Automatic user identification from button presses recorded in a feature vector

• US 8,630,963 B2
• Filed: 07/01/2011
• Issued: 01/14/2014
• Est. Priority Date: 07/01/2011
• Status: Expired due to Fees

• Current Assignee: Intel Corporation
• Original Assignee: Intel Corporation
• Inventors: Phielipp, Mariano J., Kveton, Branislav, Galn, Magdiel F.
• Primary Examiner(s): Gaffin, Jeffrey A
• Assistant Examiner(s): Bejcek, II, Robert H
• Application Number:
  US13/175,603
• Publication Number:
  US 20130006898A1
• Time in Patent Office: 928 Days
• Field of Search: 706/12
• US Class Current: 706/12
• CPC Class Codes: G06F 21/316 by observing the pattern of computer usage, e.g. typical user behaviour

Abstract

Methods, systems and computer program products to allow a user to identify himself to a content provider, without having to explicitly perform a log in process or other identification and authentication process. By manipulating a user device such as a remote control, a profile of the user may be constructed, where the profile includes a representation of how the individual user typically manipulates the device. The profile includes a feature vector that is a function of the number of times that individual buttons are pressed. The construction of the feature vector may be viewed as a training or learning phase. Once the profile and feature vector are constructed, the user'"'"'s interaction with the device in a subsequent session may be captured and compared with the profile. This may allow identification of the user, in turn allowing content to be tailored in a manner specific to this user.

Claims

1. A method, comprising:

receiving a user identifier at a user device;

training the device by constructing a profile of the user associated with the user identifier, wherein the profile comprises a feature vector that is a function of the user'"'"'s pressing of buttons on the device, the function represented by

button_isf_vec[i]=log((float)(last ₁₃ session+2)/(button_cum_vec[i]+1))

where

button _isf_vec is the feature vector,

i is the index of the feature vector, such that 0 ≦

i <

the number of buttons on the device,

last_session is an index referring to the number of sessions that the user has spent interacting with the device, and

button_cum vec[i] is the number of times that the user has pressed the ith button of the device,

where the function incorporates an inverse frequency of the presses of each button;

associating the profile with the user identifier; and

subsequent to said training, predicting the user identifier, given one or more instances of the user'"'"'s pressing of buttons on the device, given the profile, and given the association of the profile with the user identifier.

2. The method of claim 1, wherein the device comprises a remote control.

3. The method of claim 1, wherein said predicting identifies the user identifier from among a plurality of other identifiers associated with a respective plurality of other users.

4. The method of claim 1, wherein said training comprises operation of a support vector machine.

5. The method of claim 4, wherein said operation comprises on-line convex programming.

6. A system, comprising:

a processor; and

a memory in communication with said processor, wherein said memory stores a plurality of processing instructions configured to direct said processor to

receive a user identifier at a user device;

train the device by constructing a profile of the user associated with the user identifier, wherein the profile comprises a feature vector that is a function of the user'"'"'s pressing of buttons on the device, the function represented by

button_isf_vec[i]=log((float)(last₁₃ session+2)/(button_cum _vec[i]+1))

where

button_isf_vec is the feature vector,

i is the index of the feature vector, such that 0 ≦

i <

the number of buttons on the device,

last_session is an index referring to the number of sessions that the user has spent interacting with the device, and

button_cum_vec[i] is the number of times that the user has pressed the ith button of the device,

where the function incorporates an inverse frequency of the presses of each button;

associate the profile with the user identifier; and

subsequent to said training, predict the user identifier, given one or more instances of the user'"'"'s pressing of buttons on the device, given the profile, and given the association of the profile with the user identifier.

7. The system of claim 6, wherein the device comprises a remote control.

8. The system of claim 6, wherein said processing instructions configured to direct said processor to predict comprise:

processing instructions configured to direct said processor to identify the user identifier from among a plurality of other identifiers associated with a respective plurality of other users.

9. The system of claim 6, wherein said processing instructions configured to direct said processor to train comprise:

processing instructions configured to direct said processor to operate a support vector machine.

10. The system of claim 9, wherein said processing instructions configured to direct said processor to operate the support vector machine comprise:

processing instructions configured to direct said processor to perform on-line convex programming.

11. A computer program product including a non-transitory computer readable medium having computer program logic stored therein, the computer program logic comprising:

logic to cause a processor to receive a user identifier at a user device;

logic to cause the processor to train the device by constructing a profile of the user associated with the user identifier, wherein the profile comprises a feature vector that is a function of the user'"'"'s pressing of buttons on the device, the function represented by

button_isf_vec[i]=log((float)(last₁₃ session+2)/(button_cum_vec[i]+1))

where

button_isf_vec is the feature vector,

i is the index of the feature vector, such that 0≦

i <

the number of buttons on the device,

last₁₃session is an index referring to the number of sessions that the user has spent interacting with the device, and

button_cum_vec[i] is the number of times that the user has pressed the ith button of the device,

where the function incorporates an inverse frequency of the presses of each button;

logic to cause the processor to associate the profile with the user identifier; and

logic to cause the processor to predict the user identifier, subsequent to said training, given one or more instances of the user'"'"'s pressing of buttons on the device, given the profile, and given the association of the profile with the user identifier.

12. The computer program product of claim 11, wherein the device comprises a remote control.

13. The computer program product of claim 11, wherein said logic to cause the processor to predict comprises:

logic to cause the processor to identify the user identifier from among a plurality of other identifiers associated with a respective plurality of other users.

14. The computer program product of claim 11, wherein said logic to cause the processor to train comprises:

logic to cause the processor to operate a support vector machine.

15. The computer program product of claim 14, wherein said logic to cause the processor to operate the support vector machine comprises:

logic to cause the processor to perform on-line convex programming.