US 2023/0409463 A1 · MACHINE-LEARNING-BASED TECHNIQUES FOR PREDICTIVE MONITORING OF A SOFTWARE APPLICATION FRAMEWORK

Abstract

Various embodiments of the present invention provide methods, apparatuses, computing devices, and/or the like that are configured to perform software application framework monitoring using alert signatures for the software applications that are generated by at least one of: (i) a conditional ensemble machine learning framework comprise one or more alert priority score generation machine learning models, one or more alert priority explanation generation machine learning models, and a conditional ensemble machine learning model that is configured to generate an explanation-inclusive alert signature if a deep-learning-based alert priority score generated by the alert priority score generation machine learning models is identical to a decision-tree-based alert priority designation generated by the alert priority explanation generation machine learning models, and (ii) a set of alert priority score adjustment models such as an entity-based alert priority score adjustment model, a temporal alert priority score adjustment model, and a similarity-based alert priority score adjustment model.

Claims20

Dependent claims

An apparatus for predictive monitoring of a software application framework, the apparatus comprising at least one processor and at least one non-transitory memory comprising a computer program code, the at least one non-transitory memory and the computer program code configured to, with the at least one processor, cause the apparatus to:

identify a software alert data object for the software application framework, wherein the software alert data object is associated with one or more alert attribute data fields and a deep-learning-based alert priority score;
generate, based on the one or more alert attribute data fields and using one or more alert priority score adjustment models, one or more alert priority score adjustment scores for the software alert data object;
generate an alert priority score for the software alert data object based on the deep-learning-based alert priority score and the one or more alert priority score adjustment scores;
generate an alert signature for the software alert data object based on the alert priority score;
and
perform one or more incident management actions based on the alert signature.

3 more elements

A computer-implemented method for predictive monitoring of a software application framework, the computer-implemented method comprising:

identifying a software alert data object for the software application framework, wherein the software alert data object is associated with one or more alert attribute data fields and a deep-learning-based alert priority score;
generating, based on the one or more alert attribute data fields and using one or more alert priority score adjustment models, one or more alert priority score adjustment scores for the software alert data object;
generating an alert priority score for the software alert data object based on the deep-learning-based alert priority score and the one or more alert priority score adjustment scores;
generating an alert signature for the software alert data object based on the alert priority score;
and
performing one or more incident management actions based on the alert signature.

3 more elements

Dependencies are inferred from the claim text.

Machine-learning-based techniques for predictive monitoring of a software application framework

Life of this patent

Machine-learning-based techniques for predictive monitoring of a software application framework

Life of this patent

Abstract

Claims20