I fail to see how any of this is novel. They basically describe TFIDF with a cosine vector product. This taught in undergraduate classes! What am i missing!
> I fail to see how any of this is novel. [..] this whole thing seems suspicious.
It would be suspicious if it wasn't granted, as it would mean the patent office is unusually hostile to the claimant. Granting junk patents is par for the course, and the patent office is a bad joke, the 1-click patent [1] is just one of countless examples [2]. The most obvious thing in the world is "novel" to them if they can't find a patent (they only bother looking at patents) for it in their cursory prior-art search.
[1] https://en.wikipedia.org/wiki/1-Click - patents the what, not the how. Equivalent to getting a patent on "machine that flies", and then suing anyone that actually went through the trouble of figuring out aerodynamics, or hot air balloons, or rockets.
I was explicitly told by a patent lawyer that the patent office only considers other patents as prior art. Science papers and textbooks don't count and they don't look for prior work in them.
The vast majority of software patents likely shouldn't be granted either because of prior art or because they are obvious to practitioners of the art. The boat to fix the system has long sailed, and there is little point in talking about reforming it under the new US regime.
otterdude•5mo ago
I believe you're analysis is ultimately correct, but we wont fix things by staying silent.
We should be building technology and organizing to fight this BS
josefritzishere•5mo ago
I expect this is the start of a patent troll story. Patents are supposed to be about informing the public. In reality patents are full of obfuscation and other legal gamesmanship, even as to who holds the patent. Anyone familiar with the "Unified Patents" story from 2017? https://ipwatchdog.com/2017/08/27/duplicitous-nature-unified...
otterdude•5mo ago
You're completely right, the supremacy of the legal domain over the scientific i think is ultimately to blame. We need specialized courts for technological matters instead of letting seniors who can't even open their email make these important, nearly permanent common-law precidents. Or maybe just switch over to civil law.
otterdude•5mo ago
The grantee doesn't have a CS degree, this whole thing seems suspicious. Here is the company website https://www.linkedin.com/company/bundleiq/posts/?feedView=al...
Here is a post of his claiming to have worked at Mar-A-Largo for 6yrs https://x.com/mohnacky/status/1901033511307325735
like_any_other•5mo ago
It would be suspicious if it wasn't granted, as it would mean the patent office is unusually hostile to the claimant. Granting junk patents is par for the course, and the patent office is a bad joke, the 1-click patent [1] is just one of countless examples [2]. The most obvious thing in the world is "novel" to them if they can't find a patent (they only bother looking at patents) for it in their cursory prior-art search.
[1] https://en.wikipedia.org/wiki/1-Click - patents the what, not the how. Equivalent to getting a patent on "machine that flies", and then suing anyone that actually went through the trouble of figuring out aerodynamics, or hot air balloons, or rockets.
[2] https://wiki.endsoftwarepatents.org/wiki/Example_software_pa...
tensor•5mo ago
The vast majority of software patents likely shouldn't be granted either because of prior art or because they are obvious to practitioners of the art. The boat to fix the system has long sailed, and there is little point in talking about reforming it under the new US regime.
otterdude•5mo ago
We should be building technology and organizing to fight this BS