GraphDL: A visibility graph-based structural representation framework for flow-based cyber-attack detection
Ali Melih Kanca, İlker Türker · 2026
Read GraphDL: A visibility graph-based structural representation framework for flow-based cyber-attack detection by Ali Melih Kanca, İlker Türker (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13578295.txt.
Description
Ali Melih Kanca and İlker Türker built a detection system called GraphDL that turns network traffic into graphs before hunting for cyber-attacks. Flow data were cut into time windows, and a visibility graph was drawn for each traffic feature. Measures such as centrality and clustering were then pulled from those graphs and fed into a convolutional neural network classifier.
Tested on the CICIDS2018 data set, accuracy ranged from about 76 percent to about 97 percent depending on window and feature choices. A simpler model using raw flow features performed strongly at far lower computing cost. GraphDL therefore offers a structural way to capture timing and nonlinear patterns, but its extra computation must be weighed against that advantage.
Catalog record
| CREATOR | Ali Melih Kanca, İlker Türker |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 128 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |