An Ensemble Approach for Detection of Malicious URLs Using SOM and Tabu Search Optimization
The fast spread of malicious URLs is a significant risk to online safety, since it makes assaults like spam, phishing, virus distribution, and vandalism of websites easier to carry out. The dynamic nature of these threats makes traditional detection techniques unable to keep up. Enhanced methods tha...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Erbil Polytechnic University
2025-07-01
|
| Series: | Polytechnic Journal |
| Subjects: | |
| Online Access: | https://polytechnic-journal.epu.edu.iq/home/vol15/iss2/2 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | The fast spread of malicious URLs is a significant risk to online safety, since it makes assaults like spam, phishing, virus distribution, and vandalism of websites easier to carry out. The dynamic nature of these threats makes traditional detection techniques unable to keep up. Enhanced methods that can handle large-scale datasets and identify new attack patterns are needed for the real-time identification of malicious URLs. This research presents a strong hybrid machine learning approach that combines accurate classification with effective feature extraction. For feature extraction, we provide a Self-Organizing Map based Radial Movement Optimization (SOM-RMO); for classification, we present an Ensemble Radial Basis Function Network (ERBFN) optimized by Tabu Search. An RBFN tuned via Tabu Search guarantees high accuracy in the harmful URL categorization; meanwhile, the SOM-RMO efficiently performs dimensionality reduction, accentuating vital features. Our model performs better than other models in a variety of malicious URL attack scenarios. It substantially outperformed conventional detection techniques, attaining an accuracy of 97.1 %, precision of 96.4 %, recall of 95.8 %, and an F1-score of 96.0 % on a benchmark dataset. |
|---|---|
| ISSN: | 2707-7799 |