Aim

Task is to predict the sentiment of the tweet i.e. positive, negative or neutral.

Dataset

Twitter US Airline Sentiment Dataset, which contains data for over 14000 tweets.

In this repository

• dataset.csv - Dataset used for training and testing the model.
• twitter_sentiments.ipynb - Jupyter Notebook.
• emogi_sentiment.csv - CSV file for getting the sentiment of the emogi used in tweets.

Required external libraries

• numpy
• pandas
• sklearn
• nltk

Process

• Remove the airline name.
• Remove/Replace(with sentiments) the emogi.
• Remove punctuations and stop words.
• Applying PCA.
• Building model.

Future

I hope to further increase its efficiency by training it on a larger dataset. Also planning to use different classifiers and observing the effect of emogis.