SMS Spam Classifier

Dataset on kaggle

Contents

Data cleaning

• Drop null columns
• Drop duplicates
• Label encoding
• Drop NaN values

Data understanding and EDA

• Pie chart
• Bar chart
• Histogram
• Pair plot
• Correlation matrix

Data preprocessing

• To lower case
• Tokenization
• Removing special characters
• Removing stopwords and punctuation
• Stemming

Modeling

• Single classifiers
• Voting classifier

Evaluation

Here we choose the best model from the following measures

• Accuracy
• Precision
• F1 score

Deployment

Deploy our model on streamlit

---

Sample