This is a project to predict the sentiment of a users based on Twitter US Airline Sentiment.The dataset can be downloaded from kaggle link.
- Download the data set from kaggle
- Choose attributes containg text and sentiment
- Map the sentiment as -1 for negative , +1 for positive and 0 for neutral
- Show sentiment of users for each airline using bar graph
- Create bag of words using count vectorizer
- Use Random Forest Classifier for classification
- Take a string as user input and check its polarity for implementation purpose.
Following image is a screenshot of a bar graph of polarity on a user defined sentance.
The project can also be seen on kaggle link.