Wikipedia Citations in Patents Up 59 Percent

Patent Librarian Michael White recently posted an interesting graph on the number of U.S. patents referencing Wikipedia articles:

The number of U.S. patents issued last year that contain one or more references to Wikipedia articles totaled 809, a 59 percent jump from 2008. Several years ago the USPTO banned patent examiners from using Wikipedia as a source of information for determining patentability of inventions. However, examiners and applicants continue to cite it.

White's graph is shown below.

Of course, I would also offer my own graph (below) that shows the percentage of patents citing Wikipedia grouped by issue year.

Dennis Crouch

About Dennis Crouch

Law Professor at the University of Missouri School of Law. Co-director of the Center for Intellectual Property and Entrepreneurship.

33 thoughts on “Wikipedia Citations in Patents Up 59 Percent

  1. Red Monkey

    You can generate this sort of link for any version of any article – look at the options on the left hand side of the page: under “Toolbox” you find “permanent link”. Click that and you get the fixed URL in your navigation bar. As I write, link to en.wikipedia.org is the current version of the Soap article.

  2. Dave

    If when drafting you used that cite directly in the specification I think that would be helpful. How do you do that? Would you be able to generate such a link for the current page? Or would you have to wait for the page to be edited first?

  3. The 2006 article By Lorraine Woellert has its interesting pieces.

    The hyperbole of “Wikipedia has been cited in patent decisions on everything from car parts to chip designs” seems a bit outlandish, given Dennis’s graph above.

    The article does have a nice quote from John Doll about the problem of ever-changing (or at least the capability of Wiki to be ever-changing – which would include date stamps for those so maliciously inclined.

    And you always gotta love an Aharonian quote concerning the legality of anything: “From a legal point of view, a Wiki citation is toilet paper.”

    Hmm, no wonder 6 is so drawn to Wiki – must be that famoose legal logic of his seeking the appropriate gift wrap. For a bow, I’ll bet 6 uses rope.

  4. Mentioning Wikipedia in a Background section or something is a far cry from “citing” it. But then I don’t get the sense of rigorous citation discipline around here. Heck, its like a Citation Lottery…

  5. “just note the date of the page that you are citing, and keep a copy of the page in your records”

    That won’t help. First, the page can change at any time, so the granularity of 24 hours is too coarse. I suppose you could fix that by including time. But a bigger problem is that in modern content delivery systems, pages are often cached at secondary servers for load balancing. An origin server that holds the master page periodically updates the cache. Therefore, if two people ask for the same page at the same instant from two different locations, they may get different cached copies with different content.

    Printing and saving a copy gives you a snapshot to produce for litigation. But the specification is supposed to teach the public how to make and use your invention. If part of your teaching (i.e. your printed page) is locked up in your office, the specification will not be doing its job and may be deemed defective under 112.

  6. The USPTO’s decision to ban Wikipedia was reported in the Sept. 4, 2006 issue of BusinessWeek. (See
    “Kicking Wiki Out of the Patent Office” – link to businessweek.com)

    True, the total number of patents that cite Wikipedia is tiny, but impressive for a source that has only been around a few years. It compares well to more traditional publications. For example, the number of patents issued in 2009 that cite one or more IEEE publications is about 17,000 (a rough count).

    Mike White

  7. No problem with citing to Wikipedia, just note the date of the page that you are citing, and keep a copy of the page in your records in case the Wayback Machine or other archivers somehow fail to capture the page.

  8. I use Wikipedia as a technical dictionary to define background terms or to look up a spec (e.g., 802.11n).
    To solve the changing nature of Wikipedia, since I stole 1 or 2 sentences from it. I print the page out to PDF and then cite it in the IDS.

  9. “I think Mooney’s approach of lifting stuff verbatim would be smarter than trying to refer to Wikipedia in an application. It’s too much of a moving target.

    Well of course, I thought we were assuming the damage was already done. If I were to, for some reason, incorporate a wiki by reference then I would include a date in the citation. That’s just basic citation practice which everyone should do with any webpage.

  10. ” Is this rocket science or something? ”

    Yes it is. How do you know that the content that was present on the web page on the filing date is the same content that the inventor saw on that web page the last time he looked at it?

    I think Mooney’s approach of lifting stuff verbatim would be smarter than trying to refer to Wikipedia in an application. It’s too much of a moving target.

  11. On the other hand, I suspect the number of patents which recite verbatim text or images that were taken directly from Wikipedia is substantial.

  12. “I think the premise is that the wikipedia reference is actually in the patent specification ”

    “In that case, there would probably not be a citation in the file history. ”

    Then look at the time the app was filed? Is this rocket science or something?

  13. 6

    I think the premise is that the wikipedia reference is actually in the patent specification (as in “the steps for making soap are described in the wikipedia article found at //wikipedia/soap”). In that case, there would probably not be a citation in the file history.

    You can’t look at the wikipedia page or the references cited in it because if you did, you would only see how the page looked on the day you happened to look. It would not necessarily be the same as the way the page looked at the time the inventor incorporated it into the specification.

  14. “It is true there is some backup inherent in Wikipedia, and it’s also true that it is imperfect, as Noise suggests. But even if it were perfect, how exactly would someone who is reading the issued patent know how far behind the patent’s filing date to go to retrieve a snap shot of what the page looked like at the time the inventor saw it? ”

    Might they look at when it was cited in the pros. history? Or perhaps look on the page itself? References cited?

  15. Important small print from the Wayback site:

    “While the Archive endeavors to enforce its Terms of Use (link to archive.org) and maintain standard computer security, it is important for both those who visit the site (“Visitors”) and Researchers (collectively, “Users”) to be aware of the open nature of the Archive.”

    “It is possible that the computers at the Archive could become compromised by others and that the information on the Archive’s computers could be collected and disseminated without the knowledge or consent of the Archive. While the Archive endeavors to block “crackers” from breaking into its machines, the Archive is not responsible or liable for any such unauthorized uses of the Archive or its data.”

    And from the FAQ page:

    “How can I get pages authenticated from the Wayback Machine? How can use the pages in court?

    The Wayback Machine tool was not designed for legal use. We do have a legal request policy found at our legal page. Please read through the entire policy before contacting us with your questions. We do have a standard affidavit as well as a FAQ section for lawyers. We would prefer that before you contact us for such services, you see if the other side will stipulate instead. We do not have an in-house legal staff, so this service takes away from our normal duties. Once you have read through our policy, if you still have questions, please contact us for more information.”

    The legal page link currently results in a 404 Not Found message, while the standard affidavit contains the following:

    “The date assigned by the Internet Archive applies to the HTML file but not to image files linked therein. Thus images that appear on the printed page may not have been archived on the same date as the HTML file. Likewise, if a website is designed with “frames,” the date assigned by the Internet Archive applies to the frameset as a whole, and not the individual pages within each frame.”

    and finally, from back at the FAQ page:

    “Does the Internet Archive’s affidavit mean that the printout was actually the page posted on the Web at the recorded time?

    The Internet Archive’s affidavit only affirms that the printed document is a true and correct copy of our records. It remains your burden to convince the finder of fact what pages were up when.”

    Since the records are not guarenteed safe from mischief, and the affidavit only affirms the Internet Archive’s record copy and not the original, please be careful in any legal use, anywhere in the world of information obtained from the Wayback Machine.

  16. Night Writer

    Wayback machine pubilcations are admissible in oppositions to show what was known in the art prior to a particular date.

    But incorporating by reference in a disclosure raises a totally different issue. The issue there is what the applicant regards as his invention, and whether the disclosure actually discloses it well enough. The EPO doesn’t allow you to incorporate essential material by reference. The PTO lets you only if it is found in another patent.

  17. Krok

    It is true there is some backup inherent in Wikipedia, and it’s also true that it is imperfect, as Noise suggests. But even if it were perfect, how exactly would someone who is reading the issued patent know how far behind the patent’s filing date to go to retrieve a snap shot of what the page looked like at the time the inventor saw it?

  18. I often use Wikipedia in my examination work. However, I almost never use it as a reference to reject claims. Instead, I cite Wikipedia to show the general state of the art, or to demonstrate that something is well-known or conventional in the art. Additionally, more often than not, the Wikipedia entry will usually have links to other articles, specifications, standards, etc. that I *do* use to reject claims.

  19. “Several years ago the USPTO banned patent examiners from using Wikipedia as a source of information for determining patentability of inventions.”
    – seconding plurality, Dennis, can you provide more info on this?

    krok,
    The mechanism you allude to is not tamper proof, and is rather a slipshod control. Likewise, the wayback machine as an archival tool (not mentioned in the article) has been documented to be similarly substandard and not worthy of legal consideration – at least they do state that in their tiny print.

  20. I didn’t know about the banning. I don’t really see how they could ban it though, seeing as it is published material. If a copy of the article being used is provided at the time of rejection, dated to prior to filing of the application (via Wayback Machine), then what would be the problem?

  21. “Several years ago the USPTO banned patent examiners from using Wikipedia as a source of information for determining patentability of inventions.”

    [citation needed]

  22. “The trouble with Wikipedia citations is they can change after they are made”.
    WRONG
    Every version of a wikipedia article is saved an can be accessed via the history tab on top of each article.

  23. The trouble with Wikipedia citations is they can change after they are made. So whatever is in an article at the time of filing may differ from what’s in the article when the patent issues. Hopefully these citations to Wikipedia are not attempts to incorporate essential material by reference.

Comments are closed.