A sentiment analysis app that uses natural language processing to categorize tweets into two categories - positive sentiment and negative sentiment.
It is a web application, the frontend is built using HTML, CSS, and JavaScript whereas for backend I've used Flask, Python. CSS animations library is also implemented for adding animations to buttons.
This app uses Twitter's API to fetch tweets from Twitter.
The algorithm used for sentiment analysis used is Naive Bayes algorithm. I have also compared two different libraries for Naive Bayes algorithm in python viz nltk and scikit learn.
As per my observation, I found that scikit learn is much better in terms of training and prediction. I was able to train on 20,000 data while uding the nltk library whereas I trained on about 1.6 million data using the scikit learn library. The accuracy of the nltk model is about 65 % while the cikit learn model gives about 75% of accuracy.