IMDB Movie Review Sentiment Analysis

A Streamlit application that leverages a fine-tuned BERT model to analyze the sentiment of IMDB movie reviews. The app not only classifies reviews as Positive or Negative, but also provides detailed metrics such as class probabilities and token count.

Table of Contents

• Introduction
• Features
• Installation
• Usage

Introduction

IMDB Movie Review Sentiment Analysis is an intuitive web app built using Streamlit and TensorFlow with a fine-tuned BERT model. The application allows users to input a movie review and instantly obtain a sentiment prediction along with detailed performance metrics.

Features

• Real-Time Sentiment Prediction: Classify reviews into positive or negative sentiments using a fine-tuned BERT model.
• Detailed Metrics: In addition to the overall prediction, the app displays:
  • Overall Confidence: Confidence level of the predicted sentiment.
  • Class Probabilities: Individual probabilities for the positive and negative classes.
  • Token Count: The number of tokens present in the review.
• Modern UI: Attractive and responsive interface with an external CSS file for unique styling.
• Multi-Page Navigation: Easy navigation between the sentiment analysis page and an informative "About" page.
• Sample Reviews: Preloaded sample reviews to quickly test and demonstrate the app's functionality.

Installation

Prerequisites

• Python 3.9+
• pip

Steps

1. Clone the repository:

   git clone https://github.com/yourusername/imdb-sentiment-analysis.git
   cd imdb-sentiment-analysis
   

2. Set Up a Virtual Environment

   python -m venv venv
   source venv/bin/activate
   

3. Install Required Packages:

   pip install -r requirements.txt
   

Usage

 streamlit run app.py