People are consuming more information than ever before, but with curated facebook feeds, optimized google searches and even the new twitter 'highlights' encroaching on the firehose, it is hard to avoid unintentionally ending up in an echo chamber where it seems like you are getting a broad view of the world, but you are hearing what you want to hear. Eli Pariser talked about these filter bubbles in his 2011 TED talk:

How it works

Using the Bluemix Web IDE, we wrote javascript to parse several political news rss feeds and store the URLs in a Cloudant DB. We then use AlchemyAPI's Author Extraction, Entity Extraction and Knowledge Graph to identify politically aligned people and sentiments about them from the articles. We apply a scoring algorithm to place the authors on a liberal-conservative spectrum based on their sentiment toward the identified entities and display the results using bootstrap and d3.js.

Challenges we ran into

Time was the big limiting factor for us on this project.

Accomplishments that we're proud of

It was the first hackathon for most of us, so we're glad to have accomplished as much as we did.

What we learned

Bluemix, AlchemyAPI and the Watson ecosystem were very new for us, so we learned a lot.

What's next for know your bias

We didn't quite get our aggregation by author working on the cloudant db in time, so we couldn't yet hook the data into our lovely visualizations.

We would like to hook in twitter, so users can have a more personalized experience and see the biases of the journalists they follow.

It would also be nice to turn this into a chrome extension, so users can see the 'bias' score of the journalists they are reading in real time.

Built With

Share this project: