Spam Message Detection (TF-IDF + Naive Bayes)

1) Project Title

Spam Message Detection

2) Problem Statement

The goal of this project is to automatically classify text messages into two categories:

• ham: normal, non-spam message
• spam: unwanted promotional or fraudulent message

This is a classic text classification problem. We solve it with:

• TF-IDF (Term Frequency - Inverse Document Frequency) for converting text into numbers
• Multinomial Naive Bayes for fast and effective spam classification

3) Dataset Description

This project uses a small custom CSV dataset located at:

data/spam_sample.csv

It contains two columns:

• label: target class (ham or spam)
• message: raw text message

The dataset is intentionally small and beginner-friendly so the full workflow is easy to understand.

4) Full Python Code

The complete runnable code is in:

spam_message_detection.py

It includes:

• data loading
• preprocessing with TF-IDF
• model training with Naive Bayes
• evaluation metrics
• visualization plots
• sample predictions

5) Explanation

A) Data Loading

• We use pandas.read_csv() to load the CSV file into a DataFrame.
• We print the first rows, shape, and class distribution to inspect the data.

B) Preprocessing

• Split the dataset into train/test sets using train_test_split.
• Use TfidfVectorizer to transform text into numerical feature vectors.
• Fit TF-IDF on training data only, then transform test data.

C) Model Training

• Train a MultinomialNB model on TF-IDF features.
• This model is simple, fast, and effective for bag-of-words style text data.

D) Evaluation

• Predict on test data.
• Compute:
  • Accuracy
  • Classification report (precision, recall, F1-score)
  • Confusion matrix

E) Visualization

The script saves two plots:

• class_distribution.png: bar chart of ham vs spam samples
• confusion_matrix.png: model prediction quality by class

6) How to Run

Requirements

• Python 3.10+
• Libraries:
  • numpy
  • pandas
  • matplotlib
  • scikit-learn

Install dependencies

pip install numpy pandas matplotlib scikit-learn

Run the project

python spam_message_detection.py

Expected output

• Console logs for:
  • dataset preview
  • model metrics
  • sample predictions
• Image files:
  • class_distribution.png
  • confusion_matrix.png