First, sign up on Twitter website and get a developer account
Then, use crawl.py to obtain the twitter contents of each candidates, store the data in txt files named as candidate name + ".txt"
Use clean.py to clean the data, in this file we named the output file as candidate name + "without_http.txt", since the last step in data cleaning is removing the links in twitters
Use data_visualize.py to do the plotting, and get word cloud figures as outputs. We can also use other tools such as seaborn