jkl.io is a customisable user-centric news aggregator. We are pursuing a user-supported model to avoid "optimising for engagement".
We draw from diverse quality sources, public broadcasters, universities, think tanks, traditional media, as well as the web.
Our default algorithm favours serious news are there are controls that allow you to tune how scores decline over time, the clickbait penalty, and how much of a boost to give to coverage.
You can customise the interface: themes, fonts, columns, section order and size.
The optional personalisation algorithm runs completely locally on your computer. Easy to interpret scores, out of 100, are paired with a dedicated section for model examination.
There is no user tracking, no analytics, no user accounts, no cookies, no cookie popups, and no server-based data storage or personalisation.
The defaults are based on and optimised for Canberra, Australia.