Our inspiration was that we wanted to build something that we would want use, and could use immediately after the event ended. The two core principles behind Verify are that we want information fast, and we want to know if it is biased. This means we don't want to leave our browser, don't want to scroll over ads and don't want to lookup every site we go to. We knew that if we could create a project that could fit these parameters, we would (and others as well) would use it daily.
What it does
Verify is a chrome extension that analysis the current article you are looking at and produces the sites political leaning/bias, article sentiment and a concise summary of the article, only a click away. Previously analyzed site results are cached in chrome to save precious seconds when you visit again (clearable in settings).
How we built it
Our back end was a Node server deployed via Standard Library that integrated with the Google Learning API, Article Outliner API from Diffbot.com, a NLP summarizer API from DeepAI.org and https://mediabiasfactcheck.com/ for the bias analysis by scraping their website (please add an API mediabiasfactcheck!).
The saving grace of our project was Standard Library. the smart code editor, automated package imports and installations, and one click deploy saved hours in creating and hosting our analysis endpoint and allowed us to reach our stretch goal of sentiment analysis.
Challenges we ran into
The biggest challenge we faced was finding a good source for checking website bias. No API's were available, but we found a great, reputable website with thousands of entries so we ended up scraping that site for our results. The usual challenge of setting up and deploying a server without being forced to use certain libraries was removed by Standard Library's no hassle deployment
Accomplishments that we're proud of
We are proud that our project has immediate, practical uses, and that we know we will use this project for months or years to come.
What we learned
We learned the power of good prototyping tools (Standard Library) and gained more experience with Google Cloud Services and our first look into Chrome Extension development
What's next for Verify
We plan on using our project ourselves. We will look into a wide release of the extension. As the server costs and API calls are only pennies per user per month, revenue models could include advertisement based, subscription based, or based on a small up front fee (or based on burning VC funding). We can also add quality of life features such as the full article in an easily copyable format and connectivity features.