Likes, Retweets, and Karma: News in the 21st Century

Likes, Retweets, and Karma: News in the 21st Century


For my final project for ICM I was inspired by the data week, and decided to tackle something in the big data space. I wanted to explore how news article trickle down from the Associated Press to various news outlets and then to social media and then to the blogs and forums. How does an informed, fact-checked, detailed report turn into petty arguments about conspiracy theories? Could this be traced? Could the rabbit hole of information on the web be visualized or simulated?


In a unrelated project earlier in the semester, I used Python to scrap data from an API and I loved learning a new program and it seemed to me Python was made for data scraping. I thought it would be an interesting technical challenge to combine Python with Processing and use each language for its strength, Python would parse the data and Processing would visualize it.

Play Testing

In my initial version I had a user select an article headline from the New York Times and then I had one screen after another show data pertaining to the article from different sources. I thought it allowed the user to create their own rabbit hole into the madness that is the internet. However, after user testing, I realized that most people did not agree. A lot of my classmates commented on how it was confusing that the initial screen required user interaction and then you were suppose to be hands off and just sit back. Initially, I was very resistant to their comments and wanted to keep moving ahead with my original idea, but part of my education at ITP is to realize that great things that are not easily understood by your intended audience really aren’t that great.

Final Version

For my second iteration (and final version) I decided to give control back to the user. Not only could the user select an article from the New York Times, but they could select which source to view the article, and even select a new story. I received much more positive feedback with this version although I think it has much more room to grow. I’m excited to play with the next several iterations in the future.

Here’s a link to the github repo: (You will need to create a folder with files containing your own API keys to get this to work).


And here’s a video of the final version:  (NOTE – between now and my ICM final presentation Facebook has deprecated public search feature so I was unable to show that in the video)