Mitigating use of machine solvable HIPs

• US 8,885,931 B2
• Filed: 01/26/2011
• Issued: 11/11/2014
• Est. Priority Date: 01/26/2011
• Status: Expired due to Fees

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Srivastava, Kumar S.
• Primary Examiner(s): Park, Chan
• Assistant Examiner(s): Kholdebarin, Iman K
• Application Number:
  US13/013,968
• Publication Number:
  US 20120189194A1
• Time in Patent Office: 1,385 Days
• Field of Search: 382/161, 382/100, 726/21, 726/7
• US Class Current: 382/161
• CPC Class Codes: G06F 18/28 Determining representative reference patterns, e.g. by averaging or distorting; Generating dictionaries
• CPC Class Codes: G06F 21/316 by observing the pattern of computer usage, e.g. typical user behaviour
• CPC Class Codes: G06F 21/36 by graphic or iconic representation

Abstract

One or more techniques and/or systems are disclosed for mitigating machine solvable human interactive proofs (HIPs). A classifier is trained over a set of one or more training HIPs that have known characteristics for OCR solvability and HIP solving pattern from actual use. A HIP classification is determined for a HIP (such as from a HIP library used by a HIP generator) using the trained classifier. If the HIP is classified by the trained classifier as a merely human solvable classification, such that it may not be solved by a machine, the HIP can be identified for use in the HIP generation system. Otherwise, the HIP can be altered to (attempt to) be merely human solvable.

Claims

1. A computer-based method for mitigating machine solvable human interactive proofs (HIPs), comprising:

determining a HIP classification for a HIP using a trained classifier where the classifier is trained over a set of one or more training HIPs having known characteristics for;

HIP optical character recognition (OCR) solvability; and

a HIP solving pattern based upon a characteristic of how one or more characters are entered for a training HIP, the characteristic corresponding to at least one of back-tracking, deleting and replacing or deleting and re-entering; and

identifying the HIP for use in a HIP generation system if the HIP classification comprises a merely human solvable classification.

2. The method of claim 1, comprising subjecting the HIP to OCR using one or more OCR engines to determine OCR solvability for the HIP.

3. The method of claim 1, comprising collecting HIP solving pattern data from a HIP use log.

4. The method of claim 3, collecting HIP solving pattern data comprising collecting data indicating two or more of:

one or more mistakes made when using the training HIP;

a time duration for entering at least one character of the training HIP;

or

the characteristic of how one or more characters are entered.

5. The method of claim 1, comprising training the classifier over a set of one or more training HIPs using characteristics of HIPs solved by humans.

6. The method of claim 1, comprising training the classifier over a set of one or more training HIPs using characteristics of HIPs solved by machines.

7. The method of claim 1, comprising training the classifier over a set of one or more training HIPs using characteristics of HIPs solved by humans and machines.

8. The method of claim 1, comprising classifying the HIP as one of the following classifications:

merely machine solvable;

merely human solvable;

both human and machine solvable;

or

neither human nor machine solvable.

9. The method of claim 1, comprising classifying the HIP using a desired threshold for a classification confidence score for respective classifications.

10. The method of claim 1, comprising removing the HIP for use in the HIP generation system if the HIP comprises a merely machine solvable HIP.

11. The method of claim 1, comprising, if the HIP comprises a both human and machine solvable HIP:

identifying one or more machine solvable parts of the HIP; and

modifying the identified one or more machine solvable parts into a human solvable part to generate a modified HIP.

12. The method of claim 1, comprising:

determining a HIP classification for a modified HIP using the trained classifier, where the modified HIP comprises a both human and machine solvable HIP that is modified to be a merely human solvable HIP; and

using the modified HIP in the HIP generation system.

13. The method of claim 1, comprising:

running one or more HIPs in an existing HIP store through the trained classifier.

14. A system for mitigating machine solvable human interactive proofs (HIPs), comprising:

one or more processing units; and

memory comprising instructions that when executed by at least one of the one or more processing units, perform a method comprising;

determining a HIP classification for a HIP using a trained classifier where the classifier is trained over a set of one or more training HIPs having known characteristics for;

HIP optical character recognition (OCR) solvability; and

a HIP solving pattern based upon a characteristic of how one or more characters are entered for a training HIP, the characteristic corresponding to at least one of back-tracking, deleting and replacing or deleting and re-entering; and

identifying the HIP for use in a HIP generation system if the HIP classification comprises a merely human solvable classification.

15. The system of claim 14, the method comprising not identifying the HIP for use in the HIP generation system if the HIP classification comprises a machine solvable classification.

16. The system of claim 14, the method comprising modifying the HIP, if the HIP classification comprises a machine solvable classification.

17. The system of claim 14, the method comprising identifying one or more portions of the HIP that are machine solvable.

18. The system of claim 17, the method comprising altering at least some of the one or more portions of the HIP that are machine solvable to be merely human solvable.

19. The system of claim 14, the HIP solving pattern based upon an Internet-based address reputation for a location where one or more HIPs were solved.

20. A tangible computer readable storage device comprising instructions that when executed perform a method for mitigating machine solvable human interactive proofs (HIPs), comprising:

determining a HIP classification for a HIP using a trained classifier where the classifier is trained over a set of one or more training HIPs having known characteristics for;

HIP optical character recognition (OCR) solvability; and

a HIP solving pattern based upon a characteristic of how one or more characters are entered for a training HIP, the characteristic corresponding to at least one of back-tracking, deleting and replacing or deleting and re-entering; and

identifying the HIP for use in a HIP generation system if the HIP classification comprises a merely human solvable classification.