An End to End Python Web App using Flask to Predict Emotions with Natural Language Processing
I want to create an app where user can write his/her diary and analyze thier emotions with the help of sentiment analysis algorithm.
-I have developed emotion detection model using TFIDF vectorization method . used word2vec and doc 2 vec too but got more accuracy with TFIDF vectorizer.
TFIDF vectorizer is easy to use method https://towardsdatascience.com/natural-language-processing-feature-engineering-using-tf-idf-e8b9d00e7e76
Removed stop words using NLTK and cleaned text before vectorizing the text. Used Random forest classification as it is good at large data sets.
I’ve used Flask to develop the web app. And chose SQLITE as database connecting method as I’m restricting maximum no of users . Created Two tables login and diary for data storage and created triggers to keep both tables in sync
I’ve created js file along with html to run various functions within the page .
Below figure is the working design of this app
When user enters the home page of the app, they can go to different links(home.html)
User can go to login page to either login or register and login to the profile page . Here login.html calls login python func (app.py)for authentication
User can add , edit , delete notes in their diary and app creates the feelings pie chart Updates the user data through app.py in database.db
Check out my app at https://know-thyself.onrender.com/