CAS | Devpost

Screenshot

Inspiration

To build a better-informed society, it is of utmost importance for us to be able to empathize with opinions of others --- especially of those with whom we disagree. At the same time, to be better informed we should also learn more from those around us who share our opinion. So, we created a service to show people both sides of arguments on something that they care about so that they may broaden their perspectives and better contribute to the world.

What it does

Users can submit their opinions (through blog posts for now) to our web service to search a repository of 1700 articles and find ones on the same topic written by people who share similar and dissimilar opinions to that of the user.

How we built it

We aggregated articles from a variety of APIs and then processed them using MS Cognitive Services. These articles were clustered based on theme and sentiment to be able to provide users similar and dissimilar opinions on the same theme. Our web service runs on the Google Cloud Platform, which facilitated the pre-computation of article clustering algorithm and hosts our server.

We also used a bootstrap powered templating engine served on Flask.

Challenges we ran into

We spent a lot of time finding an optimum scope for our problem to be able to build a proof of concept within the limited time for the hack. Otherwise, it would have been impossible to aggregate, process, cluster, and test within a limited time. So, we limited our clustering to one theme, Brexit. In addition to those, 'Microsoft COG Services' is not as precise as it should be.