Anomaly Detection on System Generated Logs-A Survey Study

Recently, the application and systems are generating huge amounts of log data. With the rise of different applications in different domains that are deployed in every environment, it has become inevitable that a system may run into either functional or performance faults. System analyst and administrators have to look at system log data at the time of fault, and analysing millions of lines of logs manually is an impossible task to find the root cause for any faults. Timely detection of any abnormal behaviour is very important to prevent negative impact on the service, thereby helping to build a secure and trustworthy system. The aim of this survey is to provide a comprehensive overview of recent research in supervised and unsupervised anomaly detection methods. A detailed review on the various adoption methods and data sets on which it has been applied across various application domains are discussed in this paper.