twitter data visualization using maps and ggplot2

Although the data might appear to be biased and shows that the Asian countries are more sympathetic towards Nepal, but this does not takes into fact that out of 10,000 tweets mined, only ~300 had latitude and longitude values( only 205 had non zero values). And there’s a pretty good possibility that most of the tweets from western countries have no geographical information, because of privacy.

Information on the twitter data visualization and mining

Twitter data mining and visualization using Maps API and ggplot 2.
Keyword used to mine data: “earth quake”.
Total tweets mined = 10,000
Filtered data frame resulted into 205 observations of importance.

Tweets were mined at 2:30 AM, IST.

Click to enlarge the image

twitter data visualization of tweets

2 comments

  • Parikshit, very cool piece of research. How did you mine the tweets?
    Do you think time zone would be a good approximation of user’s location?

    • Hi Ela, I wrote some R scripts to mine tweets. Yes, In places where lat-long information wasn’t available, using time zone would be a very good approximation.

Leave a Reply

Your email address will not be published. Required fields are marked *