Abstract: Information technology service management (ITSM) incident reports are converted from textual data to multiple vectors using an encoder and parameters are selected, where the parameters include a base cluster number and a threshold value. A base group of clusters is generated using an unsupervised machine learning clustering algorithm with the vectors and the parameters as input. A cluster quality score is computed for each of the base group of clusters. Each cluster from the base group of clusters with the cluster quality score above the threshold value is recursively split into new clusters until the cluster quality score for each cluster in the new clusters is below the threshold value. A final group of clusters is output, where each cluster from the final group of clusters represents ITSM incident reports related to a same problem.

Abstract: Described techniques determine performance metric values of a performance metric characterizing a performance of a system resource of an information technology (IT) system, and determine driver metric values of a driver metric characterizing an occurrence of an event that is at least partially external to the system resource. A correlation analysis may confirm a potential correlation between the performance metric values and the driver metric values as a correlation. A graph relating the performance metric to the driver metric may be generated. A plurality of extrapolation algorithms may be trained to obtain a plurality of trained extrapolation algorithms using a first subset of data points of the graph, and the plurality of trained extrapolation algorithms may be validated using a second subset of data points of the graph. A driver metric threshold corresponding to the performance metric threshold may be determined using a validated extrapolation algorithm.

Abstract: 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, and prevent undesired outcomes, even in complicated, dispersed, interconnected systems.