Method and system for detecting abnormal online user activity
The present teaching generally relates to detecting abnormal user activity associated with an entity. In a non-limiting embodiment, baseline distribution data representing a baseline distribution characterizing normal user activities for an entity may be obtained. Information related to online user activities with respect to the entity may be received, distribution data representation a dynamic distribution may be determined based, at least in part, on the information. One or more measures characterizing a difference between the baseline distribution and the dynamic distribution may be computed, and in real-time it may be assessed whether the information indicates abnormal user activity. If the first information indicates abnormal user activity, then output data including the distribution data and the one or more measures may be generated.
1. A method for detecting abnormal online user activities, the method being implemented on a computer comprising at least one processor, storage, and communications circuitry, the method comprising:
determining at least one measure characterizing a difference between baseline information related to normal user activities with respect to an entity and user activity information related to online user activities with respect to the entity;
assessing in real-time whether the user activity information indicates abnormal user activity behavior based on the at least one measure, wherein an indication of the abnormal user activity behavior comprises detecting time-to-click abnormalities based on a temporal duration between when content is rendered on a user device and when a user interacts with the rendered content; and
generating, in response to determining that the user activity information indicates the abnormal user activity behavior, output data comprising the user activity information and the at least one measure.
2. The method of claim 1 , wherein the baseline information is generated offline and comprises a baseline distribution representing the normal user activities with respect to the entity during a temporal duration.
3. The method of claim 1 , wherein the user activity information is generated in real-time and comprises a dynamic distribution representing the online user activities with respect to the entity during the temporal duration.
4. The method of claim 1 , wherein the abnormal user activity behavior signifies fraudulent activities.
5. The method of claim 1 , wherein determining the at least one measure comprises:
performing a Kolmogorov-Smirnov (KS) measure,
performing an Information Value (IV) measure, or
performing the KS measure and the IV measure.
6. The method of claim 1 , wherein the output data further comprises an abnormal user activity flag indicating (i) that abnormal user activity behavior has been detected, (ii) an entity identifier indicating the entity with which the abnormal user activity behavior was detected with respect to, and (iii) a range of times associated with the online user activities.
7. A non-transitory computer readable medium having information recorded thereon for detecting abnormal online user activities, wherein the information when executed by one or more processors effectuate operations comprising:
determining at least one measure characterizing a difference between baseline information related to normal user activities with respect to an entity and user activity information related to online user activities with respect to the entity;
assessing in real-time whether the user activity information indicates abnormal user activity behavior based on the at least one measure, wherein an indication of the abnormal user activity behavior comprises detecting time-to-click abnormalities based on a temporal duration between when content is rendered on a user device and when a user interacts with the rendered content; and
generating, in response to determining that the user activity information indicates the abnormal user activity behavior, output data comprising the user activity information and the at least one measure.
8. The medium of claim 7 , wherein the baseline information is generated offline and comprises a baseline distribution representing the normal user activities with respect to the entity during a temporal duration.
9. The medium of claim 7 , wherein the user activity information is generated in real-time and comprises a dynamic distribution representing the online user activities with respect to the entity during the temporal duration.
10. The medium of claim 7 , wherein the abnormal user activity behavior signifies fraudulent activities.
11. The medium of claim 7 , wherein determining the at least one measure comprises:
performing a Kolmogorov-Smirnov (KS) measure,
performing an Information Value (IV) measure, or
performing the KS measure and the IV measure.
12. The medium of claim 7 , wherein the output data further comprises an abnormal user activity flag indicating (i) that abnormal user activity behavior has been detected, (ii) an entity identifier indicating the entity with which the abnormal user activity behavior was detected with respect to, and (iii) a range of times associated with the online user activities.
13. A system for detecting abnormal online user activities, the system comprising:
memory comprising one or more computer program instructions; and
at least one processor that, in response to executing the one or more computer program instructions, is configured to:
determine at least one measure characterizing a difference between baseline information related to normal user activities with respect to an entity and user activity information related to online user activities with respect to the entity;
assess in real-time whether the user activity information indicates abnormal user activity behavior based on the at least one measure, wherein an indication of the abnormal user activity behavior comprises detecting time-to-click abnormalities based on a temporal duration between when content is rendered on a user device and when a user interacts with the rendered content; and
generate, in response to determining that the user activity information indicates the abnormal user activity behavior, output data comprising the user activity information and the at least one measure.
14. The system of claim 13 , wherein the baseline information is generated offline and comprises a baseline distribution representing the normal user activities with respect to the entity during a temporal duration.
15. The system of claim 13 , wherein the user activity information is generated in real-time and comprises a dynamic distribution representing the online user activities with respect to the entity during the temporal duration.
16. The system of claim 13 , wherein the abnormal user activity behavior signifies fraudulent activities.
17. The system of claim 13 , wherein the at least one measure being determined comprises:
performing a Kolmogorov-Smirnov (KS) measure,
performing an Information Value (IV) measure, or
performing the KS measure and the IV measure.
18. The system of claim 13 , wherein the output data further comprises an abnormal user activity flag indicating (i) that abnormal user activity behavior has been detected, (ii) an entity identifier indicating the entity with which the abnormal user activity behavior was detected with respect to, and (iii) a range of times associated with the online user activities.