Machine Learning-based Email Spam Filtering and Classification System
DOI:
https://doi.org/10.62643/ijerst.v22i2.2320Keywords:
Email Spam Detection; Naive Bayes Classifier; Text Classification; TF-IDF Vectorization; Machine Learning; Gradio Web Application; Natural Language Processing; Ham ClassificationAbstract
Spam messages have become a significant challenge in digital communication, leading to security threats, unnecessary advertisements, and decreased user productivity. This paper presents a Machine Learning-based Email Spam Filtering and Classification System that automatically classifies messages as spam or non-spam using the Naive Bayes algorithm. The system is trained on a labeled dataset containing email messages, where each message is labeled as spam (1) or ham (0). The textual data undergoes preprocessing steps such as tokenization, removal of stop words, and transformation into numerical representations using TF-IDF vectorization with bigram support. The Multinomial Naive Bayes classifier is chosen due to its simplicity, computational efficiency, and proven effectiveness in text classification tasks. Class imbalance is addressed through oversampling of the minority class. After training, the model is deployed within a Gradio-based web interface, enabling users to input message content and obtain real-time predictions. The proposed system demonstrates reliable performance with an overall accuracy of 94.26% and fast prediction speed, making it suitable for practical spam detection applications
Downloads
Published
Issue
Section
License

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













