Skip to content

README.md

Latest commit

 

History

History
81 lines (55 loc) · 1.93 KB

README.md

File metadata and controls

81 lines (55 loc) · 1.93 KB
Raw

Sentiment Analysis on Restaurant Reviews

This project performs Natural Language Processing (NLP) to classify restaurant reviews as positive or negative using a Naive Bayes classifier.

🧠 Model Overview

  • Goal: Predict sentiment (positive/negative) of restaurant reviews.
  • Dataset: 1,000 reviews from Restaurant_Reviews.tsv.

✨ Features

  • Text cleaning and preprocessing pipeline.
  • Bag of Words model with feature limitation.
  • Sentiment prediction using Gaussian Naive Bayes.
  • Performance evaluation using accuracy and confusion matrix.

🧹 Data Preprocessing

  • Removed non-alphabetic characters.
  • Converted to lowercase.
  • Tokenized into words.
  • Removed stopwords (excluding "not").
  • Applied Porter Stemming.
  • Reconstructed into cleaned text corpus.

🧰 Feature Extraction

  • Used Bag of Words model with CountVectorizer.
  • Limited features to top 1500 words.

🤖 Modeling

  • Classifier: Gaussian Naive Bayes
  • Train/Test Split: 80/20

📊 Evaluation

  • Confusion Matrix and Accuracy Score used to assess performance.

🛠 Installation

  1. Clone this repository:
git clone https://github.com/MaddyRizvi/Natural-Language-Processing_sentiment_analysis.git
cd Natural-Language-Processing_sentiment_analysis
  1. Install dependencies:
pip install -r requirements.txt

🚀 Usage

  1. Ensure you have the dataset Restaurant_Reviews.tsv in the project directory.
  2. Run the script:
python natural_language_processing.py

📁 Files

  • Restaurant_Reviews.tsv: Dataset file.
  • natural_language_processing.py: Main script for training and testing.
  • README.md: Project overview and instructions (this file).
  • CONTRIBUTING.md: Guidelines for contributing.

✅ Requirements

  • numpy
  • pandas
  • matplotlib
  • scikit-learn
  • nltk

🙌 Contributing

We welcome contributions! Please read the CONTRIBUTING.md file for guidelines.