Twitter US Airline Sentiment

About

Analyze how travelers in February 2015 expressed their feelings on Twitter

Data

Download the *.zip file from here.

https://www.kaggle.com/crowdflower/twitter-airline-sentiment

Only use the airline_sentiment and text columns during the analysis.

Part I

Build a KNN Model.
Split the data into train and test.
Use the similarity score to find the nearest neighbor. Remember to adjust k to optimize your model.
Use accuracy as your primary metric.

Part II

Cluster the tweets into 3 groups, using KMeans.
Perform PCA and lower the dimensionality of the clusters to a 2-d representation.
Graph the clusters as a scatter plot.