Creating tests to identify fraudulent users

1. A method for generating a test to identify fraudulent users of an online group, comprising:

obtaining, for a control group, information associated with receptiveness to being matched with someone who has children, wherein the control group is representative of non-fraudulent users in the online group;

obtaining, for a general group, information associated with receptiveness to being matched with someone who has children, wherein the general group is representative of the online group as a whole;

obtaining, for a fraudulent group, information associated with receptiveness to being matched with someone who has children, wherein the fraudulent group is representative of fraudulent users in the online group; and

using a processor to generate a test that identifies fraudulent users of the online group, wherein:

the test includes (1) an equation which generates a fraud score using one or more pieces of user information as input and (2) a threshold against which the fraud score is compared; and

generating the test includes:

determining the equation, including by comparing the information associated with receptiveness to being matched with someone who has children for the control group with the information associated with receptiveness to being matched with someone who has children for the fraudulent group; and

determining the threshold, including by applying the equation to the information associated with receptiveness to being matched with someone who has children for the general group in order to obtain fraud scores for each member of the general group.

2. The method of claim 1 , wherein the online group is associated with dating, and the user demographic related information is based on user input to a questionnaire.

3. The method of claim 1 , wherein the fraudulent group includes at least one known fraudulent user.

4. The method of claim 1 , wherein the fraudulent group includes at least one suspected fraudulent user.

5. The method of claim 1 , wherein the control group includes at least one known legitimate user.

6. The method of claim 1 , wherein the control group only includes known legitimate users.

7. The method of claim 1 , wherein generating a test includes performing statistical analysis on the data associated with the control group and the data associated with the fraudulent group.

8. The method of claim 1 , wherein determining the equation further includes:

analyzing the data associated with the control group and the data associated with the fraudulent group;

determining based at least in part on the analysis one or more variables to be included in the equation; and

determining, for one of said one or more variables, an amount to add or subtract from the fraud score in the event said variable is a certain value or within a certain range of values.

9. The method of claim 1 , wherein determining the threshold further includes comparing an amount of false positives against an amount of properly identified fraudulent users for a potential threshold value.

10. A system for generating a test to identify fraudulent users of an online group, comprising:

a processor; and

a memory coupled with the processor, wherein the memory is configured to provide the processor with instructions which when executed cause the processor to:

obtain, for a control group, information associated with receptiveness to being matched with someone who has children, wherein the control group is representative of non-fraudulent users in the online group;

obtain, for a general group, information associated with receptiveness to being matched with someone who has children, wherein the general group is representative of the online group as a whole;

obtain, for a fraudulent group, information associated with receptiveness to being matched with someone who has children, wherein the fraudulent group is representative of fraudulent users in the online group; and

generate a test that identifies fraudulent users of the online group, wherein:

the test includes (1) an equation which generates a fraud score using one or more pieces of user information as input and (2) a threshold against which the fraud score is compared; and

generating the test includes:

determining the equation, including by comparing the information associated with receptiveness to being matched with someone who has children for the control group with the information associated with receptiveness to being matched with someone who has children for the fraudulent group; and

determining the threshold, including by applying the equation to the information associated with receptiveness to being matched with someone who has children for the general group in order to obtain fraud scores for each member of the general group.

11. The system of claim 10 , wherein generating the test includes performing statistical analysis on the data associated with the control group and the data associated with the fraudulent group.

12. A computer program product for generating a test to identify fraudulent users of an online group, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:

obtaining, for a control group, information associated with receptiveness to being matched with someone who has children, wherein the control group is representative of non-fraudulent users in the online group;

obtaining, for a general group, information associated with receptiveness to being matched with someone who has children, wherein the general group is representative of the online group as a whole;

obtaining, for a fraudulent group, information associated with receptiveness to being matched with someone who has children, wherein the fraudulent group is representative of fraudulent users in the online group; and

generating a test that identifies fraudulent users of the online group, wherein:

the test includes (1) an equation which generates a fraud score using one or more pieces of user information as input and (2) a threshold against which the fraud score is compared; and

generating the test includes:

determining the equation, including by comparing the information associated with receptiveness to being matched with someone who has children for the control group with the information associated with receptiveness to being matched with someone who has children for the fraudulent group; and

determining the threshold, including by applying the equation to the information associated with receptiveness to being matched with someone who has children for the general group in order to obtain fraud scores for each member of the general group.

13. The computer program product of claim 12 , wherein the computer instructions for generating a test include computer instructions for performing statistical analysis on the data associated with the control group and the data associated with the fraudulent group.

14. The method of claim 1 , wherein (1) the members of the control group do not overlap with the members of the general group and (2) each member of the general group is either a known fraudulent user or a known non-fraudulent user.

15. The system of claim 10 , wherein (1) the members of the control group do not overlap with the members of the general group and (2) each member of the general group is either a known fraudulent user or a known non-fraudulent user.

16. The computer program product of claim 12 , wherein (1) the members of the control group do not overlap with the members of the general group and (2) each member of the general group is either a known fraudulent user or a known non-fraudulent user.

17. The method of claim 1 , wherein:

determining the equation further includes comparing information associated with receptiveness to having (more) children in the future for the control group with information associated with receptiveness to having (more) children in the future for the fraudulent group; and

determining the threshold further includes applying the equation to information associated with receptiveness to having (more) children in the future for the general group.

18. The system of claim 10 , wherein:

the instructions for determining the equation further include instructions for comparing information associated with receptiveness to having (more) children in the future for the control group with information associated with receptiveness to having (more) children in the future for the fraudulent group; and

the instructions for determining the threshold further include instructions for applying the equation to information associated with receptiveness to having (more) children in the future for the general group.

19. The computer program product of claim 12 , wherein:

the computer instructions for determining the equation further include computer instructions for comparing information associated with receptiveness to having (more) children in the future for the control group with information associated with receptiveness to having (more) children in the future for the fraudulent group; and

the computer instructions for determining the threshold further include computer instructions for applying the equation to information associated with receptiveness to having (more) children in the future for the general group.

20. The method of claim 1 , wherein:

determining the equation further includes comparing information associated with receptiveness to a potential match outside of a user's metropolitan area with information associated with receptiveness to a potential match outside of a user's metropolitan area for the fraudulent group; and

determining the threshold further includes applying the equation to information associated with receptiveness to a potential match outside of a user's metropolitan area for the general group.

21. The system of claim 10 , wherein:

the instructions for determining the equation further include instructions for comparing information associated with receptiveness to a potential match outside of a user's metropolitan area with information associated with receptiveness to a potential match outside of a user's metropolitan area for the fraudulent group; and

the instructions for determining the threshold further include instructions for applying the equation to information associated with receptiveness to a potential match outside of a user's metropolitan area for the general group.

22. The computer program product of claim 12 , wherein:

the computer instructions determining the equation further include computer instructions for comparing information associated with receptiveness to a potential match outside of a user's metropolitan area with information associated with receptiveness to a potential match outside of a user's metropolitan area for the fraudulent group; and

the computer instructions for determining the threshold further include computer instructions for applying the equation to information associated with receptiveness to a potential match outside of a user's metropolitan area for the general group.