Google working on Knowledge-Based Trust, anti-fake news algorithm: how it works?

Time for updates at Google, strictly linked to the search results proposed by the engine.
It is learned that the staff who work every day to propose the best search results for users’ queries are working on a new algorithm.
It seems that the latter is about to be rolled out soon, in order to provide increasingly relevant and, above all, increasingly accurate answers.
It would be an “anti-hoax” algorithm, capable of evaluating the reliability of the information and notions contained in a specific page and – more generally – in a site.
If up to now good ranking in SERP has been based mostly on a series of factors, among which the quantity of incoming links and the number of shares of a given page stood out, in the future Google may use new parameters to update search results.
The war on clickbait sites has already begun and not just by Google: on Facebook too, an attempt has been made to introduce a tool to limit the visibility of all those pages that, using sensational or “catchy” titles, induce users to seek more information, thus entering a page that in fact contains no valuable news.
Thus, in the Google labs, the algorithm named “Knowledge-Based Trust” is being developed: it will be based on the Knowledge Vault, a database containing an incredibly vast amount of data.
In this enormous virtual archive, sites and pages considered reliable sources by Google have been cataloged over the years.

The new algorithm will therefore cross-reference the information reported on a given page with this constantly updated archive, in order to estimate the veracity of the site being “analyzed.”
The fewer wrong information pieces there seem to be, the more a site’s credibility will automatically increase; this will complement the list of other parameters on which SERP positioning is based, thus providing a new qualitative evaluation criterion and not just quantitative (based solely and exclusively on incoming links).
A significant paradigm shift that certainly could not escape even gross errors.
Its application, however, is yet to be evaluated.