Big log data stream processing: Adapting an anomaly detection technique

Abstract

With the continuous increase in data velocity and volume nowadays, preserving system and data security is particularly affected. In order to handle the huge amount of data and to discover security incidents in real-time, analyses of log data streams are required. However, most of the log anomaly detection techniques fall short in considering continuous data processing. Thus, this paper aligns an anomaly detection technique for data stream processing. It thereby provides a conceptual basis for future adaption of other techniques and further delivers proof of concept by prototype implementation.