Phone Review Sentiment Analysis

The Opening Page for the webapp!
The landing page when a review is negative.
The landing page when a review is positive.
The landing page when a review is neutral.

Inspiration

I wanted to be build a project to sense the tone of text! Given that product reviews are important for all businesses, I thought of doing sentiment analysis on popular products, then my hands fill on a 400,000 phone reviews data set and there was that.

What it does

** The Machine Learning Model**

** The Webapp **

The Webapp takes in the saved SVM model and the final vocabulary. It vectorizes the review entered using the vocabulary provided, then it uses the SVM model to predict the sentiment. Once the sentiment is detected, the webapp directs the user to webpage showing a message that tells the user whether it appears that they like their phone or not.

How I built it

I built the webapp using flask. Following the first few tutorials in this playlist: https://www.youtube.com/watch?v=4mrq8PYQ0Zg&list=PLzWaSuz38oGvFklb9kUCjT2FGS95chHcz !

Challenges I ran into

Building the webapp sucked the life out of my soul and drove me to the edge of madness! I first tried Django and CGI, both were very tedious and took lots of time without any results. I then turned to flask which was very straightforward and simple!