This is a simple sentiment analysis program which uses various reviews from yelp, amazon and IMDB to train and test a model before predicting if a statement is positive or negative. The user can then insert a statement which would be analysed to see if it was positive or negative.
There are two ways for the user to input their statement, either by typing it out or speaking into a microphone. If the script predicts that the statement was negative, it will ask the user if they want to relax and a breathing exercise would ensue. The idea behind this program is a simple app that analyses users mental state or his emotion at the moment, allows user to speak or type how they feel about their day and the app would be able to calm or destress the user through several means, whether it be by listening to calming music or doing breathing exercises.
- Logistic Regression
- Multinomial Naive-Bayes
Few Point About The Project
- IMDB movie reviews, Yelp reviews and reviews on the Amazon website were used as data
- Train : Test is 0.75 : 0.25
- These accuracies are then compared to see which was better at predicting one's sentiment, before narrowing it down to only one algorithm thus leading to good result in analysing the emotion of the individual and further responding on the basis of prediction
Snapshots 👇
