A graph-based approach to detect unexplained sequences in a log

作者：

Highlights：

• A graph mining approach has been designed for recognizing anomalous sequences.

• It supports both real time and batch processing for large scale data analysis.

• A probabilistic penalty graph has been used for modeling log temporal sequences.

• The approach’s effectiveness has been evaluated for different system configurations.

摘要

•A graph mining approach has been designed for recognizing anomalous sequences.•It supports both real time and batch processing for large scale data analysis.•A probabilistic penalty graph has been used for modeling log temporal sequences.•The approach’s effectiveness has been evaluated for different system configurations.

论文关键词：Log analysis,Event detection,Graph mining

论文评审过程：Received 16 October 2020, Revised 23 December 2020, Accepted 30 December 2020, Available online 7 January 2021, Version of Record 21 January 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114556