Network virtual user risk control method and system
Embodiments of the present application relate to a method of controlling user risk, a system for controlling user risk, and a computer program product for controlling user risk. A method is provided. The method includes retrieving association data of a first user and association data of a second user, the association data including multidimensional data, and data relating to each dimension identifying a user and serving as an association dimension, based on the association data, computing an association value between the first user and the second user for an association dimension, gathering the association value to obtain a degree of real association, and determining that the other user is malicious.
1. A network virtual user risk control method, comprising:
retrieving raw association data of a first virtual user and raw association data of a second virtual user, the raw association data comprising multidimensional data, and data relating to each dimension being capable of identifying a user and serving as an association dimension, wherein one of the first virtual user and the second virtual user is deemed to be a malicious user;
based on the raw association data of the first virtual user and the second virtual user, computing at least one association value between the first virtual user and the second virtual user for an association dimension;
gathering the at least one association value of the association dimension to obtain a degree of real association between the first virtual user and the second virtual user;
determining whether the degree of real association between the first virtual user and the second virtual user exceeds a predetermined threshold value; and
determining that the other of the first and second virtual users is a malicious user.
2. The method as described in claim 1 , wherein the computing of the at least one association value between the first virtual user and the second virtual user for the association dimension comprises:
for the association dimension, computing the at least one association value between the first virtual user and the second virtual user based on a quantity factor, a time factor, and a cascade factor.
3. The method as described in claim 2 , wherein the quantity factor employs a summation function.
4. The method as described in claim 3 , wherein the time factor and the cascade factor both employ an inverse function.
5. The method as described in claim 4 , wherein the at least one association value between the first virtual user and the second virtual user for the association dimension is computed using the following formula:
Σ x Σ level Σ t (1/t)*(1/level), wherein Σ corresponds to the summation function, x corresponds to the association quantity for an association dimension, level corresponds to a cascade level, and t corresponds to time.
6. The method as described in claim 1 , wherein the multidimensional data comprises an IP address, a cookie, a device fingerprint, a mobile phone number, a telephone number, a fax number, an email address, an address, a user name, or any combination thereof.
7. The method as described in claim 1 , wherein the retrieving of the raw association data of the first virtual user and the raw association data of the second virtual user comprises:
retrieving the raw association data of the first virtual user; and
based on the raw association data of the first virtual user, locating the second virtual user associated with the first user and the raw association data of the second virtual user.
8. The method as described in claim 7 , wherein the locating of the second virtual user associated with the first virtual user and the raw association data of the second virtual user comprises:
for each association dimension, finding dimension data used by the first virtual user based on an identifier of the first virtual user;
utilizing the found dimension data to find a user list corresponding to the dimension data; and
removing a duplicate user from the user list found for the each association dimension to identify a user in the user list as the second virtual user associated with the first virtual user.
9. The method as described in claim 1 , wherein the gathering of the at least one association value comprises summing the at least one association value for the association dimension.
10. A network virtual user risk control system, comprising:
at least one processor configured to:
retrieve raw association data of a first virtual user and raw association data of a second virtual user, the raw association data comprising multidimensional data and data relating to each dimension being capable of identifying a user and serving as an association dimension, wherein one of the first virtual user and the second virtual user is deemed to be a malicious user;
based on the raw association data of the first virtual user and the second virtual user, compute at least one association value between the first virtual user and the second virtual user for an association dimension;
gather the at least one association value for the association dimension to obtain a degree of real association between the first virtual user and the second virtual user;
determine whether the degree of real association between the first virtual user and the second virtual user exceeds a predetermined threshold value; and
determine that the other of the first and second virtual users is a malicious user; and
a memory coupled to the at least one processor and configured to provide the at least one processor with instructions.
11. The system as described in claim 10 , wherein the computing of the at least one is association value between the first virtual user and the second virtual user for the association dimension comprises:
for the association dimension, computing the at least one association value between the first virtual user and the second virtual user based on a quantity factor, a time factor and a cascade factor.
12. The system as described in claim 11 , wherein:
the quantity factor employs a summation function;
the time factor and the cascade factor both employ an inverse function; and
the at least one association value between the first virtual user and the second virtual user is computed using the following formula:
Σ x Σ level Σ t (1/t)*(1/level), wherein Σ corresponds to the summation function, x corresponds to the association quantity under an association dimension, level corresponds to a cascade level, and t corresponds to time.
13. The system as described in claim 10 , wherein the multidimensional data comprises an IP address, a cookie, a device fingerprint, a mobile telephone number, a telephone number, a fax number, an email address, an address, a user name, or any combination thereof.
14. A system as described in claim 10 , wherein the retrieving of the raw association data of the first virtual user and the raw association data of the second virtual user comprises:
retrieving the raw association data of the first virtual user; and
based on the raw association data of the first virtual user, locating the second virtual user associated with the first user and the raw association data of the second virtual user.
15. The system as described in claim 14 , wherein the locating of the second virtual user associated with the first virtual user and the raw association data of the second virtual user comprises:
for each association dimension, finding dimension data used by the first virtual user based on an identifier of the first virtual user;
utilizing the found dimension data to find a user list corresponding to the dimension data; and
removing a duplicate user from the user list found for the each association dimension to identify a user in the user list as the second virtual user associated with the first virtual user.
16. The system as described in claim 10 , wherein the gathering of the at least one association value comprises summing the at least one association value for the association dimension.
17. A computer program product for controlling network virtual user risk, the computer program product being embodied in a non-transitory computer readable storage medium and comprising computer instructions for:
retrieving raw association data of a first virtual user and raw association data of a second virtual user, the raw association data comprising multidimensional data, and data relating to each dimension being capable of identifying a user and serving as an association dimension, wherein one of the first virtual user and the second virtual user is deemed to be a malicious user;
based on the raw association data of the first virtual user and the second virtual user, computing at least one association value between the first virtual user and the second virtual user for an association dimension;
gathering the at least one association value of the association dimension to obtain a degree of real association between the first virtual user and the second virtual user;
determining whether the degree of real association between the first virtual user and the second virtual user exceeds a predetermined threshold value; and
determining that the other of the first and second virtual users is a malicious user.