Please support us on Show HN today: Personalized Hackernews front page (using Astro, Svelte)

Smart News

Welcome to Personalized Hacker News! This website offers a custom Hacker News frontend that uses AI to optimize article rankings based on your personal interests. Enjoy a tailored news feed that highlights the most relevant and engaging content for you, ensuring you never miss out on the topics you care about the most.

June 2024: In the first release we are using a simple keyword-based approach. Each article is tagged with a set of keywords. The keywords are calculated from the title with the llama3 model.

July 2024: We are identifying topics in the hackernews post titles and show a top ranking of the most popular topics. Also we calculate related topics using the cf4j recommender library.

July 10th 2024: Added a personal history tab, the history is stored in the browser only. Clicking on links will automatically add the article to the history.

Roadmap: We are working on a more advanced approach also taking the content of the URL and the comments into account. Also allow filtering by keywords and keeping a personal keyword profile to recommend stories.

Tech Stack

Links & Attribution