Splunk Inc.

Telefonaktiebolaget Lm Ericsson (publ)

Ericsson Ab

Redback Networks Inc.

George Apostolopoulos

Zhuxuan Jin

Rahul Aggarwal

Liming Wei

Bo Lei

Marios Iliofotou

Karthik Kannan

Joseph Auguste Zadeh

Rodolfo Soto

Jeswanth Manikonda

Essam Zaky

John Clifton Pierce

Sitaram Venkatraman

Brian Russell Kean

Peter J Owens

Diamantis Kourkouzelis

Ignacio Nicolas Bermudez Corrales

Ramanathan Lakshmikanthan

Chandramouli Balasubramanian

Peter Owens

Christian Stabel Daudt

A method is disclosed that includes receiving, at a computing device, an event log including events derived from machine data, and determining a score by comparing an event from the event log with frequent patterns of features. Determining the score includes determining a length of a frequent pattern within the event in the event log and a count of occurrences of the frequent pattern within the events, determining a contribution of the frequent pattern based on the length and the count, determining a penalty for an unmatched feature of the first event based on a cardinality of the events, and averaging the contribution and the penalty to obtain the score. The method further includes issuing an alert identifying the first event as an anomaly using the first score and an anomaly score threshold.

Anomaly event detection

One or more embodiments are directed multistage device clustering. A log including network traffic of multiple devices in a network is received. From the log, features of the devices are extracted and an aggregated feature matrix generated. A traffic behavior subset of the features in the aggregated feature matrix is selected, and a topic modeling algorithm applied thereto to obtain traffic behavior device groups. An application behavior subset of the features in the aggregated feature matrix is selected. On a per traffic behavior device group basis, the topic modeling algorithm is applied to the application behavior subset to obtain application behavior device subgroups. One or more devices are assigned to at least one of the plurality of application behavior device subgroups to obtain an assignment.

Multistage device clustering

Embodiments of the present invention are directed to facilitating detection of suspicious access to resources. In accordance with aspects of the present disclosure, an access graph is generated. The access graph contains access data that includes observed accesses between entities and resources. Access scores can be determined for entity-resource pairs in the access graph by applying a set of access rules to the entity-resource pairs in the access graph. The access scores indicate an extent of relatedness between the corresponding entity and resource. Thereafter, the access scores can be used to train a probabilistic prediction model that predicts suspiciousness of accesses between entities and resources.

Predicting suspiciousness of access between entities and resources

A network connection between a server group of a data intake and query system and each of one or more source network nodes is established. Source data at the server group is received from at least one of the one or more source network nodes via the respective network connections and transformed, by the indexer server, to timestamped data entries of machine data. A model management server detects data constraints for a security model that include a data element used by the security model and an availability requirement set. Using the timestamped data entries, the data constraints are validated, and the validation used to determine a data availability assessment of the security model.

Systems data availability validation

A method is disclosed that includes receiving, at a computing device, an event log including multiple events, where the events are derived from machine data, determining a first score associated with a first granularity level by comparing an event from the event log with a first frequent patterns generated for the first granularity level, and determining a second score associated with a second granularity level by comparing the event with a second frequent patterns generated for the second granularity level. The method further includes determining an aggregate score for the event based on the first score and the second score, and comparing the aggregate score for the event with an anomaly score threshold. Further, the method includes issuing an alert identifying the event as an anomaly based on the aggregate score exceeding the anomaly score threshold.

Frequent pattern based anomaly event detection

Techniques are described for analyzing data regarding activity in an IT environment to determine information regarding the entities associated with the activity and using the information to detect anomalous activity that may be indicative of malicious activity. In an embodiment, a plurality of events reflecting activity by a plurality of entities in an IT environment are processed to resolve the identities of the entities, discover how the entities fit within a topology of the IT environment, and determine what the entities are. This information is then used to generate an entity relationship graph that includes nodes representing the entities in the IT environment and edges connecting the nodes representing interaction relationships between the entities. In some embodiments, baselines are established by monitoring the activity between entities. This baseline information can be represented in the entity relationship graph in the form of directionality applied to the edges. The entity relationship graph can then be monitored to detect anomalous activity.

Anomaly detection based on changes in an entity relationship graph

A network connection between a server group of a data intake and query system and each of one or more source network nodes is established. Source data at the server group is received from at least one of the one or more source network nodes via the respective network connections and transformed, by the indexer server, to timestamped entries of machine data. A model management server detects data constraints for a security model. Using the timestamped entries, the data constraints are validated to obtain a validation result, where validating the data constraints includes determining whether the timestamped entries satisfy the availability requirement set for the data element. The model management server determines a data availability assessment of the security model based on the validation result.

Validation of systems data

One or more embodiments are directed behavioral based device clustering. A network traffic log of devices in the network is received. Features of devices are extracted from the network traffic log and aggregated into an aggregated feature matrix on a per device basis. By applying a topic modeling algorithm to the aggregated feature matrix, the devices are clustered into device groups according to behavior groups. A device is assigned to the device group to create an assignment.

Behavioral based device clustering

Facilitating detection of suspicious access to resources

A method is disclosed. The method includes: receiving, at a computing device, an event log including a plurality of events, where the plurality of events are derived from machine data generated by components of an information technology environment; determining a first score associated with a first granularity level by comparing a first event from the event log with a first plurality of frequent patterns generated for the first granularity level; determining a second score associated with a second granularity level by comparing the first event with a second plurality of frequent patterns generated for the second granularity level; determining an aggregate score for the first event based on the first score and the second score; comparing the aggregate score for the first event with an anomaly score threshold; and issuing an alert identifying the first event as an anomaly based on the aggregate score exceeding the anomaly score threshold.

Anomaly event detection using frequent patterns

Behavioral based device clustering system and method

Anomaly detection based on information technology environment topology

A network connection between a server group of a data intake and query system and each of one or more source network nodes is established. The server group includes an indexer server and a model management server. Source data at the server group is received from at least one of the one or more source network nodes. A model management server detects data constraints for a security model. The data constraints include a data element used by the security model and an availability requirement set. Using the timestamped entries, the data constraints are validated to obtain a validation result. The model management server determines a data availability assessment of the security model based on the validation result. The data availability assessment of the security model is stored in computer storage.

Systems data validation

Techniques are described for analyzing data regarding activity in an IT environment to determine information regarding the entities associated with the activity and using the information to detect anomalous activity that may be indicative of malicious activity. In an embodiment, a plurality of events reflecting activity by a plurality of entities in an IT environment are processed to resolve the identities of the entities, discover how the entities fit within a topology of the IT environment, and determine what the entities are. This information is then used to generate a entity relationship graph that includes nodes representing the entities in the IT environment and edges connecting the nodes representing interaction relationships between the entities. In some embodiments, baselines are established by monitoring the activity between entities. This baseline information can be represented in the entity relationship graph in the form of directionality applied to the edges. The entity relationship graph can then be monitored to detect anomalous activity.

Fingerprinting entities based on activity in an information technology environment

A source edge router initiates point-to-point (P2P) label switched paths (LSPs) to multiple receiver edge routers, sends one or more PATH messages to one or more nodes downstream of the source edge router for the P2P LSPs, wherein each one of the set of PATH messages includes a same identifier of a point-to-multipoint (P2MP) LSP that includes a session object that contains a tunnel identifier and a P2MP identifier, and receives from each of those nodes downstream of the source edge router, a set of one or more RESV messages, wherein each one of the received RESV messages received from a respective node downstream of the source edge router includes the identifier of the P2MP LSP and a label allocated by that node downstream of the source edge router.

Growing community of inventors

San Jose, CA, United States of America

George Apostolopoulos