Anomaly detection for online endorsement event
View Patent ↗A system for protecting endorsement events against anomalous behaviors is disclosed. The system includes a data collecting module, an activity recording module, an aggregation module, a detection module and an action module. The data collecting module collects event data describing one or more endorsement events. The activity recording module records activity data associated with the one or more endorsement events. The aggregation module aggregates the activity data. The detection module detects, based at least in part on the aggregation of the activity data, one or more anomalies associated with the one or more endorsement events. The action module takes one or more protection actions responsive to the detection of the one or more anomalies. The present disclosure also includes a method for protecting endorsement events against anomalous behaviors.
1. A method comprising:
retrieving, using one or more computing devices, event data describing one or more endorsement events;
determining, using the one or more computing devices, activity data associated with the one or more endorsement events;
aggregating, using the one or more computing devices, the activity data to an aggregation level, wherein aggregating the activity data comprises: determining the aggregation level; generating an aggregation value for the aggregation level from the activity data, the aggregation value representing a number of endorsement events performed in the aggregation level; and storing the aggregation value in an aggregation table;
detecting using the one or more computing devices, based at least in part on the aggregation level of the activity data, one or more anomalies, wherein the one or more anomalies represent suspicious behavior associated with the one or more endorsement events; and
taking, using the one or more computing devices, one or more protection actions responsive to the detection of the one or more anomalies.
2. The method of claim 1 , wherein aggregating the activity data comprises:
aggregating the activity data to create one or more aggregation tables, wherein the one or more aggregation tables include one or more of a user table and a target table.
3. The method of claim 1 , wherein detecting the one or more anomalies comprises:
detecting the one or more anomalies based at least in part on one or more aggregation tables, wherein the one or more aggregation tables include one or more of a user table and a target table.
4. The method of claim 1 , wherein detecting the one or more anomalies comprises:
determining an aggregation level;
retrieving an aggregation value associated with the aggregation level from an aggregation table; and
detecting the one or more anomalies based at least in part on the aggregation value.
5. The method of claim 1 , wherein taking the one or more protection actions comprises:
generating one or more review leads for the one or more anomalies;
providing the one or more review leads to a reviewer;
receiving one or more review verdicts for the one or more review leads; and
updating a decision history based at least in part on the one or more review verdicts.
6. The method of claim 1 , wherein taking the one or more protection actions comprises:
determining one or more suspicious targets associated with the one or more anomalies; and
adding the one or more suspicious targets to a blacklist.
7. The method of claim 1 , further comprising:
identifying, using the one or more computing devices, one or more targets associated with the one or more endorsement events;
determining, using the one or more computing devices, whether the one or more targets are in a blacklist; and
responsive to the determination that the one or more targets are in the blacklist, blocking the one or more endorsement events using the one or more computing devices.
8. A computer program product comprising a non-transitory computer usable medium including a computer readable program, wherein the computer readable program when executed on a computer causes the computer to:
retrieve event data describing one or more endorsement events; determine activity data associated with the one or more endorsement events;
aggregate the activity data to an aggregation level, wherein aggregating the activity data comprises: determining the aggregation level; generating an aggregation value for the aggregation level from the activity data, the aggregation value representing a number of endorsement events performed in the aggregation level; and storing the aggregation value in an aggregation table;
detect, based at least in part on the aggregation level of the activity data, one or more anomalies, wherein the one or more anomalies represent suspicious behavior associated with the one or more endorsement events; and
take one or more protection actions responsive to the detection of the one or more anomalies.
9. The computer program product of claim 8 , wherein aggregating the activity data comprises:
aggregating the activity data to create one or more aggregation tables, wherein the one or more aggregation tables include one or more of a user table and a target table.
10. The computer program product of claim 8 , wherein detecting the one or more anomalies comprises:
detecting the one or more anomalies based at least in part on one or more aggregation tables, wherein the one or more aggregation tables include one or more of a user table and a target table.
11. The computer program product of claim 8 , wherein detecting the one or more anomalies comprises:
determining an aggregation level;
retrieving an aggregation value associated with the aggregation level from an aggregation table; and
detecting the one or more anomalies based at least in part on the aggregation value.
12. The computer program product of claim 8 , wherein taking the one or more protection actions comprises:
generating one or more review leads for the one or more anomalies;
providing the one or more review leads to a reviewer;
receiving one or more review verdicts for the one or more review leads; and
updating a decision history based at least in part on the one or more review verdicts.
13. The computer program product of claim 8 , wherein taking the one or more protection actions comprises:
determining one or more suspicious targets associated with the one or more anomalies; and
adding the one or more suspicious targets to a blacklist.
14. The computer program product of claim 8 , wherein the computer readable program when executed on the computer also causes the computer to:
identify one or more targets associated with the one or more endorsement events;
determine whether the one or more targets are in a blacklist; and
responsive to the determination that the one or more targets are in the blacklist, block the one or more endorsement events.
15. A system comprising:
one or more processors, the one or more processors being configured to:
retrieve event data describing one or more endorsement events;
determine activity data associated with the one or more endorsement events;
aggregate the activity data to an aggregation level, wherein the one or more processors aggregate the activity data by: determining the aggregation level; generating an aggregation value for the aggregation level from the activity data, the aggregation value representing a number of endorsement events performed in the aggregation level; and storing the aggregation value in an aggregation table;
detect, based at least in part on the aggregation level of the activity data, one or more anomalies, wherein the one or more anomalies represent suspicious behavior associated with the one or more endorsement events; and
take one or more protection actions responsive to the detection of the one or more anomalies.
16. The system of claim 15 , wherein the one or more processors aggregate the activity data by:
aggregating the activity data to create one or more aggregation tables, wherein the one or more aggregation tables include one or more of a user table and a target table.
17. The system of claim 15 , wherein the one or more processors detect the one or more anomalies associated with the one or more endorsement events by:
detecting the one or more anomalies based at least in part on one or more aggregation tables, wherein the one or more aggregation tables include one or more of a user table and a target table.
18. The system of claim 15 , wherein the one or more processors detect the one or more anomalies associated with the one or more endorsement events by:
determining an aggregation level;
retrieving an aggregation value associated with the aggregation level from an aggregation table; and
detecting the one or more anomalies based at least in part on the aggregation value.
19. The system of claim 15 , wherein the one or more processors take the one or more protection actions by:
generating one or more review leads for the one or more anomalies;
providing the one or more review leads to a reviewer;
receiving one or more review verdicts for the one or more review leads; and
updating a decision history based at least in part on the one or more review verdicts.
20. The system of claim 15 , wherein the one or more processors take the one or more protection actions by:
determining one or more suspicious targets associated with the one or more anomalies; and
adding the one or more suspicious targets to a blacklist.
21. The system of claim 15 , wherein the one or more processors are further configured to:
identify one or more targets associated with the one or more endorsement events;
determine whether the one or more targets are in a blacklist; and
responsive to the determination that the one or more targets are in the blacklist, block the one or more endorsement events.