Now, more than ever, it is imperative to stay engaged with the world around us. One potential factor that detracts students and others with busy lifestyles from reading trusted news sites is that the news that is presented first isn't what interests them.

Hermès was created to address this exact issue. Powered by a dynamic News Delivery system, Hermès encourages a greater engagement with the news by creating personalized web pages for each user, so that they only see news about topics they care about. Use Hermès to gain a greater understanding of the world around you, all through the lens of your own interests.

We strive for collective intelligence in information. Recent advancements in Machine Learning and Unsupervised Learning have made it possible to create a collective system, using each user's individual experiences with Hermès, that is able to make news suggestions based on user-inputted preferences. This is largely powered by DBSCAN, or Density Based Clustering, an Unsupervised Learning algorithm that, based on a density parameter, groups users into clusters on the basis of their respective interests.

Visually, these clusters of our users can be represented using t-distributed Schochastic Neighbor Embedding: