MODELING, DETECTING, AND PREDICTING USER BEHAVIOR WITH HIDDEN MARKOV MODELS

• US 20090164395A1
• Filed: 12/21/2007
• Published: 06/25/2009
• Est. Priority Date: 12/21/2007
• Status: Active Grant

• Current Assignee: R2 Solutions LLC (Acacia Research Corporation)
• Original Assignee: Yahoo! Inc. (Apollo Global Management, Inc.)
• Inventors: Heck, Larry P.
• Granted Patent: US 7,941,383 B2
• Time in Patent Office: Days
• US Class Current: 706/16
• CPC Class Codes: G06F 16/954 Navigation, e.g. using categorised browsing

Abstract

Mechanisms model, detect, and predict user behavior as a user navigates the Web. In one embodiment, mechanisms model user behavior using predictive models, such as discrete Markov processes, where the user'"'"'s behavior transitions between a finite number of states. The user'"'"'s behavior state may not be directly observable (e.g., a user does not proactively indicate what behavior state he is in). Thus, the behavior state of a user is usually only indirectly observable. Mechanisms use predictive models, such as hidden Markov models, to predict the transitions in the user'"'"'s behavior states.

Claims

1. A method for providing targeted data to a set of users, the method comprising:

maintaining state transition information that indicates state-to-state transitions made by the set of users;

updating the state transition information to record that a first user navigated from a first state to a second state;

in response to a request by a second user to navigate to the first state, inspecting the state transition information to determine a next state associated with said first state;

wherein said first state and said next state include references to one or more web resources;

selecting at least one web resource from said next state to be presented to the second user; and

presenting to the second user controls for navigating to said at least one web resource.

2. The method of claim 1, wherein the state-to-state transitions include captured state-to-state transition information.

3. The method of claim 1, wherein said at least one web resource is a web page.

4. The method of claim 1, wherein inspecting the state transition information to determine a next state associated with said first state comprises executing a predictive model to select the next state.

5. The method of claim 4, wherein the predictive model is a hidden Markov model.

6. The method of claim 1, wherein the state-to-state transitions are specific to a user in the set of users.

7. The method of claim 1, wherein presenting to the second user controls for navigating to said at least one web resource includes displaying the at least one web resource in a web browser toolbar.

8. The method of claim 1, wherein presenting to the second user controls for navigating to said at least one web resource includes displaying at least one web resource in a web page.

9. The method of claim 1, wherein updating the state transition information to record that a first user navigated from a first page to a second page is performed anonymously.

10. The method of claim 1, wherein updating the state transition information to record that a first user navigated from a first state to a second state includes receiving an indication of consent from the first user before state transition information is recorded.

11. The method of claim 10, wherein receiving the indication of consent may be selectively revoked by the first user.

12. The method of claim 1, wherein inspecting the state transition information to determine a next state associated with said first state includes evaluating the state-to-state transitions to find a common pattern of state access among the set of users.

13. The method of claim 1, wherein the second state corresponds to the next state.

14. A machine-readable storage medium carrying instructions for providing targeted data to a set of users, wherein the instructions, when executed by one or more processors, cause the one or more processors to perform:

maintaining state transition information that indicates state-to-state transitions made by the set of users;

updating the state transition information to record that a first user navigated from a first state to a second state;

in response to a request by a second user to navigate to the first state, inspecting the state transition information to determine a next state associated with said first state;

wherein said first state and said next state include references to one or more web resources;

selecting at least one web resource from said next state to be presented to the second user; and

presenting to the second user controls for navigating to said at least one web resource.

15. The machine-readable storage medium of claim 14, wherein the state-to-state transitions include captured state-to-state transition information.

16. The machine-readable storage medium of claim 14, wherein said at least one web resource is a web page.

17. The machine-readable storage medium of claim 14, wherein inspecting the state transition information to determine a next state associated with said first state comprises executing a predictive model to select the next state.

18. The machine-readable storage medium of claim 17, wherein the predictive model is a hidden Markov model.

19. The machine-readable storage medium of claim 14, wherein the state-to-state transitions are specific to a user in the set of users.

20. The machine-readable storage medium of claim 14, wherein presenting to the second user controls for navigating to said at least one web resource includes displaying the at least one web resource in a web browser toolbar.

21. The machine-readable storage medium of claim 14, wherein presenting to the second user controls for navigating to said at least one web resource includes displaying at least one web resource in a web page.

22. The machine-readable storage medium of claim 14, wherein updating the state transition information to record that a first user navigated from a first page to a second page is performed anonymously.

23. The machine-readable storage medium of claim 14, wherein updating the state transition information to record that a first user navigated from a first state to a second state includes receiving an indication of consent from the first user before state transition information is recorded.

24. The machine-readable storage medium of claim 23, wherein receiving the indication of consent may be selectively revoked by the first user.

25. The machine-readable storage medium of claim 14, wherein inspecting the state transition information to determine a next state associated with said first state includes evaluating the state-to-state transitions to find a common pattern of state access among the set of users.

26. The machine-readable storage medium of claim 14, wherein the second state corresponds to the next state.