Deleted Tweets: taking back what was said

For my master’s project I am looking at deleted Tweets. More specifically, I am catching tweets that have been marked deleted by their sender and visualizing how they connect to the user themselves.

To make this all possible I had to make some software that gathers and stores a high volume of tweet data for real-time processing and display. That part forms the technical side of my multimedia engineering degree but how I choose to present the data will be the more challenging conceptual portion of my project.

tw_recallNow that I have the data gathering happening I made a quick visualization prototype that shows how long each tweet was free in the wild before it was deleted (X-axis) and how many people likely could have seen the tweet (Y-axis), based on that user’s follower count. Thus the triangle acts as a sort of megaphone, showing how the tweet’s impact is amplified the longer it is published and the more people are registered as followers.

As the visualization runs, more tweets are added which helps see trends over time. This may not end up in the final project but it was interesting to see the deleted tweets stack up.

posted on Thursday, April 22nd, 2010 by Pehr in MAT Projects