PTO Informative Decisions: Patent Eligibility Rejection — Look to what is expressly “RECITED” in the claims

The USPTO has identified three recent PTAB decisions as “INFORMATIVE.” All three will be helpful to patent prosecutors and so should be considered:

  • Obviousness: Hulu, LLC v. Sound View Innovations, LLC, Case IPR2018-00582, Paper 34 (Aug. 5, 2019) (designated informative on Dec. 11, 2019) (finding insufficient justification to combine references);
  • Obviousness: Johns Manville Corp. v. Knauf Insulation, Inc., Case IPR2018-00827, Paper 9 (Oct. 16, 2018) (designated informative on Dec. 11, 2019) (denying institution based upon insufficient justification to combine references); and
  • Eligibility: Ex parte Linden, 2018-003323 (Apr. 1, 2019) (designated informative on Dec. 11, 2019) (reversing examiner rejection based upon 2019 Examination Guidelines for eligibility).

This post will focus on Linden, which is a patent application owned by BAIDU spun out of Prof. Andrew Ng’s lab at Stanford.

In Linden, the BAIDU patent application claims a method using a trained neural network to transcribing speech.  The claimed method involves several data processing steps: normalizing the input (to the training data); generating a “jitter set” of audio files (time-distorted versions of the original); generating a spectrogram for each time-jiggered audio file; predicting character probabilities with the neural network; and transcribing the audio based upon character probabilities and a language model.

The examiner rejected the claims as directed to an abstract idea of manipulating data; creating information sets (based upon prior information sets); and decoding data.  On appeal, the PTAB reversed — finding that the claim should not be classified as directed to an abstract idea under the 2019 Examination Guidelines.  Under the Guidelines, abstract ideas must be classified as either:

  1. Mathematical concepts;
  2. Certain methods of organizing human activities; or
  3. Mental processes.

* The guidelines note claims outside the enumerated groupings can be treated as abstract ideas, but only in “rare circumstance.” *

Here, the Board found that the transcription claims are not directed at these categories. Focusing first on mental processes and human activities, although transcription is a human/mental activity, the Board found that the specific process here (jiggering/spectrogram sets/etc.) cannot be “practically performed mentally” nor do they recite “organizing human activity” — which is defined as activities such as using fundamental economic principles, commercial and legal interactions, and managing relationships.

Regarding mathematical concepts.  The claims here require several mathematical operations to be performed: jiggering; creating spectrogram; predicting character probabilities; etc.  However, the Board found that none of these operations actually “recite” a mathematical algorithm or formula.

[T]he Examiner identifies that the Specification discloses an algorithm to obtain the predicted character probabilities.  The mathematical algorithm or formula, however, is not recited in the claims.  As such, under the recent Memorandum, the claims do not recite a mathematical concept.

Linden. In case you didn’t see it – this is bonkers. The Board here is saying that the claim would be problematic only if it actually and expressly recited the algorithm that it uses.  Since the claim is drafted more broadly (i.e., at a higher level of abstraction), it cannot be seen as abstract.

The PTAB went on to explain that even if the claims did recite a mathematical concept — they are still not problematic because the claims as a whole are not “directed to an abstract idea” but rather any abstractions are “integrated into a practical application.”

Regarding Alice Step 2, the Board also faulted the examiner for failing to provide evidence that the claims do not include an inventive concept:

[T]he examiner concludes the claims do not include “any additional elements that amounts to significantly more than a judicial exception” but fails to provide sufficient factual support.  Berkheimer.


= = = =

Typo by the PTO?.  The Board identifies the case as “Ex PARTE LEE LINDEN, BENJAMIN LEWIS, AND ABHEEK ANAND.”  This is confusing to me because none of these individuals are listed as inventors on the patent (Awni Hannum is the first named inventor).

33 thoughts on “PTO Informative Decisions: Patent Eligibility Rejection — Look to what is expressly “RECITED” in the claims

  1. 9

    The link to the lindin case currently leads to a 404 error – anyone have the correct link? Thanks!

  2. 8

    Another easy one in my scheme.

    Method? Yes. Is the results of the method an item of information?

    Does the utility of the information arise by human consumption of the meaning of the information? No. The consumer is the compiler. The algo has no human meaning.


    Obvious? Maybe or maybe not.

    1. 8.1

      The utility of the printing press arises by human consumption, so I guess a printing press would not be eligible for patentability under your scheme.

      1. 8.1.1

        The utility of a printing press is faster, easier production of configured ink on paper.

        The press has utility without reference to what is being printed. The result of improved printing methods is not an item of information.

        With that reasoning power of yours, law school must have been a LOOOONG time ago….


          … says the guy who cannot understand the terrain of patent law and what “utility” means in this terrain…

          Marty, you really shouldn’t criticize your betters on a topic that you refuse to grasp.


              You wouldn’t know your betters’ because you are so mired in your own schemes (and do so without understanding the terrain upon which you would do battle).


          This sounds like a break through. The utility of shadow accounts is faster/easier transactions without reference to what is being transacted. Eligible!

  3. 7

    Laughable below that a claim is a mathematical concept.

    So a machine is built and it has utility. The claims pass 112. Novel and non-obvious, but it somehow has this mathematical concept that stops it from getting a patent.

    Information processing is a physical process. It takes time, energy, and space. The “mathematical concept” is a method for performing information processing and limits the claims as much as specifications of gears and levers.

    The ignorance is just astounding…

  4. 6

    “While we greatly respect the PTO’s expertise on all matters relating to patentability, including patent eligibility, we are not bound by its guidance. And, especially regarding the issue of patent eligibility and the efforts of the courts to determine the distinction between claims directed to natural laws and those directed to patent-eligible applications of those laws, we are mindful of the need for consistent application of our case law.” Cleveland Clinic v True Health (case 2018-1218, decided April 1, 2019) at 13 (nonprecedential).

    Simply put, the 2019 USPTO Guidelines are, to quote a previous post, bonkers. The fact that the CAFC has already said those Guidelines are meaningless, and now a PTAB decision is designated as “informative” based on those very meaningless Guidelines completely boggles the mind. Whether you like the Guidelines or not it besides the point. I’m at a complete loss as to how this kind of decision could be “informative” except to inform everyone in the field that the USPTO has no clue how to interpret the mess the courts have made out of 101. It’s time for Congress to step in and fix this mess. The courts and USPTO have already proven that they aren’t capable of it.

    1. 6.1

      Long reply lost because Internet blinked…

      In short, the Executive Branch cannot follow the lead of the Judicial Branch in the disregard for the Separation of Powers by misapplying the Common Law law writing t001.

      And make no mistake, what the Supreme Court set forth was a Separation of Powers violating power grab to continuously use Common Law for every case of contested eligibility.

      Common Law law writing has its place — but that place is NOT reaching a desired Ends at the expense of the words of Congress (and statutory law). What the Supreme Court has done is create a Gordian Knot of “ever-growing” mess with its “Gist/Abstract” Alice-Mayo two-step.

      Iancu (correctly, and objectively) noted that the Common Law Gordian Knot regarding eligibility was a contradictory mess.

      He tried to provide off-ramps so that his Executive Branch employees could avoid (as much as possible) contributing to the mess of added Common Law Alice-Mayo law writing.

      The elephant in the room is NOT the guidance. The elephant in the room is what the guidance is trying to follow.

    2. 6.2


      What your post does not draw attention to in that Cleveland Clinic case was that the claim that the courts rejected was directly modeled on an example claim from a previous Office guidance document.

      That previous Office guidance document example claim was modeled on a court case.

      The problem is the courts (and the Office guidance — both past and present) can only reflect the Gordian Knot created by the courts.

  5. 5

    In case you didn’t see it – this is bonkers. The Board here is saying that the claim would be problematic only if it actually and expressly recited the algorithm that it uses. Since the claim is drafted more broadly (i.e., at a higher level of abstraction), it cannot be seen as abstract.

    This is indeed the “correct” (in double plus good newspeak terms) way of thinking about the claim per the guidance. The claim asserting the mathematical formula as the main improvement is directed to that mathematical formula and may be ineligible, but a claim that is limited by functional result coupled with a specification that MERELY teaches that performing a mathematical algorithm as the only enabled means of achieving the result is not directed to ineligible subject matter and is eligible.

    In other words, while you cannot patent e=mc^2, you can patent the function of converting a measured mass reading into an energy value, provided that you have no dependent claim which makes the mistake of saying “wherein the conversion follows the form of e=mc^2” and leave e=mc^2 in the specification.

    The fact that this results in claims being eligible while embodiments of that same claim are ineligible is irrelevant. The fact that this turns eligibility into a nose of wax that may be twisted in any direction is irrelevant. The only way you can call it bonkers is if you assume the function of the guidance is to apply the law to the process of examination. That is not the function of the guidance, and it says as much.

    1. 5.2

      … and you are absolutely missing the point of the guidance BEING the attempt to apply a contradictory Gordian Knot law to the process of examination.

      That is exactly what Iancu is trying with the guidance.

      AND he is trying to do it in a way that respects Separation of Powers (even as the Judicial Branch stomps on that concept), and in a way (through the off-ramps) that does not ADD more rope to the Gordian Knot.

      But you, Random, are too invested in your own belief structure and engage only in understanding things that feed your world view to grasp what is going on.

    2. 5.3

      “That is not the function of the guidance, and it says as much.”

      How so? This sounds like anon-style reading.


          If your point is that your unfounded assertions extend beyond the literal meaning on text to the identities of other posters, I agree wholeheartedly.


            Your attempted strawman as to what my point may be only impugns you.

            Ben, your attempted insults find no basis.

  6. 3

    I do think the case caption is in error. In addition to Hannum, Public PAIR discloses ten (!) other named inventors, but none of them are the three amigos of Linden et al. Per LinkedIn, Hannum is now at Facebook, while Linden and his crew were all previously there, so maybe it was some kind of mix-up based on the Facebook connection. Also, given that the decision was issued on Apr. 1, 2019, it could be an obscure April Fools’ joke that none of us can appreciate.

  7. 2

    “In case you didn’t see it – this is bonkers. The Board here is saying that the claim would be problematic only if it actually and expressly recited the algorithm that it uses.”

    They just followed the 2019 Guidance. The 2019 Update confirms their interpretation, stating: “A claim does not recite a mathematical concept (i.e., the claim limitations do not fall within the mathematical concept grouping), if it is only based on or involves a mathematical concept. For example, a limitation that is merely based on or involves a mathematical concept described in the specification may not be sufficient to fall into this grouping,
    provided the mathematical concept itself is not recited in the claim

    It’s the January 2019 Guidance that is bonkers, not the board.

    1. 2.1

      It’s the January 2019 Guidance that is bonkers, not the board.

      Not mutually exclusive. They can both be bonkers.

      1. 2.1.1

        Sure, but anticipating your boss’s expectations and conforming your work to those expectations is hardly indicative of being bonkers.

    2. 2.2

      Yeah, it even had an example that followed this stupidity.

      The 2019 guidelines left us in a bad state. Claim 1 and 2 below would be eligible, while claim 3 below– which is dependent and narrower than claims 1 and 2– would not be eligible.

      1. A method comprising:
      measuring voltage drop across an electrical element; measuring current flow through the element; based on the measurements, determining resistance of the element.

      2. The method of claim 1, wherein a mathematical model is used to determine the resistance.

      2. The method of claim 1, wherein the model is R=V/I.


          This is NOT the “fault” of the guidance (per se), and is more the “fault” of the Gordian Knot of contradictory common law (re)writing 35 USC 101.

          ALL of the “fault” here stems from the original source of the Supreme Court.

      1. 2.2.2

        I would consider your hypothetical claim 3 to be patent eligible, since claim 1 integrates it into a practical application. YMMV.

    3. 2.3

      >>mathematical concept

      The problem with this gobbledygook is that it has nothing to do with reality.

  8. 1

    As has been distinguished:

    applied math (i.e., most any engineering);
    Math(S) (philosophy)

    These really are different things.

Comments are closed.