Host abnormal behavior detection method based on subgraph matching
The invention discloses a host abnormal behavior detection method based on subgraph matching, which comprises the following steps: collecting abnormal behavior log data to construct a first heterogeneous graph, and extracting behavior meta-graphs based on the first heterogeneous graph to form a beha...
Saved in:
Main Authors | , , , , , , |
---|---|
Format | Patent |
Language | Chinese English |
Published |
21.11.2023
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | The invention discloses a host abnormal behavior detection method based on subgraph matching, which comprises the following steps: collecting abnormal behavior log data to construct a first heterogeneous graph, and extracting behavior meta-graphs based on the first heterogeneous graph to form a behavior meta-graph library; operating system kernel events are collected to construct a second heterogeneous graph; extracting a behavior meta-graph from the behavior meta-graph library, and searching candidate nodes matched with meta-graph nodes in the second heterogeneous graph for each meta-graph node in the behavior meta-graph; selecting a meta-graph node with the least number of candidate nodes as a seed node; traversing from each candidate node in the candidate node set of the seed nodes, simplifying a matching result by using an influence score between the nodes, and determining a matching sub-graph; and calculating the similarity between the behavior meta-graph and the matched sub-graph, and if the similarity |
---|---|
Bibliography: | Application Number: CN202310804303 |