The idea is to scrape the most recent HN article titles and use TF-IDF and cosine similarity as metrics to rank articles that you're interested in to filter out only relevant things. Hopefully a fun project for beginners to get inspiration to jump into this world.