I test the code of user-timeline-tools from casmlab and collect the tweets from ten famous person, suck like Taylor Swift and Kobe Bryant, the famous singer and basketball player.
And I create another two files:
cloudWord.py tweetAmount.py
cloudWord.py will make a word cloud of the tweets, and we can find the most frequency word in the tweets are "RT" and "U". If the stopwords contain "RT", we can get a more accurate result.
tweetAmount.py will make a bar graph based on the data of tweets and fullname in the Database. We can find the relative frequency of tweets among the ten users.
I have push the graph of cloudword and tweetAmount on Github.
Conclusion
As the data that we collect getting larger, we can make a more accurate analysis and predict. We can find out who use the Twitter most and how people post a tweet with, Phone or other Social Network? I have try to combine the data to analysis when people prefer to post tweets and what kind of topic people pay more attention to on the Twitter.
parseUserTimeLine.py, there was an error in the main function when it load the data from the json file with a for-loop.- I am confused that I can not map the tweets to the twitter name when I query the database. So we can not find out which tweet is posted by the user.
