-discovery tree further includes:
for each level two node of the plurality of level two nodes:
determining, using the contingency table, whether the level two node is to be expanded; and
generating, for each level two node that is to be expanded, one or more level three node, the one or more level three node corresponding to a combined interaction characteristic, the combined interaction characteristic including a corresponding interaction characteristic of a level two node and another level two node or a level one node.