Sentiment Analysis on Amazon Review Dataset using MLOps

This project showcases an end-to-end MLOps pipeline for sentiment analysis using the Amazon Review dataset.

Dataset Overview

Source: Amazon Review Dataset
Version: 3 (Updated 09/09/2015)

Origin

The dataset contains Amazon reviews spanning 18 years, including ~35 million reviews up to March 2013. Each review includes product information, user details, ratings, and plaintext review content.

For more details, refer to:

• J. McAuley and J. Leskovec, RecSys 2013: Hidden factors and hidden topics: understanding rating dimensions with review text.

The polarity dataset for this project is constructed by Xiang Zhang as a benchmark for text classification. For more details, refer to:

• Xiang Zhang, Junbo Zhao, Yann LeCun, NIPS 2015: Character-level Convolutional Networks for Text Classification.

Dataset Details

• Reviews are classified as positive (scores 4 & 5) or negative (scores 1 & 2). Neutral reviews (score 3) are excluded.
• Each class contains:
  • 1,800,000 training samples
  • 200,000 testing samples

Files:

• train.csv and test.csv (comma-separated values with the following columns):
  • Class index (1 = Negative, 2 = Positive)
  • Review title
  • Review text

Special Formatting:

• Review title and text are enclosed in double quotes ("), and internal double quotes are escaped ("").
• Newlines are represented as \n.

Deployment

• Docker Image: aswaths/sentimentanalysis
• Streamlit App: Streamlit Sentiment Analysis App

Instructions

1. Prepare the Dataset

   • Download the dataset and place it in the data folder.

2. Run the FastAPI Server

   • Execute serve.py to launch the FastAPI server.
   • Access API documentation at http://localhost:8000/docs.

3. Make Predictions

   • Use the following curl command to fetch predictions:

     curl -X POST "http://localhost:8000/predict" -H "Content-Type: application/json" -d '{"query": "This is a great product!"}'

4. Build and Run Docker Image

   • To build the Docker image:

     docker build -t <image_name> .
     docker run <image_name>

   • Modify the Dockerfile as needed.

5. Push to DockerHub

   • Create a DockerHub repository, then upload your image:

     docker login
     docker tag <image_name> <username>/<repo>:<tag>
     docker push <username>/<repo>:<tag>

6. Monitor Models with MLflow

   • Access the MLflow UI to track and monitor model performance.