🎭 Text Sentiment Analysis

A machine learning pipeline to classify IMDB movie reviews as positive or negative using NLP preprocessing, TF-IDF vectorization, and a Logistic Regression model.

---

🧠 Overview

This project builds a text sentiment analysis model using the IMDB Reviews Dataset. The pipeline involves preprocessing, vectorization, training, evaluation, and live prediction.

---

📁 Dataset

This project uses the IMDB Reviews Dataset:

➡️ Download from Kaggle

Steps:

1. Download and extract the dataset.
2. Rename or ensure the file is named IMDB_Dataset.csv.
3. Place it in the project root directory.

⚠️ The dataset is not included in the repo due to GitHub file size limits.

---

🛠️ Tech Stack

• Python 3
• Pandas
• Scikit-learn
• NLTK
• TF-IDF Vectorizer
• Logistic Regression

---

🚀 Features

• Clean and normalize text using NLP techniques
• Convert reviews into numerical features using TF-IDF
• Train and evaluate a logistic regression model
• Save trained model and vectorizer for reuse
• Predict sentiment of custom reviews in real-time

---

📊 Results

Metric	Score
Accuracy	85.13%
F1-Score	85%

---

📦 Project Structure

text-sentiment-analysis/
├── IMDB\_Dataset.csv
├── sentiment\_analysis.py
├── sentiment\_model.pkl
├── tfidf\_vectorizer.pkl
└── README.md

---

▶️ Getting Started

1. Clone the repo

   git clone https://github.com/ahsankhizar5/text-sentiment-analysis.git
   cd text-sentiment-analysis

2. Install dependencies

   pip install -r requirements.txt

3. Run the script

   python sentiment_analysis.py

4. Enter your own review for live prediction!

---

📌 Example

📝 Try your own review:
Enter a movie review: this seems to be bad one
Predicted Sentiment: Negative 😞

---

📑 License

MIT License

---

🤝 Contact

For queries or collaboration, feel free to reach out: Ahsan Khizar GitHub — LinkedIn

“Code is not just about solving problems. It’s about building trust, clarity, and real-world impact — one line at a time.”> — Ahsan Khizar