Keyword Meta-Analysis Level 2
The news is filled with too much information and bias. We wanted a getting to the heart of things by allowing people to see the underlying key terms and concepts pervading news topics of their choosing.
What it does
A user enters a topic they're interested in and the application surveys a wide range of news articles from varying (and often contrasting) news sources. It performs natural language processing on these many articles and narrows them down to 3-4 key terms for users to select and dive deeper into. We let our users keep going up to three levels of depth, and when they're ready, we pull articles and compare them in terms of overall emotional tone and summarize them for our users.
Challenges We Faced
A lot of the problem was UI/UX and deciding what would be most beneficial to people at each stage of them digging deeper into the news. There's a lot of data out there, and while the technical challenges were certainly challenging (the data-sifting and animations were particularly hard at times), the overall just not-scariness of the way we've brought a meta-analytical natural language processing news method into being was particularly tough to do.
Sourcing for freely available information was also a significant challenge. We didn't have the time to build our own huge network of scrapers and cachers which we envision the project to eventually have, and we had to look hard to find good APIs that gave us access to enough information to both prove our concept and develop our product.
Accomplishments that we're proud of
It looks great and feels so simple to use! We got a lot of natural language processing and scraping done in a short amount of time. And it works surprisingly well, giving all of us 'aha' moments and interesting results that we just would not have otherwise seen - which in itself proves our initial concepts regarding the idea.
What's next for 036 - reNsearch: Redefining Research
We want to bring our tool to research, allowing people to understand large masses of cluttered information much more quickly and effectively. We hope to empower the public through making the means of acquiring information more convenient, intuitive and productive.