Impact scoring and reducing false positives
Some embodiments of the present invention provide systems and methods for detecting anomalies in network traffic. Some embodiments detect anomalies based on time-series activity in network traffic. Upon detection of an anomaly, significant changes can be analyzed to identify abnormal changes in network traffic across different network entities. The identified changes can then be used to determine the cause and the impact of the detected anomaly on the network traffic.
1. A non-transitory computer readable storage medium storing instructions that when executed by a computer cause the computer to perform a method for determining the impact of an anomalous event on a network entity, the method comprising:
detecting an anomalous event based on an outlier at a first network entity;
determining a time T 1 when the outlier occurred at the first network entity;
determining a first difference in a network traffic metric between time T 1 and time T 2 , wherein time T 2 is prior to time T 1 , and wherein the first difference is based on overall observed network traffic in a network;
determining a second difference in the network traffic metric between time T 1 and time T 2 , wherein the second difference is based on observed network traffic at a second network entity for time T 1 and time T 2 , wherein the second network entity is different from the first network entity; and
determining an impact score for the second network entity based on the first difference and the second difference.
2. The non-transitory computer readable storage medium of claim 1 , the method further comprising:
ranking the second network entity in the set of network entities based on the determined impact score for the second network entity.
3. The non-transitory computer readable storage medium of claim 1 , wherein the impact score for the second network entity is correlated to the contribution of the second network entity to the determined difference in the network traffic metric.
4. The non-transitory computer readable storage medium of claim 1 , the method further comprising:
displaying the impact score for the second network entity.
5. The non-transitory computer readable storage medium of claim 1 , the method further comprising determining the impact score using one of a 1-10 scale or a 1-100 scale.
6. The non-transitory computer readable storage medium of claim 1 , wherein time T 2 is one day prior to time T 1 .
7. The non-transitory computer readable storage medium of claim 1 , wherein time T 2 is one week prior to time T 1 .
8. The non-transitory computer readable storage of claim 1 , wherein time T 2 is one month prior to time T 1 .
9. A system for determining the impact of an anomalous event on a network entity, comprising:
a computing device configured to:
detect an anomalous event based on an outlier at a first network entity;
determine a time T 1 when the outlier occurred at the first network entity;
determine a first difference in a network traffic metric between time T 1 and time T 2 , wherein time T 2 is prior to time T 1 , and wherein the first difference is based on overall observed network traffic in a network;
determine a second difference in the network traffic metric between time T 1 and time T 2 , wherein the second difference is based on observed network traffic at a second network entity for time T 1 and time T 2 , wherein the second network entity is different from the first network entity; and
determine an impact score for the second network entity based on the first difference and the second difference.
10. The system of claim 9 , wherein the computing device is further configured to rank the second network entity based on the determined impact score.
11. The system of claim 9 , wherein the impact score for the second network entity is correlated to the contribution of the second network entity to the determined difference in the network traffic metric.
12. The system of claim 9 , wherein the computing device is further configured to display the impact score for the second network entity.
13. The system of claim 9 , wherein time T 2 is one day prior to time T 1 .
14. The system of claim 9 , wherein time T 2 is one week prior to time T 1 .
15. The system of claim 9 , wherein time T 2 is one month prior to time T 1 .
16. A method for determining the impact of an anomalous event on a network entity, comprising:
detecting an anomalous event based on an outlier at a first network entity;
determining a time T 1 when the outlier occurred at the first network entity;
determining a first difference in a network traffic metric between time T 1 and time T 2 , wherein time T 2 is prior to time T 1 , and wherein the first difference is based on overall observed network traffic in a network;
determining a second difference in the network traffic metric between time T 1 and time T 2 , wherein the second difference is based on observed network traffic at a second network entity for time T 1 and time T 2 , wherein the second network entity is different from the first network entity; and
determining an impact score for the second network entity based on the first difference and the second difference.
17. The method of claim 16 , further comprising ranking the second network entity based on the determined impact score.
18. The method of claim 16 , wherein the impact score for the second network entity is correlated to the contribution of the second network entity to the determined difference in the network traffic metric.
19. The method of claim 16 , further comprising displaying the impact score for the second network entity.
20. The method of claim 16 , wherein time T 2 is one day prior to time T 1 .
21. The method of claim 16 , wherein time T 2 is one week prior to time T 1 .
22. The method of claim 16 , wherein time T 2 is one month prior to time T 1 .