Skip to content

Latest commit

 

History

History
45 lines (31 loc) · 1.43 KB

README.md

File metadata and controls

45 lines (31 loc) · 1.43 KB
Raw

Movie Review Sentiment Analysis

Notebook
Flask App

Description

This Flask app deploys a Naive Bayes classifier model built from scratch, to Docker to be used for instant-classification of inputted reviews.

A model working in a Jupyter notebook is only 10% of the battle when it comes to full-stack Data Science & MLOps. I deploy a model e2e.

Goals

  • ML Model deployment e2e
  • Docker Deployment
  • Flask app deployment
  • Software Engineer a usable, interactive app beyond just deployment and classification
    • Interactive animations based on classified sentiment
    • Sleek, attractive design
  • Get hands-on experience applying a Naive Bayes classifier by making it from scratch

Demo

positive

positive.png

negative (sad)

negative-sad.png

negative (angry)

negative-angry.png

Results

With a balanced dataset of exactly 50% positive and 50% negative class (feeds into the Naive prior probability) and a custom NB model, we get weak results. Of course major improvements can be made by leveraging more powerful models that don't have the inherent naivety of prior probabilities.

image.png

References:

Face Emoji with Emotion https://codepen.io/benjyring/pen/YmPRZx

Contact Form UI https://codepen.io/krisantuswanandi/pen/KxrgeZ