Back to Volume
Paper: A Novelty Method for Delay Detection based on Software Logs
Volume: 522, Astronomical Data Analysis Software and Systems XXVII
Page: 607
Authors: Gil, J. P.; Russo, B.; Avila, A.
Abstract: This work introduces a new method to extract typical intervals of times between events based on the information contained in software logs, and an application to identify outliers representing unexpected delays in the generated dataset. This research was motivated by real examples on difficult detection of processes that suddenly become slow and the problem goes unnoticed to operations engineers. The contribution of this work is to propose a procedure to efficiently analyze delays between pairs of related events combining techniques from process mining, data science and text mining.
Back to Volume