Re:Mozilla hit with takedown request for anti-paywall addons
by jpfed on Sunday December 02, 2018 @11:39PM
(#18586357)
>They need to publish their full article content for search crawlers to read
At first, I thought it might be possible to solve this by getting search engines to standardize on a vector format that they could accept for protected content. So the crawler sees a 300-dimensional vector that effectively gives a semantic summary of the document.
But then I thought content providers could achieve a substantially-similar effect by just serving their documents to search engines in scrambled (e.g. alphabetized) form. They could still provide normal headlines to get them clicks.
BUT THEN I thought it would be a really cool and bizarre problem to circumvent this by attempting to devise a method for finding the most-probable original document given its alphabetized version.
At first, I thought it might be possible to solve this by getting search engines to standardize on a vector format that they could accept for protected content. So the crawler sees a 300-dimensional vector that effectively gives a semantic summary of the document.
But then I thought content providers could achieve a substantially-similar effect by just serving their documents to search engines in scrambled (e.g. alphabetized) form. They could still provide normal headlines to get them clicks.
BUT THEN I thought it would be a really cool and bizarre problem to circumvent this by attempting to devise a method for finding the most-probable original document given its alphabetized version.