Inspiration
Thanksgiving is a holiday for a family to get together and bond. Unfortunately, in arguably the most polarized political climate ever in the history of this great country, Thanksgiving has become a time where you meet distant relatives who spew vitriol rhetoric on the dinner table. What was once the great American holiday has been reduced to an ideological battleground.
What happened? What radicalized and emboldened once pleasant and demure relatives to spew divisive words? Some studies have concluded that the exposure of fake and biased news via Facebook and other social media platforms have contributed to this phenomenon.
Thus the problem at hand. We cannot control which articles people click and open. However, what if we could identify articles which are potentially biased and fake, and replace, without the users knowledge, the contents of such articles. Thus, Pak! the fake news destroyer was born.
How I built it
- Interface with the Mercury Web Parser (MWP) API to obtain headline, author, body text, relevant images.
- Obtain the article body and send to Google Cloud Platform Natural Language API to obtain named entities.
- Selected top entities and used as search term to look for similar news from Associated Press from GDELT API.
- Take url of the returned article from GDELT API and submit to MWP API.
- Replace old article and inject the AP article content returned by MWP API. TADAAAA!
Challenges I ran into
- Don't know how to code
- Not very smart
- Allergic to fake news
Accomplishments that I'm proud of
- Finishing hacking
- Sleeping a little
What I learned
- Believe in yourself
- Believe in your teammates
- Red Bull gives you wings
What's next for Pak!
- Saving the world one step at a time.
Log in or sign up for Devpost to join the conversation.