Hacker News Reader

How is the clustering of these articles actually work?

Pretty interesting!

Thanks! I'm using a td-idf algorithm to rank similarity between stories. It's evolved over time and I'm constantly refining it.