Continuous knowledge graph generation using causal event graph feedback
Described systems and techniques determine causal associations between events that occur within an information technology landscape. Individual situations that are likely to represent active occurrences requiring a response may be identified as causal event clusters, without requiring manual tuning to determine cluster boundaries. Consequently, it is possible to identify root causes, analyze effects, predict future events, continuously generate a knowledge graph, and prevent undesired outcomes, even in complicated, dispersed, interconnected systems.
1 . A computer-implemented method for generating a knowledge graph, the method comprising:
requesting feedback on a causal graph;
collecting and aggregating the feedback on the causal graph;
processing the feedback on the causal graph and a spatiotemporal context of the causal graph using a machine learning model;
generating a knowledge graph based on output of the machine learning model;
generating a new causal graph using the knowledge graph;
requesting new feedback on the new causal graph, the new feedback occurring at a first level of the knowledge graph;
collecting and aggregating the new feedback on the new causal graph;
processing the knowledge graph, the new feedback on the new causal graph, and a new spatiotemporal context of the new causal graph using the machine learning model;
updating the knowledge graph based on new output of the machine learning model, wherein updating the knowledge graph occurs at a second level of the knowledge graph, wherein the first level is different than the second level; and
determining, by an Information Technology (IT) landscape manager, a root cause from the knowledge graph and predicting an occurrence of similar situations in the future to prevent undesired network events without human intervention.
2 . The computer-implemented method as in claim 1 , wherein generating the knowledge graph comprises generating a hierarchical knowledge graph.
3 . The computer-implemented method as in claim 1 , wherein the machine learning model comprises a graph neural network.
4 . The computer-implemented method as in claim 1 , wherein the feedback includes both positive feedback and negative feedback on the causal graph.
5 . The computer-implemented method as in claim 1 , wherein requesting the feedback on the causal graph comprises:
displaying the causal graph; and
displaying a binary input associated with the causal graph, wherein the binary input includes a positive feedback input and a negative feedback input.
6 . A computer program product for generating a knowledge graph, the computer program product being tangibly embodied on a non-transitory computer-readable medium and including executable code that, when executed, causes a computing device to:
request feedback on a causal graph;
collect and aggregate the feedback on the causal graph;
process the feedback on the causal graph and a spatiotemporal context of the causal graph using a machine learning model;
generate a knowledge graph based on output of the machine learning model;
generate a new causal graph using the knowledge graph;
request new feedback on the new causal graph, the new feedback occurring at a first level of the knowledge graph;
collect and aggregate the new feedback on the new causal graph;
process the knowledge graph, the new feedback on the new causal graph, and a new spatiotemporal context of the new causal graph using the machine learning model;
update the knowledge graph based on new output of the machine learning model, wherein updating the knowledge graph occurs at a second level of the knowledge graph, wherein the first level is different than the second level; and
determine, by an Information Technology (IT) landscape manager, a root cause from the knowledge graph and predict an occurrence of similar situations in the future to prevent undesired network events without human intervention.
7 . The computer program product of claim 6 , wherein generating the knowledge graph includes causing the computing device to generate a hierarchical knowledge graph.
8 . The computer program product of claim 6 , wherein the machine learning model comprises a graph neural network.
9 . The computer program product of claim 6 , wherein the feedback includes both positive feedback and negative feedback on the causal graph.
10 . The computer program product of claim 6 , wherein requesting the feedback on the causal graph includes causing the computing device to:
display the causal graph; and
display a binary input associated with the causal graph, wherein the binary input includes a positive feedback input and a negative feedback input.
11 . A system for generating a knowledge graph, comprising:
at least one processor; and
a non-transitory computer readable medium comprising instructions that, when executed by the at least one processor, cause the system to:
request feedback on a causal graph;
collect and aggregate the feedback on the causal graph;
process the feedback on the causal graph and a spatiotemporal context of the causal graph using a machine learning model;
generate a knowledge graph based on output of the machine learning model;
generate a new causal graph using the knowledge graph;
request new feedback on the new causal graph, the new feedback occurring at a first level of the knowledge graph;
collect and aggregate the new feedback on the new causal graph;
process the knowledge graph, the new feedback on the new causal graph, and a new spatiotemporal context of the new causal graph using the machine learning model;
update the knowledge graph based on new output of the machine learning model, wherein updating the knowledge graph occurs at a second level of the knowledge graph, wherein the first level is different than the second level; and
determine, by an Information Technology (IT) landscape manager, a root cause from the knowledge graph and predict an occurrence of similar situations in the future to prevent undesired network events without human intervention.
12 . The system of claim 11 , wherein generating the knowledge graph includes causing the system to generate a hierarchical knowledge graph.
13 . The system of claim 11 , wherein the machine learning model comprises a graph neural network.
14 . The system of claim 11 , wherein the feedback includes both positive feedback and negative feedback on the causal graph.
15 . The system of claim 11 , wherein requesting the feedback on the causal graph includes causing the system to:
display the causal graph; and
display a binary input associated with the causal graph, wherein the binary input includes a positive feedback input and a negative feedback input.