Skip to content
All work
Machine LearningPythonNLPML

Spam Email Classifier

A complete machine-learning pipeline, from raw email text to a spam prediction.

Source Code
SpamCheck dashboard marking a prize email as likely spam with a spam score and key indicators

Overview

A full machine-learning lifecycle in one application: clean and vectorize email text, train several models, keep the best one automatically, and classify new messages from a desktop app that stays responsive while it works.

Built with

  • Python
  • NLP
  • TF-IDF
  • Naive Bayes
  • Logistic Regression
  • SVM
  • Tkinter

What I built

01

Preprocessing

  • Tokenization
  • Stopword removal
  • Stemming
  • TF-IDF features

02

Modeling

  • Naive Bayes, Logistic Regression, and SVM
  • Automatic model selection
  • Evaluation metrics
  • Model persistence

03

Application

  • Tkinter desktop GUI
  • Threaded execution so the UI never freezes

Next project

Gitify