SMS Spam Classification using NLP

Overview

This project focuses on building a Natural Language Processing (NLP) model to classify SMS messages as Spam or Ham (Non-Spam).

The project demonstrates various text preprocessing techniques and feature extraction methods commonly used in NLP applications.

---

Problem Statement

Spam messages are unwanted messages that can affect user experience and security. The objective of this project is to build a machine learning model capable of automatically identifying whether an SMS message is spam or ham.

---

Objectives

• Perform text preprocessing.
• Clean and normalize text data.
• Convert text into numerical features.
• Train machine learning models.
• Classify messages into Spam or Ham categories.

---

NLP Techniques Used

Text Preprocessing

• Lowercase conversion
• Tokenization
• Stopword Removal
• Stemming
• Lemmatization
• Regular Expressions

Feature Engineering

• Bag of Words
• CountVectorizer
• TF-IDF Vectorizer

---

Libraries Used

• Python
• Pandas
• NumPy
• NLTK
• Scikit-Learn
• Matplotlib
• Seaborn

---

Workflow

1. Data Cleaning
2. Exploratory Data Analysis
3. Text Preprocessing
4. Tokenization
5. Stemming and Lemmatization
6. Feature Extraction
7. Model Building
8. Model Evaluation

---

Project Structure

sms-spam-classification-nlp
│
├── data/
├── sms_spam_classifier.ipynb
├── requirements.txt
├── README.md
└── images/

---

Skills Demonstrated

• Natural Language Processing
• Text Cleaning
• Tokenization
• Regular Expressions
• Stemming
• Lemmatization
• Count Vectorization
• TF-IDF Vectorization
• Machine Learning
• Feature Engineering

---

Applications

• Email Spam Filtering
• SMS Spam Detection
• Chatbots
• Sentiment Analysis
• Text Classification
• Information Retrieval

---

Future Improvements

• Word2Vec Embeddings
• GloVe Embeddings
• LSTM Models
• BERT Transformers
• Hyperparameter Tuning
• Model Deployment using Streamlit

---

Author

Deebesh Sundar

Machine Learning & Data Science Practitioner