Twitter-US-Airline-Sentiment

In this project, we will analyze the dataset from Kaggle (https://www.kaggle.com/crowdflower/twitter-airline-sentiment) which contains 14,640 tweets posted in February 2015. These tweets are the passenger comments after their flight with different airline companies. This dataset is slightly reformatted and already contains a sentiment analysis which divides the comments into positive, neutral, and nagative. Since the detail of classifying the tweets is absent, we will perform our own sentiment analysis using TextBlob after the Exploratory Data Analysis (EDA). After EDA, our next step is to create our own sentiment analysis. We extract the polarity of each tweet from the result of TextBlob as our new sentiment and further compare our analysis with the original one. We also create wordclouds for the negative and positive comments. In the last part of our project, we build the text classification model using naive bayes classifier.