Thursday, July 16, 2009

Prior Art Combinator: A tool to preemptively invalidate troll patents.

As I understand it, in order to invalidate a claim in a patent, all the elements of the claim must appear together in the same published article. From what I have read, blog posts count as well. I have not been able to determine if all the concepts listed together in a book would count. I imagine that the Patent and Trademark Office (PTO) would require that the book specifically mention that all the elements listed in the book actually belong to the same invention.

Also, as I understand it, a patent claim may still be considered valid if it lists a limitation that is not listed in the prior art. What I can't figure out so far is: If the prior art lists more elements than the patent claim, does that invalidate the claim or can the very act of not listing elements be considered a limitation within the claim. In effect, allowing patent trolls to simply patent anything and everything they find in prior art simply by removing one element from the prior art and calling that their claim. This does not seem reasonable but then very little in patent law these days does.

Finally, I have not seen anything that says the prior art publication had to be written by a human.

So I have an invention to prevent the patenting of inventions. While that may sound onerous, I suspect that it will actually alleviate a lot of the trouble we have been having with patent trolls and return patents to the realm of real innovation.

I propose that we create a database of all possible discrete elements and limitations that have appeared in all existing patents and any other documentation anywhere. Heck, people could even add any additional elements or limitations as they see fit. Then use a simple computer program to generate and "publish" articles with every possible combination of element and limitation. I know, we are talking about trillions and trillions of articles. The program could never finish.

Therefore, I also propose the following additional, optional features:

  • Users of the system could vote on which would be the most important elements and limitations to make sure are covered in our "articles."
  • Users could add tags to elements indicating which fields they are most likely to be used in. This would reduce the number of "articles" about "Umbrellas with USB connectors sewn into the seats of their pants for control and monitoring of garage doors."
  • Users could then also vote on which tags are the most important to focus on.

I realize that this would likely require more web document storage space than is currently indexed by Google. Therefore, I also propose the following additional options for storing and "publishing" the "articles":

  • The web pages could actually be automatically generated by PHP or JSP scripts using variables to represent and insert the particular combination of elements and limitations specified for that "article." Each element and limitation in the database would necessarily have a unique code number or index key. The body of the web page could then simply list the codes as parameters and call the function to build the page. The script would then simply insert the appropriate text from the database in the appropriate location in the web page and send it out. Presto, we have a web page listing a specific set of elements and limitations. As far as the browser is concerned, the web page is static and has been "published."
  • It would also be possible to just use HTML server side includes and generate static web pages with various combinations of elements and limitations included. Naturally, this would require that all the elements and limitations be in separate .shtml files on the web server.
  • Here is an even more efficient, if ethereal proposition: Simply create one web page with some PHP or JSP in it to generate a page based on the codes given as parameters, just as before. However, the codes for the elements and limitations would actually be part of the URL and sent to the server in an HTML GET request. The script then takes the codes specified in the URL and automatically generates whatever combination of elements and limitations are specified in the URL. We get an actual web page sent back even though one never really existed before. Simply by posting that one web page and putting the elements and limitations in the database, we have statistically published all possible combinations at once. Anyone who wants to view an article with a particular combination of elements and limitations simply constructs the proper URL and they can see that the page is "there." It's like quantum computing without all those pesky sub-atomic particles. Like Schrodinger's Cat, every combination would "exist" simultaneously but only by observing it does it then become, well, observable. For prior art dating purposes, I suppose we would also have to list when each element and limitation was added to the database. Only if all the elements predate the offending patent would the "Schrodinger's Article" in question apply. Now, I don't know if the PTO would go for this reasoning but it would be worth a try. And it would certainly be fun.

Finally, assuming that the PTO would not fall for "Schrodinger's Article" idea, and that we might not be able to store all of these different combinations of elements on one server, we could spread the love around by providing a web site where users can generate source code for "articles" using any of the dynamic or static methods described above, then allow users to copy and paste the HTML code for that particular combination onto a page on their own web site or blog. People all over the world could set aside some spare space on their web server to "publish" a bunch of these articles. The database could remain on one central server (with read-only access allowed by all) so that volunteers would only need to devote the storage space for the codes rather than the entire elements. Volunteers who cannot use PHP on their site can download a static version of the article.

Unfortunately, I do not have the web-design or coding skills necessary to create the proposed database or web site. I don't know if anyone else has thought of this before. Hopefully, it hasn't been patented. Therefore, as far as I am legally able and as long as the idea doesn't already "belong" to someone else, I am hereby authorizing anyone to use this idea to build such a site. The text of the article is released under the following Creative Commons license:
Creative Commons License
Prior Art Combinator: A tool to preemptively invalidate troll patents. by Grant Sheridan Robertson is licensed under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 United States License.
Permissions beyond the scope of this license may be available at

No comments:

Post a Comment