This project encompasses the design, construction, and implementation of a sentiment analysis service for movie reviews. It demonstrates the deployment of a machine learning model as both a Function-as-a-Service (FaaS) using AWS Lambda and as a containerized microservice using Docker. A simple web interface is provided to showcase the functionality to stakeholders. The primary objective of this project was to gain hands-on experience in deploying machine learning models as scalable and efficient microservices.
- Natural Language Processing (NLP)
- AWS Lambda for serverless deployment
- Docker for containerization
- Flask as the web server framework
- Basic HTML and CSS for front-end presentation
- Model deployment and operationalization
The company's platform allows for user-generated movie reviews that directly influence the overall product ratings. The objective is to automate the evaluation of these reviews, accurately classifying them as positive or negative. This automation will significantly refine the rating process, ensuring consistency and objectivity while also accommodating the growing volume of user feedback.
-
Model Training: Developed a simple sentiment analysis model using Python's NLTK library to classify movie reviews as positive or negative. The model achieved an accuracy score of 87.75%.
-
API Endpoint: Leveraged AWS Lambda to host the model, creating a callable API endpoint for real-time sentiment analysis.
-
Model Deployment: Deployed the model within a Docker container, utilizing Flask as the web server to handle incoming requests.
-
Front-End Demonstration: Constructed a mini-website that interacts with the model, allowing stakeholders to input movie reviews and instantly view the sentiment analysis results.
- Install Docker. Check that it has installed correctly:
docker --version- Pull repository and cd into the project:
git clone https://github.com/GregoryTomy/sentiment-service.git
cd sentiment-service- Run docker compose
docker-compose up --build-
Open the address by copy-pasting the url. This is what it will look like: "Running on http://127.0.0.1:4000"
-
Enter a review and click submit.