written by The Law Offices of James David Busch LLC. 

James is Licensed to Practice before the USPTO, in Illinois and Arizona.

Using Deep Neural Networks to Seed Forward Citations with Third-Party Submissions

Using Deep Neural Networks to Seed Forward Citations with Third-Party Submissions

Forward patent citations of your specification can make your patent more valuable. In my prior article, Finding Forward Citations and Getting Claims Your Competitors Want Using Continuations, I explained:

What are forward citations?  Forward citations are instances where later filed patent applications cite back to your patent specification.  This is similar to a journal article citing to prior noteworthy or famous journal articles. 

Whether you are the author of a journal article, or the inventor of a patent, your work becomes more "valuable" each time a subsequent work cites back to yours.  Many patent industry commentators have stated that the number of forward citations is an important indicator of the “value” or importance of a patent.

I also noted in the article, that third parties are able to provide citations in patent applications:

“Cited by Third Party” (CBT).  A forward cite that is CBT is one that is provided by a third party on a third party submission form during prosecution.  Some patent owners may proactively submit these forms in competitor’s patent filings.  Like a CBA forward citation, it creates a record that the forward patent owner has actual knowledge of the cited patent, and further provides some mapping of the cited patent to the claims pending in the forward application.

Data collected by other commentators supports the conclusion that forward citations are one important component of what makes a patent valuable:

We found that forward citations (later patents that cite the subject patent) were the most significant factor in identifying patents that were likely to be purchased. In fact, the patents that were sold—or even highlighted by brokers, e.g. the representative patent—in a brokered patent package exhibited an even more extreme number of forward citations than litigated patents.

Deep neural networks can be used to help patent owners increase their forward citation count through the use of Third-Party Submission (3PS) forms in their competitors patent applications.

What is a Third-Party Submission (3PS)? A 3PS allows a third-party to submit a listing of art (Currently: 10 references / $180 fee) in a pending patent application. The requirements are set forth in MPEP 1134, 37 CFR 1.290, and 35 USC 122(e):

(e) PREISSUANCE SUBMISSIONS BY THIRD PARTIES.—

(1)_ IN GENERAL.— Any third party may submit for consideration and inclusion in the record of a patent application, any patent, published patent application, or other printed publication of potential relevance to the examination of the application, if such submission is made in writing before the earlier of—

(A) the date a notice of allowance under section 151 is given or mailed in the application for patent; or

(B) the later of—

(i) 6 months after the date on which the application for patent is first published under section 122 by the Office, or

(ii) the date of the first rejection under section 132 of any claim by the examiner during the examination of the application for patent.

(2) OTHER REQUIREMENTS.—Any submission under paragraph (1) shall—

(A) set forth a concise description of the asserted relevance of each submitted document;

(B) be accompanied by such fee as the Director may prescribe; and

(C) include a statement by the person making such submission affirming that the submission was made in compliance with this section.

After identifying relevant art for submission to application that satisfies the time requirements of 35 USC 122(e)(1), the specific portions of the art must be identified to satisfy 35 USC 122(e)(2). These submissions are made electronically, or in paper on form PTO/SB/429. See also USPTO Preissuance Submission FAQ.

Using Deep Neural Networks to Generate Third-Party Submissions.

Specifically, methods employing deep neural networks can help patent owners increase their forward citation count, and protect their intellectual property rights by preventing competitors from getting claims that read on the patent owner’s disclosure.

Increasing your Forward Citation Count. Deep neural network methods can help patent owners:

  • find relevant later patent application publications within the time period specified in 35 USC 112(e)(1), based on the published claims of those later-filed competitor-owned applications; and,

  • identify relevant portions of the patent owners specification mapped onto the claims of the later competitor-owned applications as required in 35 USC 122(e)(2).

Based on this information a 3PS form can be completed and submitted in the third-party application. In time, the patent owner’s application will then show an increased forward citation count based on that submission.

Citing Additional Art Against Competitor Applications.

The USPTO allows the filing of 3PS without a fee if a fee exemption applies to the submission.  A third party is exempt from paying a fee for a submission of three or fewer documents, provided it’s the party's first such submission and the party files a "first and only" statement.

To the extent that you are not exempt from paying the fee, the current USPTO fee (at the time of writing) to file a 3PS is $180 for each group, or partial group, of 10 documents.

To maximize the opportunity to submit art, and provide some level of anonymity, a patent owner may consider conducting a second deep neural network invalidity search for the third-party application limited to documents that are not prior art to the patent owner’s specification, but that are prior art to the subsequent third party application. The additional references may narrow the scope of protection that the competitor is ultimately able to obtain (but the references do not apply as 35 USC 102 art against the patent owner’s specification).

Potential Drawbacks to Filing a Third-Party Submission against a Competitor’s Patent.

There are potential drawback to the use of 3PS in a competitor's application.

  • When filing, the statement of relevance cannot make arguments for invalidity of the pending claims; however a claim chart can be submitted that stops short of making the statement that the claim is anticipated etc.

  • Once filed, the third party filer does not have any participation in the proceedings.

  • The competitor will have the opportunity to amend its claims and/or argue that the cited materials do not prevent patentability of their claims.

  • To the extent the third-party is able to distinguish the art or otherwise obtain allowance over the art, their patent will be arguably stronger for it.  

  • One study found that "preissuance submissions are relied upon by examiners less than 13% of the time they are properly submitted."

If you would like to explore this process further, contact JDB IP.

Using Deep Neural Networks to Decide Whether to Copy Claims or File a Third-Party Submission

Using Deep Neural Networks to Decide Whether to Copy Claims or File a Third-Party Submission

Using Deep Neural Networks to Mine Your Patent Specification for Claims that Your Competitors Want

Using Deep Neural Networks to Mine Your Patent Specification for Claims that Your Competitors Want