Detection of user behavior using time series modeling

• US 9,152,490 B2
• Filed: 10/25/2013
• Issued: 10/06/2015
• Est. Priority Date: 04/02/2013
• Status: Expired due to Fees

• Current Assignee: Western Digital Technologies Incorporated (Western Digital Corporation)
• Original Assignee: Western Digital Technologies Incorporated (Western Digital Corporation)
• Inventors: Ahmadi-Ardakani, Arya
• Primary Examiner(s): SHIN, CHRISTOPHER B
• Application Number:
  US14/063,982
• Publication Number:
  US 20140298117A1
• Time in Patent Office: 711 Days
• Field of Search: 710/62, 713/300, 711/137, 711/154
• US Class Current: 1/1
• CPC Class Codes: G06F 11/0757 by exceeding a time limit, i.e. time-out, e.g. watchdogs
• CPC Class Codes: G06F 11/3438 monitoring of user actions tracking the activity of the user H04L67/535
• CPC Class Codes: G06F 11/3452 Performance evaluation by statistical analysis
• CPC Class Codes: G06F 11/3485 for I/O devices

Abstract

The embodiments provide a way to predict when a storage device will be accessed. In order to enhance performance, the storage device may proactively prepare for the access operation, and thus, minimize the access-time response of the storage device. The user behavior is recorded over time and collected into a dataset. In one embodiment, the intervals between the data points in the dataset are calculated and arranged into a matrix. Patterns in the matrix are recognized and used to recognize the next likely access by the user. The storage device may then take various actions, such as drive spin up, in anticipation of the next predicted access to minimize access-time response.

Claims

1. A method of predicting future access of a storage device, said method comprising:

determining respective time intervals between each access request to any subsequent access request of the storage device;

arranging the respective time intervals in a matrix;

identifying a pattern in the respective time intervals, including determining first set of intervals occurring in a column of the matrix that correspond to a second set of intervals occurring in a row of the matrix;

predicting a next access request based on the pattern in the respective time intervals; and

proactively activating the storage device based on the predicted next access request.

2. The method of claim 1, wherein proactively activating the storage device comprises proactively spinning up a hard disk drive before the predicted next access request based on a spin up time of the hard disk.

3. The method of claim 1, wherein proactively activating the storage device comprises proactively activating the storage device before the predicted next access request based on the respective intervals and a tolerance.

4. The method of claim 1, further comprising determining an error tolerance between the respective intervals.

5. The method of claim 4, further comprising grouping a plurality of access requests that occur within the error tolerance.

6. The method of claim 1, wherein each column of the matrix indicates time intervals from all other past user access requests and wherein each row of the matrix indicates time intervals to subsequent user access requests.

7. A method of managing an electronic device based on predicting user behavior, said method comprising:

recording a series of user access requests of the electronic device over a period of time;

determining respective time intervals between the series of user access requests;

arranging the respective time intervals into a matrix wherein each column of the matrix indicates time intervals from all other past user access requests and wherein each row indicates time intervals to subsequent user access requests;

identifying a pattern of user access requests based on modeling the series of user access requests based on matching a series of time intervals of the user access requests by identifying a sequence of intervals that occur in both a row and column from a selected element in the matrix;

determining a predicted time of a next user access request based on the identified pattern; and

initiating an action on the electronic device based on the predicted time of the next user access request.

8. The method of claim 7, wherein identifying the pattern of user access requests comprises identifying a sequence of intervals that occur in both a row and column from a selected element in the matrix within a tolerance.

9. The method of claim 7, wherein initiating the action on the electronic device comprises initiating spin up of a hard disk drive.

10. The method of claim 7, wherein initiating the action on the electronic device comprises initiating a spin down of a hard disk drive.

11. The method of claim 7, wherein initiating the action on the electronic device comprises transitioning a storage drive from a standby state to an active state.

12. The method of claim 7, wherein initiating the action on the electronic device comprises transitioning a storage drive from an active state to a standby state.

13. A storage device configured to anticipate user behavior based on modeling a time series of user access requests, said storage device comprising:

at least one storage medium;

a controller comprising;

an event recorder configured to record a time series of user access requests for data from the at least one storage medium and to store the time series in a matrix, wherein each column of the matrix indicates time intervals from all other past user access requests and wherein each row indicates time intervals to subsequent user access requests;

a power manager configured to manage power consumption of the at least one storage medium based on an active state and a standby state;

a time series modeler configured to detect patterns of user behavior based on intervals in the time series between user access requests, wherein the time series modeler is configured to detect patterns of user behavior based on matching sequences of intervals occurring both a row and a column of the matrix; and

a predictor configured to predict a next user access request based on the patterns of user behavior and send a signal to the power manager to transition the storage device from a first power state to a second power state in anticipation of the predicted next user access request.

14. The storage device of claim 13, wherein predictor is configured to send a signal to transition the at least on storage medium from a standby state to an active state in anticipation of the predicted next user access request.

15. The storage device of claim 13, wherein predictor is configured to send a signal to transition the at least on storage medium from an active state to a standby state in anticipation of the predicted next user access request.

16. A device configured to predict user behavior, said device comprising:

a storage storing information subject to user access requests; and

a processor configured by executable program code to record a series of user access requests of the device over a period of time, determine respective intervals between the series of user access requests, identify a pattern of user access requests based on modeling the series of user access requests based on matching a series of time intervals of the user access requests, determine a predicted time of a next user access request based on the identified pattern, and initiate an action on the device based on the predicted time of the next user access request,

wherein the processor is configured to arrange the respective intervals into a matrix stored in a memory coupled to the processor, wherein each column of the matrix indicates intervals from all other past user access requests and wherein each row indicates intervals to subsequent user access requests; and

wherein the processor is configured to identify access requests based on identifying a sequence of intervals that occur in both a row and column from a selected element in the matrix.

17. The device of claim 16, wherein the storage comprises a hard disk drive and the processor is configured to initiate spin up of a hard disk drive.

18. The device of claim 16, wherein the storage comprises a hard disk drive and the processor is configured to initiate spin down of a hard disk drive.

19. The device of claim 16, wherein the storage comprises a hard disk drive and the processor is configured to transition the hard disk drive from a standby state to an active state.

20. The device of claim 16, wherein the storage comprises a hard disk drive and the processor is configured to transition the hard disk drive from an active state to a standby state.

21. The device of claim 16, wherein the processor is configured to initiate caching content responsive to the pattern of user access.

22. The device of claim 16, wherein the processor is configured to connect to a network responsive to the pattern of user access.