Event mini-graphs in data intake stage of machine data processing platform

Patent metadata

Field	Value
Publication number	US-9596254-B1
Application number	US-201514929203-A
Country	US
Kind code	B1
Filing date	Oct 30, 2015
Priority date	Aug 31, 2015
Publication date	Mar 14, 2017
Grant date	Mar 14, 2017

Abstract

A security platform employs a variety techniques and mechanisms to detect security related anomalies and threats in a computer network environment. The security platform is “big data” driven and employs machine learning to perform security analytics. The security platform performs user/entity behavioral analytics (UEBA) to detect the security related anomalies and threats, regardless of whether such anomalies/threats were previously known. The security platform can include both real-time and batch paths/modes for detecting anomalies and threats. By visually presenting analytical results scored with risk ratings and supporting evidence, the security platform enables network security administrators to respond to a detected anomaly or threat, and to take action promptly.

First claim

What is claimed is:

1. A method for processing data at data intake for detection of an anomaly in a distributed computer environment, the method comprising:
receiving event data representing an event on a computer network, the event data being indicative of a plurality of entities and an action involved in the event;
identifying the entities and a relationship between the entities, based on the action in the event data;
creating, for the event, a record of the relationship between the entities by using a data structure representing a relationship graph, the relationship graph including at least two nodes and an edge between the two nodes, each node representing one of the entities, the edge representing the relationship between the entities; and
creating an updated event data by appending the event data representing the event with the record of the relationship;
wherein the record of the relationship is specific to the event;
sending the updated event data to an event processing engine for further processing;
generating a composite relationship graph that is combined from a plurality of the relationship graph corresponding to a plurality of the updated event data; and
using the event processing engine to perform analytics on the plurality of the updated event data and the composite relationship graph;
wherein the anomaly detection is performed based on applying a machine learning model to perform analytics on at least a portion of the composite relationship graph.

2. The method of claim 1 , wherein the relationship between the entities represents an activity, recorded in the event data, performed by one entity with respect to another entity.

3. The method of claim 1 , further comprising:
sending the updated event data to a distributed messaging system.

4. The method of claim 1 , further comprising:
using the event processing engine to identify security-oriented anomalies in the plurality of events.

5. The method of claim 1 , further comprising:
sending the record of the relationship to the event processing engine for further processing; and
using the event processing engine to use a downstream entity to perform analytics on the plurality of the updated event data and the composite relationship graph.

6. The method of claim 1 , further comprising:
repeating said receiving and identifying steps for event data representing each of a plurality of events on the computer network to generate the relationship graph for each of the plurality of events.

7. The method of claim 1 , further comprising:
repeating said receiving and identifying steps for event data representing each of a plurality of events on the computer network to generate the relationship graph for each of the plurality of events; and
combining additional relationship graphs for the plurality of events into the composite relationship graph.

8. The method of claim 1 , further comprising:
repeating said receiving and identifying steps for event data representing each of a plurality of events on the computer network to generate the relationship graph for each of the plurality of events; and
combining additional relationship graphs for the plurality of events into the composite relationship graph;
wherein the composite relationship graph includes all identified relationships among all identified entities involved in the plurality of events.

9. The method of claim 1 , wherein said identifying step comprises:
tokenizing the event data by extracting, as tokens, a key or a value or a key-value pair in the event data; and
parsing the event data based on a predetermined data format that specifies which tokens represent the entities and which tokens represent actions in the extracted tokens.

10. The method of claim 1 , wherein said identifying step comprises:
parsing the event data based on a predetermined data format that specifies which data represent the entities and which data represent the action in the event.

11. The method of claim 1 , wherein said identifying step comprises:
parsing the event data based on a predetermined data format that specifies which data represent the entities and which data represent the action in the event,
wherein the predetermined data format comprises a data format representing output from at least one of: an active directory, a proxy, a firewall, a web gateway, a virtual private network (VPN) connection, an intrusion detection system, a network traffic analyzer, or a malware engine.

12. The method of claim 1 , wherein identifying the entities and the action in the event comprises:
tokenizing the event data by extracting, as tokens, a key or a value or a key-value pair in the event.

13. The method of claim 1 , wherein the action is performed by one entity with respect to another entity, and wherein the identified relationship between the entities is identified based on the action.

14. The method of claim 1 , wherein the identified relationship between the entities is indicative of the action, and wherein the identified relationship is identified based on comparing the action with a table of identifiable relationships.

15. The method of claim 1 , wherein said identifying step comprises:
detecting a data format of the event data.

16. The method of claim 1 , further comprising:
issuing a query to a data processing system that performs the query against data stored in a distributed file system,
wherein said receiving step includes receiving the event data as a result of the query from the processing system.

17. The method of claim 1 , further comprising:
issuing a query to a data processing system that performs the query against data stored in a distributed file system,
wherein said receiving step includes receiving the event data as a result of the query from the processing system, and
wherein the data processing system includes a framework that provides methods including: a map method that performs a data processing operation on local data on distributed nodes, and a reduce method that performs a summary operation to generate a result based on the processed local data.

18. The method of claim 1 , further comprising:
requesting a machine on the computer network to transmit the event data, and wherein said receiving event data occurs in response to the machine on the computer network transmitting the event data.

19. The method of claim 1 , wherein said identifying step comprises identifying a timestamp in the event.

20. The method of claim 1 , further comprising:
configuring said identifying step by making an adjustment to a configuration file.

21. The method of claim 1 , wherein the event data comprises timestamped machine data.

22. The method of claim 1 , wherein the method is performed as part of an extract-transform-load stage of at least one of a distributed event processing system or an anomaly detection system.

23. The method of claim 1 , further comprising:
identifying a plurality of attributes of the event, based on the event data; and
adding a view identifier to the event data to allow a downstream entity, by having designated the view identifier, to receive select information extracted from and/or generated based on the plurality of attributes of the event, through an interface identified by the view identifier.

24. The method of claim 1 , further comprising:
identifying a plurality of attributes of the event, based on the event data; and
adding a view identifier to the event data to al

Assignees

• Splunk Inc

Inventors

• Muddu Sudhakar
• Tryfonas Christos
• Bulusu Ravi Prasad

Classifications

• G06N20/20 Primary

  Ensemble learning · CPC title

• H04L63/1416

  Event detection, e.g. attack signature detection · CPC title

• G06N7/01

  Probabilistic graphical models, e.g. probabilistic networks · CPC title

• G06F40/134

  Hyperlinking · CPC title

• H04L63/20

  for managing network security; network security policies in general (filtering policies H04L63/0227) · CPC title