Alert correlating using sequence model with topology reinforcement systems and methods

Abstract

Alert correlation plays an important role in IT event management. It helps reduce the number of alerts that IT staff have to act upon. The disclosure describes a method, a computer program product that applies a machine driven deep learning model to effectively correlate alerts caused by a common root cause. In addition, this method of correlation provides the user the context of the root cause. Therefore, it helps the user to quickly identify, understand and resolve the problem thereby reducing the mean time to identification and resolution. Alerts that are caused by the same root cause therefor come together. In the machine learning world, language sequence models are doing very well on learning the sequence patterns between words. For example, the machine can learn the subtle difference between choice of words and the order of words in order to fake a person's writing. The disclosed embodiments use similar technology but apply it on IT resource and application monitoring alerts across private and public clouds to learn the alert's sequence pattern. Once the sequence model is trained with alert sequences, the model is fed with a stream of new alerts, the model then identifies the two or more alerts that are together or clustered. Clustered alerts are often caused by the same root cause and should be correlated as one unit of work to understand cause, impact and resolution.

Dependencies are inferred from the claim text.