Abstract :

The Continuous Growth Of Computer Networks Has Resulted In An Enormous Volume Of Communication Data, Making Manual Inspection Of Network Traffic Increasingly Difficult. Accurate Identification Of Network Protocols Is Essential For Network Monitoring, Security Analysis, Traffic Management, And Intrusion Detection. This Study Presents A Machine Learning-based Framework For Automated Network Traffic Classification By Analyzing Communication Characteristics Extracted From Packet Data. The Proposed Approach Performs Multiple Preprocessing Operations, Including Missing Value Treatment, Categorical Feature Encoding, Normalization, Randomization, And Dataset Partitioning To Prepare Reliable Input For Model Development. Two Supervised Learning Algorithms, Namely K-Nearest Neighbors (KNN) And Naïve Bayes, Are Implemented And Evaluated To Classify Different Network Protocol Types. The Developed Application Provides An Interactive Environment For Dataset Upload, Preprocessing, Model Training, Comparative Performance Analysis, And Protocol Prediction Using Previously Unseen Traffic Records. Experimental Evaluation Demonstrates That The KNN Classifier Achieves Superior Classification Accuracy And Overall Predictive Performance Compared With The Naïve Bayes Model Across Multiple Evaluation Metrics, Including Precision, Recall, F1-score, And Confusion Matrix Analysis. The Proposed Framework Enables Efficient And Consistent Protocol Identification While Reducing The Effort Associated With Manual Traffic Examination. The Developed Solution Can Support Network Administrators And Security Analysts In Monitoring Communication Patterns, Improving Traffic Analysis, And Enhancing Decision-making For Modern Network Management Environments.

Published:

30-7-2026

Issue:

Vol. 26 No. 7 (2026)

Page Nos:

1349-1354

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite