Abstract :

With The Rapid Growth Of Online Educational Platforms, Organizing And Categorizing Learning Content Such As Images, Diagrams, Text, And Headers Has Become Essential For Improving Content Accessibility And User Experience. Manual Classification Of Educational Images Is Time-consuming And Inefficient, Especially With Large Datasets. This Project Proposes A Multilabel Image Classification System Using Convolutional Neural Networks (CNN) Optimized With Stochastic Gradient Descent (SGD) For Accurately Classifying Learning Images Into Multiple Categories. The Proposed System Utilizes A Deep Learning-based CNN Architecture Combined With SGD Optimizer, Configured With A Learning Rate Of 0.1 And Momentum Of 0.9, Along With Binary Crossentropy As The Loss Function. These Optimizations Enhance The Model’s Learning Capability And Improve Classification Performance. The Dataset Is Preprocessed Through Normalization And Splitting Into Training And Testing Sets In An 80:20 Ratio. The Trained Model Is Then Used To Classify Images Into Categories Such As Diagrams, Text, Headers, And Mathematical Content. Experimental Results Demonstrate That The Proposed CNN-SGD Model Achieves An Accuracy Of Over 96%, With Strong Performance Across Evaluation Metrics Including Precision, Recall, And F1-score. Visualization Tools Such As Confusion Matrices And Training Accuracy Graphs Further Validate The Model’s Effectiveness. This System Provides An Efficient And Scalable Solution For Automatic Content Classification In E-learning Platforms, Enhancing Content Organization And User Experience.

Published:

08-4-2026

Issue:

Vol. 26 No. 4 (2026)

Page Nos:

1885-1891

Section:

Articles

License:

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

How to Cite