The Twitter airline dataset is a collection of comments posted by Twitter users about their travel experiences with different airlines. The dataset is composed of a total of 5,313 tweets that were gathered over a period of six months.

Each tweet has been labelled as either positive or negative based on the sentiment of the tweet towards the airline.

The data has been split into training (3,963 rows) and testing (1,350 rows) datasets
airline_tweets_train.csv - The training set, we will use this dataset to train our models
airline_tweets_test.csv - The test set, we will evaluate and rank teams with this dataset
Columns
Id - Numeric Id of the tweet
airline_sentiment - Sentiment of the tweet, 'positive' 'negative' or 'neutral'
airline - Name of the airline referenced in the tweet
text - Content of the tweet
retweet_count - Number of retweets performed on tweet
user_timezone - Timezone of the user posting tweet