Our project helps curious users find the real source behind that clickbait title, spun news article, or interesting fact from an article.
What it does
NewsSauce takes a website url and find the original source of the information used to make the article. So whether it be an interview, or picture, or even a tweet, NewsSauce will find the original
How we built it
We built it using a python article library that analyzes the library and returns vital information. We used a depth first search to look through links of links of links to articles and possible sources. We optimized the searching by using buzzword filtering to cut off the rather expensive (O^N) program.
Challenges we ran into
Our program didn't always know when the source ended and/or if a source or article was completely related to the original. To overcome this, we plan to use H.I.T to teach a computer how to know what keywords and phrases to click on to find original content, and to train a computer to find relevancy between articles.
Accomplishments that we're proud of
We are proud that we got a working program that is up and running on a website in less than 36 hours. We are also proud that this is a large first step into a program that could end false information reporting and misinterpretation of sources.
What we learned
We learned that we couldn't artificially teach the computer what to look for through code, but that we needed to organically teach the computer through H.I.T
What's next for NewsSauce
We plan to use AWS to help us with H.I.T and allowing humans to teach a computer how to determine relevancy and how to find a source.