BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//cfp.securityfest.com//academic-village-2026//speaker//3M
 QPLX
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-academic-village-2026-SQNRZZ@cfp.securityfest.com
DTSTART;TZID=CET:20260527T151500
DTEND;TZID=CET:20260527T154500
DESCRIPTION:Transformer-based models have emerged as a powerful solution fo
 r network traffic classification\, achieving high accuracy by autonomously
  learning patterns in raw traffic data. However\, their high computational
  costs make real-time deployment impractical. In contrast\, industry-prove
 n tools like Snort and Suricata offer efficient network analysis but rely 
 on manually crafted signatures\, resulting in slower updates and limited a
 daptability to emerging threats.\n\nIn this work\, we propose a cascading 
 model that leverages the strengths of both approaches. During training\, a
  transformer-based model learns traffic patterns\, which are then extracte
 d using SHAP analysis to enhance the knowledge base of a signature-based I
 DS. In deployment\, the IDS handles routine classifications\, while only c
 omplex cases are escalated to the transformer model. Our experiments combi
 ning the analysis of ET-BERT with Snort demonstrate a four-fold performanc
 e improvement over running only ET-BERT without compromising false positiv
 e or false negative rates.
DTSTAMP:20260627T234110Z
LOCATION:Taube Room
SUMMARY:Snort Meets Transformers: Accelerating Transformer-Based Network Tr
 affic Classification for Real-Time Performance - Mohamed Hashim Changrampa
 di
URL:https://cfp.securityfest.com/academic-village-2026/talk/SQNRZZ/
END:VEVENT
END:VCALENDAR