Modern politics is incredibly complex and nuanced, and fundamental rights such as privacy are being brought to question every day. Unfortunately, many people only receive one side of the story by reading biased news sources.
What it does
Wüf analyzes the contents of whatever article you may be reading, and offers suggestions for supplementary articles. It sources from a wide variety of news publishers, and although each one may be biased, the variety of viewpoints offered makes it easier to get the whole story.
How we built it
Challenges we ran into
We wished to categorize topics into 4 main groups based off of sentiment analysis: positive, negative, neutral, and controversial. While the first 3 categories were fairly trivial to implement, detecting controversy was anything but! After some 3 A.M. multivariable calculus, we had finally created a satisfactory algorithm.
Accomplishments that we're proud of
The afore mentioned controversy detection algorithm was a very fun problem to solve, however learning technologies such as Redis and Node.js was also rewarding in its own right.
What we learned
Text analysis can be challenging, but certainly isn't rocket science. Separate functions are also invaluable when working with Node.js to avoid callback hell.
What's next for Wüf
If development of Wüf continues, our goal will be to further improve our algorithms and provide even more useful data. The user interface could also use a bit of cleanup.