User biometric pattern learning and prediction

• US 9,336,373 B2
• Filed: 04/15/2014
• Issued: 05/10/2016
• Est. Priority Date: 04/15/2014
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Inventors: Ramachandran, Vidhyaprakash, Jasti, Jay K., Su, Yu
• Primary Examiner(s): Pham, Luu
• Assistant Examiner(s): JACKSON, JENISE E
• Application Number:
  US14/253,068
• Publication Number:
  US 20150294097A1
• Time in Patent Office: 756 Days
• Field of Search: 726/28
• US Class Current: 1/1
• CPC Class Codes: G06F 21/316 by observing the pattern of computer usage, e.g. typical user behaviour
• CPC Class Codes: G06F 21/32 using biometric data, e.g. fingerprints, iris scans or voiceprints
• CPC Class Codes: G06F 3/04883 for inputting data by handwriting, e.g. gesture or text

Abstract

A user device collects timing data that indicates screen touch timing behavior during multiple touch events associated with at least one of a hold time for a particular button or a transition time between two particular buttons. The user device also collects force data indicating screen touch force behavior during the multiple touch events. The user device identifies a user biometric pattern for the touch event based on the timing data and the force data, and stores the user biometric pattern.

Claims

1. A method, comprising:

collecting, by a user device, timing data that indicates screen touch timing behavior during multiple touch events for a sequence of keys, wherein the timing data for each of the multiple touch events is associated with at least one of a hold time for a particular button or a transition time between two particular buttons;

collecting, by the user device, force data indicating screen touch force behavior during the multiple touch events, wherein the force data for each of the multiple touch events includes acceleration values and orientation values for the user device, wherein the acceleration values includes three directional acceleration values that are captured independently and the orientation values include three directional orientation values that are captured independently;

identifying, by the user device, a user biometric pattern for the multiple touch events, wherein the user biometric pattern is based on the timing data and the force data and includes, for each of the multiple touch events, a time interval, the acceleration values over the time interval, a first order acceleration derivative, a second order acceleration derivative, the orientation values over the time interval, a first order orientation derivative, and a second order orientation derivative;

storing, by the user device, the user biometric pattern associated with the multiple touch events;

presenting, by the user device, a challenge to solicit new touch events for the sequence of keys;

obtaining, by the user device, new timing data for the new touch events;

obtaining, by the user device, new force data for the new touch events;

comparing, by the user device, the new timing data and the new force data to the user biometric pattern;

determining, by the user device and based on the comparing, if the new timing data and the new force data correspond to the user biometric pattern; and

preventing access to one of the user device or an application stored on the user device when the new timing data and the new force data do not correspond to the user biometric pattern.

2. The method of claim 1, wherein the identifying further comprises:

generating a directed graph corresponding to a particular keypad arrangement, and identifying clusters for each vertex and each edge in the directed graph.

3. The method of claim 1, wherein the determining includes identifying an overall confidence value based on the new timing data and the new force data.

4. The method of claim 3, wherein identifying the overall confidence value includes:

identifying a confidence value for each of the multiple touch events, and

averaging the confidence values to generate the overall confidence value.

5. The method of claim 1, wherein the timing data for each of the multiple touch events includes:

a timestamp of a down click and a timestamp of an up click for a single button, or

a timestamp of an up click for a first button and a timestamp of a down click for a second button.

6. The method of claim 1, wherein the identifying the user biometric pattern includes, for each of the multiple touch events, constructing a cluster of the timing data and constructing a cluster of the force data.

7. The method of claim 6, wherein the identifying the user biometric pattern further includes identifying a centroid and a radius for each cluster of the timing data and each cluster of the force data.

8. The method of claim 1, wherein

presenting the challenge to solicit new touch events includes:

soliciting one of a password, a secret code, or an input pattern;

or

display an alphanumeric sequence that has to be entered by the user.

9. The method of claim 8, wherein determining if the new timing data and the new force data correspond to the user biometric pattern further comprises:

comparing content of the new touch events with content of the sequence of keys.

10. A device, comprising:

a memory configured to store a plurality of instructions; and

a processor configured to execute instructions in the memory to;

collect timing data that indicates screen touch timing behavior during multiple touch events for a sequence of keys, wherein the timing data for each of the multiple touch events is associated with at least one of a hold time for a particular button or a transition time between two particular buttons,

collect force data indicating screen touch force behavior during the multiple touch events, wherein the force data for each of the multiple touch events includes acceleration values and orientation values for the device, wherein the acceleration values includes three directional acceleration values that are captured independently and the orientation values include three directional orientation values that are captured independently,

identify a user biometric pattern for the multiple touch events, wherein the user biometric pattern is based on the timing data and the force data and includes, for each of the multiple touch events, a time interval, the acceleration values over the time interval, a first order acceleration derivative, a second order acceleration derivative, the orientation values over the time interval, a first order orientation derivative, and a second order orientation derivative,

store the user biometric pattern associated with the multiple touch events,

present a challenge to solicit new touch events for the sequence of keys,

obtain new timing data for the new touch events,

obtain new force data for the new touch events,

compare the new timing data and the new force data to the user biometric pattern,

determine, based on the comparing, if the new timing data and the new force data correspond to the user biometric pattern, and

prevent access to one of the device or an application stored on the device when the new timing data and the new force data do not correspond to the user biometric pattern.

11. The device of claim 10, wherein, when identifying the user biometric pattern, the processor is further configured to:

generate a directed graph corresponding to a particular keypad arrangement, and identify clusters for each vertex and each edge in the directed graph.

12. The device of claim 10, wherein when determining if the new timing data and the new force data correspond to the user biometric pattern, the processor is further configured to:

identify an overall confidence value based on the new timing data and the new force data.

13. The device of claim 12, wherein when identifying the overall confidence value, the processor is further configured to:

identify a confidence value for each of the multiple touch events, and

average the confidence values.

14. The device of claim 10, wherein the timing data for each of the multiple touch events includes at least one of:

a timestamp of a down click and a timestamp of an up click for a single button, or

a timestamp of an up click for a first button and a timestamp of a down click for a second button.

15. The device of claim 10, wherein, when identifying the user biometric pattern, the processor is further configured to:

construct, for each of the multiple touch events, a timing cluster of the timing data that identifies a centroid and a radius of the timing cluster, and

construct, for each of the multiple touch events, a force cluster of the force data that identifies a centroid and a radius of the force cluster.

16. A non-transitory computer-readable medium storing instructions executable by a computational device to:

collect timing data that indicates screen touch timing behavior during multiple touch events for a sequence of keys, wherein the timing data for each of the multiple touch events is associated with at least one of a hold time for a particular button or a transition time between two particular buttons,

collect force data indicating screen touch force behavior during the multiple touch events, wherein the force data for each of the multiple touch events includes acceleration values and orientation values for the computational device, wherein the acceleration values includes three directional acceleration values that are captured independently and the orientation values include three directional orientation values that are captured independently,

identify a user biometric pattern for the multiple touch events, wherein the user biometric pattern is based on the timing data and the force data and includes, for each of the multiple touch events, a time interval, the acceleration values over the time interval, a first order acceleration derivative, a second order acceleration derivative, the orientation values over the time interval, a first order orientation derivative, and a second order orientation derivative,

store the user biometric pattern associated with the multiple touch events,

present a challenge to solicit new touch events for the sequence of keys,

obtain new timing data for the new touch events,

obtain new force data for the new touch events,

compare the new timing data and the new force data to the user biometric pattern,

determine, based on the comparing, if the new timing data and the new force data correspond to the user biometric pattern, and

prevent access to one of the computational device or an application stored on the computational device when the new timing data and the new force data do not correspond to the user biometric pattern.

17. The non-transitory computer-readable medium of claim 16, wherein the instructions to identify the user biometric pattern further comprise instructions to:

generate a directed graph corresponding to a particular keypad arrangement, and identifying clusters for each vertex and each edge in the directed graph.

18. The non-transitory computer-readable medium of claim 16, wherein the instructions to identify the user biometric pattern further comprise instructions to:

for each of the multiple touch events, construct a cluster of the timing data and construct a cluster of the force data, and

identify a centroid and a radius for each cluster of the timing data and each cluster of the force data.

19. The non-transitory computer-readable medium of claim 16, wherein the instructions to determine if the new timing data and the new force data correspond to the user biometric pattern further comprise instructions to:

identify an overall confidence value based on the new timing data and the new force data.

20. The non-transitory computer-readable medium of claim 19, wherein the instructions to identify the overall confidence value further comprise instructions to:

identify a confidence value for each of the multiple touch events, and

average the confidence values.