HYBRID MACHINE LEARNING APPROACH FOR REAL-TIME MALICIOUS URL DETECTION USING SOM-RMO AND RBFN WITH TABU SEARCH OPTIMIZATION

Aug 6, 2026

The document addresses the escalating risk posed by malicious URLs, including spam, phishing, malware, and defacement attacks, and highlights the limitations of traditional detection when threats evolve rapidly. It proposes a hybrid real-time model that unites SOM-RMO for feature extraction with an RBFN classifier improved by Tabu Search optimization. SOM-RMO reduces dimensionality and emphasizes significant features, while the optimized RBFN increases classification accuracy. Results on a benchmark dataset report 96.5% accuracy, 95.2% precision, 94.8% recall, and a 95.0% F1-score, outperforming conventional methods.

LanguageEnglish

CategoryTechnology

FormatPDF

Pages24

Reading Time1 h 0 min

AccessFree to download

1. Introduction

Internet growth and the rise of online illegal activity

Blacklisting and its limitations for new attacks

Role of machine learning and deep learning in cybersecurity

Research objectives

Main novelty of the proposed hybrid model

HYBRID MACHINE LEARNING APPROACH FOR REAL-TIME MALICIOUS URL DETECTION USING SOM-RMO AND RBFN WITH TABU SEARCH OPTIMIZATION Swetha T,Dr Seshaiah M ,Dr Hemalatha KL, Dr.ManjunathaKumar BH, Dr Murthy SVN Research Scholar,Associate Professor,Professor&HOD,Professor &HOD,Associate Professor Dept OfCSE,SJCIT,Dept of CSE, SJCIT ,Dept ISE,Dept of CSE SJCIT,,Dept CSE, SJCIT Mail Id: urs,[merikapudi@gmail.com](merikapudi@gmail.com),hema,hod,svn Abstract The proliferation of malicious URLs has become a significant threat to internet security, encompassing SPAM, phishing, malware, and defacement attacks. Traditional detection methods struggle to keep pace with the evolving nature of these threats. Detecting malicious URLs in real-time requires advanced techniques capable of handling large datasets and identifying novel attack patterns. The challenge lies in developing a robust model that combines efficient feature extraction with accurate classification. We propose a hybrid machine learning approach combining Self-Organizing Map based Radial Movement Optimization (SOM-RMO) for feature extraction and Radial Basis Function Network (RBFN) based Tabu Search for classification. SOM-RMO effectively reduces dimensionality and highlights significant features, while RBFN, optimized with Tabu Search, classifies URLs with high precision. The proposed model demonstrates superior performance in detecting various malicious URL attacks. On a benchmark dataset, our approach achieved an accuracy of 96.5%, precision of 95.2%, recall of 94.8%, and an F1-score of 95.0%, outperforming traditional methods significantly. Keywords: Malicious URL detection, Self-Organizing Map, Radial Movement Optimization, Radial Basis Function Network, Tabu Search 1. Introduction Many offline activities have moved online as a result of the Internet's expansion and development, including general business, social networking, e-commerce, and banking. As such, there is now a higher chance that illegal activity may occur online. This emphasises how urgently action must be done to maintain internet security [1] . To get sensitive data or compromise the system, people are being tricked into accessing dangerous URLs. This means that protecting this side is becoming a critical need because [2] . Malicious people can nevertheless attack the connection between the client and the server even in the presence of laws and standards. Phishing, spam, malware, and other types of attacks are all referred to as"malicious," as one umbrella term [3] . Because malicious URLs collect needless information and trick unwary end users into falling for scams, they result in yearly losses of billions of dollars. The online security world has created blacklisting services to help identify dangerous websites [4]-[6] . The goal was to identify the risk that dangerous websites pose. The blacklist is a database including every URL that has ever been deemed possibly dangerous. Apparently, there are circumstances when URL blacklisting is effective [7] . Nevertheless, an attacker can exploit these weaknesses by modifying the URL string in a way that makes the system readily fooled. Many harmful websites will unavoidably stay online because they are either too new, never examined, or had their evaluations incorrect. A further instrument in the arsenal for identifying dangerous websites are heuristics, which are basically an improved version of the signature-based blacklist method. One can compare the signatures of an old malicious URL and a new one. An additional line of protection against dangerous websites is offered by this approach. The techniques described here will help you distinguish between benign and malicious URLs. These more traditional methods do, however, have several shortcomings, which are enumerated here: (a) Zero-hour phishing attempts cannot be stopped by the blacklist method since it can only identify and categorise 47-83% of newly found phishing URLs in a 12-hour timeframe

FAQs

What problem does the document target?

It targets real-time detection of malicious URLs that enable attacks such as spam, phishing, malware, and defacement, where traditional methods may fail as threats evolve.

How does the proposed hybrid model work?

It uses SOM-RMO to extract and highlight important URL features, then applies an RBFN classifier whose classification is optimized using Tabu Search.

What performance results are reported on the benchmark dataset?

The approach reports 96.5% accuracy, 95.2% precision, 94.8% recall, and 95.0% F1-score, exceeding traditional methods.

Aug 6, 2026

Upload by Ava Thompson