Real-Time Data Mining for Event Streams

Roudjane, Massiva; Rebaïne, Djamal; Khoury, Raphaël; Hallé, Stéphane

doi:10.1109/edoc.2018.00025

Cited by 5 publications

(1 citation statement)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…BeepBeep 3 aims to present reusable, tested, and general toolkits that reduce the development effort of continuous event processing and express this processing in a more readable way and with a higher level of abstraction. BeepBeep 3 does not present predefined processors for anomaly detection, although such extensions exist [21]. BeepBeep forms more complex computations on the data by composing (or piping) processors between them, which is achieved by letting the output of one processor be the input of another.…”

Section: Related Workmentioning

confidence: 99%

Development of monitoring systems for anomaly detection using ASTD specifications

Chaymae,

Marc,

Thibaud

et al. 2022

Preprint

View full text Add to dashboard Cite

Anomaly-based intrusion detection systems are essential defenses against cybersecurity threats because they can identify anomalies in current activities. However, these systems have difficulties providing entity processing independence through a programming language. In addition, a degradation of the detection process is caused by the complexity of scheduling the training and detection processes, which are required to keep the anomaly detection system continuously updated. This paper shows how to use the algebraic state-transition diagram (ASTD) language to develop flexible anomaly detection systems. This paper provides a model for detecting point anomalies using the unsupervised nonparametric technique Kernel Density Estimation to estimate the probability density of event occurrence. The proposed model caters for both the training and the detection phase continuously. The ASTD language streamlines the modeling of detection systems thanks to its process algebraic operators that provide a solution to overcome these challenges. By delegating the combination of anomaly-based detection processes to the ASTD language, the effort and complexity are reduced during detection models development. Finally, using a qualitative evaluation, this study demonstrates that the algebraic operators in the ASTD specification language overcome these challenges.

show abstract

Section: Related Workmentioning

confidence: 99%