-
Frontend of website
-
Analysis of reviews using NLTK and visualized with matplotlib
Inspiration
Challenge 1: Enable companies and organizations to have the ability to understand customer/user sentiments and thoughts quickly
What it does
sentimentalReviews is a website, in which users can leave reviews for a business.
Also accompanying is a mobile app with similar functions.
The reviews are then stored in a sqLite database, in which they are classified as "positive", "negative", and "neutral" according to a classifier provided by NLTK.
The data is then visualized using matPlotLib.
How we built it
We used python with the Django framework for the backend. We used html/css/bootstrap for the frontend. We used matPlotLib and NLTK to analyze and visualize the user reviews.
Challenges we ran into
-accessing the database in dJango models was actually pretty hard: took a lot of googling to figure out -learning about form processing in Django was really difficult
Accomplishments that we're proud of
Built a fully functioning website: with both a frontend and a backend. First time accomplished by all members.
What we learned
What's next for sentimentalReviews
Right now, we're using a pre-trained classifier that was trained using Twitter data. Ideally, if there was enough time, we could manually classify restaurant reviews, and train our own neural net for more accurate classification.
Log in or sign up for Devpost to join the conversation.