“SIEVE: Generating a cybersecurity log dataset collection for SIEM event classification”

The research report”SIEVE: Generating a cybersecurity log dataset collection for SIEM event classification” was published in the first-rate magazine Computer Networks, vol. 266 -July 2025.

The research, carried out by colleagues Pierpaolo Artioli and Gianluca Pellegrini of the Cyber Lab in Grottaglie, in collaboration with the University of Bari, is part of the activities provided for in the “Suite Cybersecurity and SOC” Programme.

The aim of the research is to create an effective mechanism for creating synthetic logs of security events, given the limited availability of public datasets. These logs are needed to assess the performance of automated security event classification algorithms, which are essential for improving the capacity and effectiveness of SIEM systems. For this purpose, starting from existing datasets, using an appropriate combination of NLP techniques, the synthetic generation algorithm SPICE is introduced, which is extremely effective in generating realistic logs.

Synthetic logs generated through SIEVE have already been successfully used in the integration of advanced AI-based capabilities into BV TECH’s SIEM, and will soon be available for research purposes to the entire scientific community on cybersuite.it.