More on on All Prior Art database of Alexander Reben; the USPTO responds (sort of)

Further to an earlier IPBiz post Supreme patent silliness reported by Jamie Condliffe at Gizmodo: a (randomized) prior art generator , Laura Sydell at NPR has a post on All Prior Art of Alexander Reben.

Within Sydell's text:


If someone tries to patent an invention and the idea has already been published, the U.S. Patent and Trademark Office won't grant them a patent. Reben's project is trying to cut down on the number of frivolous patents by publishing lots and lots of ideas, and he thinks his database could be used by patent examiners to challenge frivolous patents.

An explanation on Reben's site says the system works "by pulling text from the entire database of U.S.-issued and published (unapproved) patents and creating prior art from the patent language. While most inventions generated will be nonsensical, the cost to create and publish millions of ideas is nearly zero — which allows for a higher probability of possible valid prior art."

Reben says his program generates about 36,000 new ideas a minute. But, it's a bit more than pure random words.



It is doubtful that Reben's approach will be a source of prior art useful to the patent office. As a first point, Reben's "art" is better characterized as "combinations of old ideas" rather than a source of "new ideas."

Along this line, because, by definition, it creates only combinations of elements, each of which is KNOWN in the prior art, it is unlikely to be of help in obviousness rejections, which require a motivation to combine
[e.g., MPEP 2143.01 ]. It is unlikely that the USPTO will find much motivation for one of ordinary skill in a database which randomly combined fragments of known art. Does the artificial act of randomly combining into "one document" overcome this? Likely not.

Separately, in the off-chance that Reben's database created "one document" which has all the elements of a patent claim (arguably compressing an obviousness issue into an anticipation issue), would the USPTO accept this as "prior art"? An anticipatory prior art patent document is presumptively enabled, but do combinations of elements from many patent documents enjoy this? Likely not. And further: the prior art needs to enable the portions of its disclosure alleged to anticipate the claimed invention. [Antor Media (CAFC 2012)]

Sydell's post also included text of a response from the USPTO


Asked whether the agency would actually look at the All Prior Art site to help decide whether to grant a patent, USPTO Press Secretary Paul Fucito emailed this statement:

"High-quality patents enable certainty and clarity of rights, which fuels innovation and reduces needless litigation. To ensure we continue issuing high-quality patents well into the future, we established the Enhanced Patent Quality Initiative (EPQI). We are strengthening work products, processes, services, and how we measure patent quality at all stages of the patent process.

"In terms of wider access to prior art, the USPTO has a Third-Party Preissuance Submission program which provides a mechanism for third parties to submit patents, published patent applications, or other printed publications of potential relevance to the examination of a patent application with a concise description of the asserted relevance of each document submitted."




link to Sydell article: http://www.npr.org/sections/alltechconsidered/2016/04/12/473880115/can-silly-patents-help-fight-frivolous-lawsuits

**Comments posted

http://www.infoworld.com/article/3060284/intellectual-property/how-do-you-stop-patent-trolls-this-algorithm-just-might-do-the-trick.html

http://www.pcworld.com/article/3060106/how-do-you-stop-patent-trolls-this-algorithm-just-might-do-the-trick.html#comments