DETAILED ACTION
Claims 1-6, 8-14, and 16 are presented for examination.
This Office Action is in response to submission of documents on October 7, 2025.
Previous rejection of claims 1-6, 8-14, and 16 under 35 U.S.C. 112(a) as failing to comply with the written description requirement is withdrawn. A new rejection is asserted on different grounds.
Previous rejection of claims 9-15 and 16 under 35 U.S.C. 112(b) as being indefinite is withdrawn. A new rejection under 35 U.S.C. 112(b) is asserted on different grounds.
Previous objection to claim 1 for minor informalities is withdrawn. A new objection to the title of the invention is asserted herein.
Rejection of claims 1-6, 8-14, and 16 under 35 U.S.C. 101 as being directed to unpatentable subject matter is maintained.
Rejection of claims 1-6, 8-14, and 16 as being obvious over Beser in view of Lally and Trost is maintained.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Arguments
Regarding the finality of this Office Action and the arguments asserted by the Applicant in the Response, Examiner is not persuaded that the finality of this Office Action is precluded by the alleged unanswered arguments from the previous action. Response at pg. 8. Applicant cites MPEP 707.07(f) as requiring the examiner to take note of the applicant’s argument and answer the substance of it. However, Applicant should take note of the language that indicates that the examiner should take note but does not indicate that the examiner must take note, as asserted by Applicant, particularly when the argument does not properly address the rejection. Thus, the examiner is not under a mandatory requirement to take note of every argument.
The arguments set forth by Applicant in the previous Response do not correctly address the rejections and therefore do not shift the burden back to Examiner. In that Response of October 7, 2025, Applicant asserts that Beser discloses nodes that are “on-chain” and therefore does not suggest the first and second computing systems. Response at pg. 22. However, the “off-chain” aspect of the claim presented therein is not alleged to be taught by Beser. Beser discloses the limitations of a “first computing system” and a “second computing system” that are “on-chain,” as admitted by Examiner when the requirement that “said first computing system and said second computing system are off chain systems” is asserted as being taught by Lally. See Office Action of November 18, 2025 at pg. 19, pg. 22, and pg. 23. Accordingly, the argument does not correctly address what is taught by Beser because Applicant asserted that “the Beser [reference] does not disclose or suggest the recited first and second provider systems1,” Response at pg. 23, which is not correct for the reasons provided below in the rejections under 35 U.S.C. 103. The two systems are disclosed, albeit “on-chain.”
Lally, on the other hand, is the reference cited for teaching the “on-chain” aspect of the claim. See Office Action of 11/18 at pg. 23. Besides that limitation, the only other portions of claim 1 that are rejected as being disclosed by Lally are transmitting, by the first computing system, to a second computing system within the system, a request for a…data set and receiving, by the second computing system, the request from the first computing system. Thus, any limitations directed to the data being in the blockchain before or after issuance of a request is not rejected using Lally as a reference. Accordingly, the rejection is not correct and therefore was not individually addressed in the subsequent Office Action. See., e.g., Office Action at pg. 18, indicating that “in response to receiving the request from the first computing system, generating, by the second computing system, the…data set” is taught by Beser.
To summarize, regarding the previous Response of Applicant, the arguments related to Beser and Lally do not address the pending rejections. Applicant is essentially arguing the references separately and claiming that Beser does not disclose what is rejected by Lally and Lally does not disclose what is disclosed by Beser. Therefore, the detailed rejection of the claim provided with the 35 U.S.C. 103 provides ample rationale for which the Applicant had an opportunity to rebut in the Final Office Action.
Accordingly, this Action is made final.
Regarding the objection for minor formalities, Examiner was mistaken on the correct word usage that was missing from the claim. Applicant is correct in assuming the Examiner intended the term “wherein” as opposed to “whether.” Accordingly, the objection is withdrawn.
Applicant's arguments regarding rejection of claims 1-6, 8-14, and 16 under 35 U.S.C. 101 have been fully considered but they are not persuasive for the following reasons:
Applicant asserts that “the generation of the synthetic model, as recited in Applicant’s claim 1, cannot practically be performed in the human mind and therefore does not fall within the mental processes grouping.” Response at pg. 11. In response, Examiner notes that the claims do not recite a step of “generating a synthetic model.” Accordingly, the argument is not persuasive.
Applicant asserts that the claim, as amended to recite :generating, by the second computing system, using a new generative adversarial network, the synthetic data set such that it approximates real data for application to the machine learning model.” As indicated in the previous Office Action, generating synthetic data that approximates real data is a mental process that can be performed in the human mind. For example, the real data may include a plurality of values and a human can, with pencil and paper, add 1 to half of the values and subtract 1 from half of the values, resulting in a synthetic data set with the same average value as the real data set, thereby approximating the real data. By reciting that the values are generated using a GAN, the limitation is mere instructions to apply an abstract idea using a generic computer component (i.e., a machine learning model that is not recited with any specific features that indicate it is anything but a generic model). Accordingly, the argument is not persuasive.
Applicant next argues why a “computational model” is not a mathematical concept and therefore the claim is patentable subject matter. Upon reviewing the claims, as amended, it appears that the limitation of “a computational model” is not recited. Accordingly, Applicant’s arguments are not persuasive.
With regard to the limitation of a “machine learning model,” Applicant argues that a “machine learning model” is not a mathematical concept. Upon review of the Specification to ascertain the description of machine learning model to analyze whether, as the term is defined, it can be considered a mathematical concept or a potential additional element that may integrate the other judicial exceptions into a practical application, improve a technology, and/or amount to significantly more than the recited judicial exceptions. The Specification has the term “machine learning model” in it once: in the title. Thus, any arguments related to the term “machine learning model” are not relevant because the term clearly does not have support in the Specification in order for Examiner to properly evaluate the claim limitation.
With regards to the steps of “storing,” “transmitting,” and “receiving” data, Examiner is not persuaded that the terms provide an inventive step. See Response at pg. 14. Sending, storing, and/or receiving data is not a “technological solution” and the claims (and Specification) do not disclose that any novel form of performing any of those steps. The steps are necessarily required for any application that involves multiple computing systems that are in communication. Thus, these limitations do not “apply, rely on, or use the judicial exception in a manner that imposes a meaningful limit on the judicial exception, such that the claim is more than a drafting effort designed to monopolize the judicial exception." See, e.g., p. 8 of the 2019 PEG (quoting Response at pg. 13).
Applicant then argues why each of these steps is not a extra-solution activity and/or why the limitation imparts subject matter patentability on the claims that otherwise only recite judicial exceptions. A summary of the reasons why each argument is unpersuasive is as follows:
“Storing…” is an additional element that has no direct relationship with any of the recited judicial exceptions. The limitation is essentially “store a model,” the storage of which is not described in the Specification as being anything other than the accepted meaning of storing data. Thus, the limitation has no relation to generating the synthetic data set, applying the synthetic data set, nor determining a validity measurement. The step of storage is tangential, at best, and essential to any application whereby data is stored.
“Transmitting…” is recited as involving sending data that has been stored in a blockchain. The “transmitting” is not tied in any manner to generating and/or applying the synthetic data. Instead, the step is a necessary operation that is essential in any application whereby one system has data and another system requires the data. The limitation that is was stored in a block of a blockchain before transmission has no bearing on the usage of the data after transmission.
“Receiving…” is not recited with any details (nor disclosed) as being anything but the commonly accepted definition of “receiving data.” The step of receiving data is not tied to the generation nor application of the data, nor to the determination of a validity measurement. The step is essential to any application whereby data is present within a first computing system and is required by another computing system.
Regarding Desjardins, the claims included limitations that reflected improvements disclosed in the specification. The claim recited, inter alia, “adjust the first values of the plurality of parameters to optimize performance of the machine learning model on the second machine learning task while protecting performance of the machine learning model on the first machine learning task,” which was found to reflect “improvements as to how the machine learning model itself operates, including training a machine learning model to learn new tasks while protecting knowledge about previous tasks to overcome the problem of “catastrophic forgetting” encountered in continual learning systems.” See Memorandum of December 5, 2025 by Charles Kim, indicating changes to MPEP 2106.04(d), Subsection III. In the present claims, the steps of storing, transmitting, and receiving models and/or synthetic do not reflect any technological improvement disclosed in the Specification. Nothing in these steps, for example, limit data transmission to only synthetic data and/or preventing the real data from also be transmitted. Thus, in summary, Examiner is not persuaded by the argument that the additional elements “enable secure, zero-knowledge validation of a machine learning model using a blockchain-mediated synthetic data exchange” nor that “they are the mechanisms by which the technical problem identified in the Background-secure validation without disclosure of real data or models-is solved” because the claimed steps do not themselves do any of these activities (e.g., validating, limiting disclosure of real data and/or models). The steps are necessary for any data validation system and the inclusion of the limitations do not meaningfully limit the generation, application, and validation of the data and/or model.
Accordingly, rejection of the claims under 35 U.S.C. 101 are maintained.
Applicant’s arguments with respect to the rejection of claims 1-6, 8-14, and 16 under 35 U.S.C. 103 have been fully considered and are not persuasive. With regards to the arguments related to Beser and Lally, Examiner has already addressed the rejections above. With regards to Trost, [006] discloses: “One general aspect includes a method for generating mock test data for an application. The method includes providing a random input to a generator model. The random input is transformed into generated data that is then provided to a discriminator model along with production data.” The data is generated without using the real data. Although sent with production (i.e., “real” data), the actual generation of the data does not require the production data. It is only the discriminator training that requires the production data.
Further “for application to the computational model” is an intended use and therefore carries little, if any, patentable weight. A recitation of the intended use of the claimed invention must result in a structural difference between the claimed invention and the prior art in order to patentably distinguish the claimed invention from the prior art. If the prior art structure is capable of performing the intended use, then it meets the claim. Here, the synthetic data could be sent without production data (e.g., no production data exists for testing) and therefore the prior art is capable of performing the recited limitation.
Regarding the teachings of Trost and the new limitations of a “machine learning model” and a “new generative adversarial network,” Examiner disagrees because at least [0023]-[0024] disclose such limitations. The reasoning is provided below with respect to the rejection under 35 U.S.C. 103.
Accordingly, rejection of claims 1-6, 8-14, and 16 as being obvious over Beser in view of Lally and Trost is maintained.
Specification
The title of the invention is not descriptive. A new title is required that is clearly indicative of the invention to which the claims are directed. The title, “METHOD AND SYSTEM OF MACHINE LEARNING MODEL VALIDATION IN BLOCKCHAIN THROUGH ZERO KNOWLEDGE PROTOCOL,” indicates that the claims are directed to validating a machine learning model. While the claims now recite “a machine learning model,” the term is not present at all in the Specification. Thus, in addition to the rejection under 35 U.S.C. 112(a), which will require amendments removing “a machine learning model” from the claims, the title must be changed to reflect what is actually being claimed (e.g., “METHOD AND SYSTEM OF COMPUTATIONAL MODEL VALIDATION IN BLOCKCHAIN THROUGH ZERO KNOWLEDGE PROTOCOL”).
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1-6, 8-14, and 16 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claims contain subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention. Independent claims 1 and 9 recite “a machine learning model.” The terms only appears in the title of the application and not within the actual disclosure. Accordingly, the term is not supported. Appropriate correction is required. For the purposes of examination, the term “machine learning model” will be interpreted as a “computational model,” as disclosed in the Specification.
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-6, 8-14, and 16 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Independent claims 1 and 9 recite “using a new generative adversarial network.” It is unclear what is meant by “new.” In one interpretation, “new” can mean an additional GAN other than an already existing GAN. In another interpretation, “new” can mean a structure that has been recently developed. As disclosed in the Specification, “the provider system 106 may use a known or new generative adversarial network (GAN).” Spec. at [0022]. Thus, it appears that “new” is intended to by synonymous with “novel” as it is the opposite of “known.” If that is the intended meaning, the Specification lacks disclosure as to the novel network that was developed for this application. The use of “novel” in this context would necessarily encompass any future GANs used for the purposes claimed herein. Further, the term is indefinite as to when a GAN should be considered “old” versus “new.” Appropriate correction is required.
The independent claims recite “protecting real data protecting real data by generating…the synthetic data set such that it approximates real data for application to the machine learning model but without using real data.” It is unclear how the act of generating synthetic data is protecting real data. The real data still exists and is capable of being provided to one or more parties. The act of generating the data itself does not ensure any protection for the data but is instead a step in the process of transmitting only synthetic data. Accordingly, the limitation is not given any weight regarding patentability of the claims.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-6, 8-14, and 16 are rejected under 35 U.S.C. 101 because the claimed invention is directed to judicial exceptions without significantly more. The claims recite mathematical concepts and mental processes. These judicial exceptions are not integrated into a practical application because the additional elements are extra-solution activities that are not sufficient to amount to significantly more than the judicial exception because the elements are well-understood, routine, and conventional, and/or are characterized as types of elements that courts have found to be insignificantly more than recited judicial exceptions.
Claim 1
Step 1: The claim is directed to a process, falling under one of the four statutory categories of invention.
Step 2A, Prong 1: The claim 1 limitations include (bolded for abstract idea identification):
Claim 1
Mapping Under Step 2A Prong 1
A method for determining the validity of a machine learning model using a blockchain and zero knowledge principles, comprising:
storing, in a memory of a first computing system within a system, a machine learning model;
transmitting, by the first computing system, to a second computing system within the system, a request for a synthetic data set for use in validating the machine learning model;
receiving, by the second computing system, the request from the first computing system;
in response to receiving the request from the first computing system, protecting real data by
generating, by the second computing system, using a new generative adversarial network, the synthetic data set such that it approximates real data for application to the machine learning model but without using real data;
after generating the synthetic data set, transmitting, by the second computing system, the generated synthetic data set to a blockchain node in a blockchain network, wherein said first computing system and said second computing system are off-chain systems;
storing, by the blockchain node in the blockchain network, the generated synthetic data set received from the second computing system;
transmitting, by the blockchain node, a blockchain data value included in one block of a plurality of blocks of the blockchain, wherein the blockchain data value includes the generated synthetic data set
receiving, by a receiver of the first computing system, from the blockchain network, the blockchain data value that includes the generated synthetic data set previously generated by the second computing system;
receiving, by the receiver of the first computing system, an expected accuracy value;
generating, by a processor of the first computing system, a result value via application of the generated synthetic data set to the machine learning model; and
determining, by the processor of the first computing system, a validity measurement for the machine learning model based on a comparison of the generated result value and the expected accuracy value.
Abstract Idea: Mental Process and/or Mathematical Calculations
Generating synthetic data is a process that can be performed by a human using pencil and paper and/or a generic computer to facilitate the data generation. The process can include evaluating real data and creating new data that imitates the real data. Further, the process of generating synthetic data can include performing one or more calculations, such as randomly generating data, that comply with one or more statistical models that approximate real data and/or real data distribution. See MPEP § 2106.04(a)(2), Subsections I and III.
Abstract Idea: Mathematical Calculations
The step includes executing a machine learning model. The model is comprised of one or more mathematical functions and includes receiving input, performing calculations, and providing output based on the calculations. See MPEP § 2106.04(a)(2), Subsection I.
Abstract Idea: Mathematical Calculations
The step includes comparing a result (i.e., a value) from a machine learning model to a given value, and providing a result. Comparison of values is a mathematical concept that includes performing one or more mathematical calculations. See MPEP 2106.04(a)(2).
Step 2A, Prong 2: The claim 1 limitations recite (bolded for additional element identification):
Claim 1
Mapping Under Step 2A Prong 2
A method for determining the validity of a machine learning model using a blockchain and zero knowledge principles, comprising:
storing, in a memory of a first computing system within a system, a machine learning model;
transmitting, by the first computing system, to a second computing system within the system, a request for a synthetic data set for use in validating the machine learning model;
receiving, by the second computing system, the request from the first computing system;
in response to receiving the request from the first computing system, protecting real data by generating, by the second computing system, the synthetic data set such that it approximates real data for application to the machine learning model but without using real data;
after generating the synthetic data set, transmitting, by the second computing system, the generated synthetic data set to a blockchain node in a blockchain network, wherein said first computing system and said second computing system are off-chain systems;
storing, by the blockchain node in the blockchain network, the generated synthetic data set received from the second computing system;
transmitting, by the blockchain node, a blockchain data value included in one block of a plurality of blocks of the blockchain, wherein the blockchain data value includes the generated synthetic data set
receiving, by a receiver of the first computing system, from the blockchain network, the blockchain data value that includes the generated synthetic data set previously generated by the second computing system;
receiving, by the receiver of the first computing system, an expected accuracy value;
generating, by a processor of the first computing system, a result value via application of the generated synthetic data set to the machine learning model; and
determining, by the processor of the first computing system, a validity measurement for the machine learning model based on a comparison of the generated result value and the expected accuracy value.
This limitation recites an idea of a solution. See MPEP 2106.05(f)(1). The recitation is further not a step of the method but is part of the preamble. However, even if given weight, “the validity” is not recited with specificity to limit the claim to any particular type of output or result.
Storing information in memory is an extra-solution activity that involves data transmission and can be performed by generic computer components, such as a database and/or other computer memory. The claim does not recite any component nor recite the data storage with any specificity. See MPEP 2106.05(g)(3).
Providing data (i.e., transmitting data) is an extra-solution activity that does not integrate the judicial exception into a practical application. The limitation does not recite, with specificity, how the data is provided and therefore does not improve the functioning of a computer. See MPEP 2106.05(d)(II).
Receiving data is an extra-solution activity that amounts to mere data gathering and transmission. See MPEP 2106.05(g)(3).
Protecting real data using a judicial exception (mental process and/or mathematical concepts) is an idea of a solution that is not recited with specificity such that it integrates the judicial exception into a practical application and/or improves a technology. See MPEP 2106.05(f)(1).
Providing data (i.e., transmitting data) is an extra-solution activity that does not integrate the judicial exception into a practical application. The limitation does not recite, with specificity, how the data is provided and therefore does not improve the functioning of a computer. See MPEP 2106.05(d)(II).
Storing information in memory is an extra-solution activity that involves data transmission and can be performed by generic computer components, such as a database and/or other computer memory. The claim does not recite any component nor recite the data storage with any specificity. See MPEP 2106.05(g)(3).
Providing data (i.e., transmitting data) is an extra-solution activity that does not integrate the judicial exception into a practical application. The limitation does not recite, with specificity, how the data is provided and therefore does not improve the functioning of a computer. See MPEP 2106.05(d)(II).
Receiving data, regardless of the type or source, is an extra-solution activity that amounts to mere data gathering and transmission. See MPEP 2106.05(g)(3) (e.g., “Selecting a particular data source or type of data to be manipulated” is insignificant extra-solution activity).
Receiving data is an extra-solution activity that amounts to mere data gathering and transmission. See MPEP 2106.05(g)(3).
Step 2B: Regarding Step 2B, the inquiry is whether any of the additional elements (i.e., the elements that are not the judicial exception) amount to significantly more than the recited judicial exception. In claim 1, the steps of “storing,” “receiving,” and transmitting data are extra-solution activities that courts have found to be insignificantly more than the recited judicial exception. See, e.g., MPEP 2106.05(d)(II), Subsections i and iv; MPEP 2106.05(g)(3); Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93. See also Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network).
Claim 2
Claim 2 recites wherein receiving the expected accuracy value further includes receiving a cryptographic key. This step is merely data gathering facilitated by data transmission, both of which courts have found to be insignificant extra-solution activities. See MPEP 2106.05(g)(3). The claim further recites validating, by the processor of the first computing system, the data set using the cryptographic key. However, using a key to validate blockchain data is well-understood, routine, and conventional activity that does not amount to significantly more than the recited judicial exception. See, e.g., Beser et al., U.S. Patent Publication 2019/012595 at [0049]: “For example, each node 110-1, 110-2, 110-3 may use its associated public key to digitally sign transactions (e.g., transmissions of model update data) in system 100. Each node 110-1, 110-2, 110-3, and/or validation node 130-1, 130-2 may also be configured to verify digital signatures, maintain integrity of Merkle tree, verify proof of work, and/or the like, similar to typical blockchain technologies.” Accordingly, the claim is rejected under 35 U.S.C. 101 as being directed to unpatentable subject matter.
Claim 3
Claim 3 recites wherein the expected accuracy value is received from the blockchain data value. The limitation merely indicates a source for the “expected accuracy value” and does not include additional elements. Further, the limitation does not elevate any previous additional element to significantly more than the recited judicial exception. Accordingly, the claim is rejected under 35 U.S.C. 101 as being directed to unpatentable subject matter.
Claim 4
Claim 4 recites wherein the expected accuracy value is received from the second computing system. The limitation merely indicates a source for the “expected accuracy value” and does not include additional elements. Further, the limitation does not elevate any previous additional element to significantly more than the recited judicial exception. Accordingly, the claim is rejected under 35 U.S.C. 101 as being directed to unpatentable subject matter.
Claim 5
Claim 5 recites wherein the machine learning model is received from the second computing system prior to storage in the memory of the first computing system. The limitation merely indicates a source for the “machine learning model” and does not include additional elements since receiving the machine learning model is not positively recited as a step. Further, the limitation does not elevate any previous additional element to significantly more than the recited judicial exception. Accordingly, the claim is rejected under 35 U.S.C. 101 as being directed to unpatentable subject matter.
Claim 6
Claim 6 recites further comprising: receiving, by the receiver of the first computing system, one or more additional variables from the second computing system. The step recites the extra-solution activities of data transmission and data gathering, both of which have been found by courts to be well-understood, routine, and conventional insignificant extra-solution activities. See Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network); but see DDR Holdings, LLC v. Hotels.com, L.P., 773 F.3d 1245, 1258, 113 USPQ2d 1097, 1106 (Fed. Cir. 2014) (“Unlike the claims in Ultramercial, the claims at issue here specify how interactions with the Internet are manipulated to yield a desired result‐‐a result that overrides the routine and conventional sequence of events ordinarily triggered by the click of a hyperlink.”
The claim further recites refining the determined validity measurement based on the one or more additional variables. Given its broadest reasonable interpretation, “refining” includes adjusting a value to a different, possibly more accurate, value. In some instances, the step may include one or more mathematical calculations and thus be considered an abstract idea. In other instances, “refining” a measurement may include replacing a value with a different value, which can be considered a mental process that can be performed by a human using pencil and paper (or a generic computer). See MPEP 2106.04(a)(2), Subsection III(C). Accordingly, the claim is not patent eligible under 35 U.S.C. 101.
Claim 8
Claim 8 recites wherein the synthetic data set is generated using a generative adversarial network. A generative adversarial network is a machine learning model that includes performing one or more mathematical functions. Accordingly, the claim recites additional abstract ideas and therefore does not include additional elements that would amount to significantly more than the already-identified judicial exceptions. Accordingly, the claim is not patent eligible under 35 U.S.C. 101.
Claim 9
Claim 9 recites a system for performing the steps of the method recited in claim 1, comprising: a blockchain network, a first computing system, a second computing system, including a memory,…a receiver…, a transmitter, and a processor. The additional elements are generic computer components that are not recited with any specificity that would amount to significantly more than the recited judicial exceptions. Courts have found that recitation of generic computer components to perform a judicial exception do not integrate the judicial exception into a practical application and amount to merely reciting an exception with the words “apply it.” See MPEP 2106.05(f). The remining limitations of claim 9 are rejected for at least the same reasons as provided regarding claim 1. Accordingly, the claim is not patent eligible under 35 U.S.C. 101.
Claim 10
Claim 10 recites substantially the same limitations as claim 2. Accordingly, for at least the same reasons as provided for claim 2, claim 10 is rejected under 35 U.S.C. 101 for being directed to unpatentable subject matter.
Claim 11
Claim 11 recites substantially the same limitations as claim 3. Accordingly, for at least the same reasons as provided for claim 3, claim 11 is rejected under 35 U.S.C. 101 for being directed to unpatentable subject matter.
Claim 12
Claim 12 recites substantially the same limitations as claim 4. Accordingly, for at least the same reasons as provided for claim 4, claim 12 is rejected under 35 U.S.C. 101 for being directed to unpatentable subject matter.
Claim 13
Claim 13 recites substantially the same limitations as claim 5. Accordingly, for at least the same reasons as provided for claim 5, claim 13 is rejected under 35 U.S.C. 101 for being directed to unpatentable subject matter.
Claim 14
Claim 14 recites substantially the same limitations as claim 6. Accordingly, for at least the same reasons as provided for claim 14, claim 14 is rejected under 35 U.S.C. 101 for being directed to unpatentable subject matter.
Claim 16
Claim 16 recites substantially the same limitations as claim 8. Accordingly, for at least the same reasons as provided for claim 8, claim 16 is rejected under 35 U.S.C. 101 for being directed to unpatentable subject matter.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1-6, 8-14, and 16 are rejected under 35 U.S.C. 103 as being obvious over Beser (U.S. Patent Pub. No. 2019/0012595) in view of Lally, et al., (WIPO Pub. No. 2021138426, hereinafter “Lally”), and Trost, et al., (U.S. Patent Pub. No. 2021/0271591, hereinafter “Trost”).
Claim 1
Beser discloses:
A method for determining the validity of a machine learning model using a blockchain and zero knowledge principles, comprising:
A neural network may comprise one or more nodes and validation nodes. Each node may execute a computing model and, in response to detecting a model update event in the computing model, may generate model update data and transmit the model update data to one or more validation nodes. The validation nodes may generate an updated computing model based on the model update data. The validation nodes may validate and consent to the model update data and/or the updated computing model using blockchain technologies. Beser at Abstract.
storing, in a memory of a first computing system within a system, a machine learning model;
In various embodiments, each computing node in the ANN may initially be loaded with an initial computing model. Beser at [0018].
In various embodiments, system 100 may comprise one or more computing nodes 110 (e.g., a first node 110-1, a second node 110-2, a third node 110-3, etc.) and one or more validation nodes 130 (e.g., a first validation node 130-1, a second validation node 130-2, etc.). Beser at [0027].
The “validation node” is analogous to “the first computing system.”
Validation nodes 130 may also be configured to validate the model update data by locally testing the model update data to determine whether the model update data cures the prediction error, detects the new model requirement, or the like. For example, validation nodes 130 may locally implement the test data and/or model update, and may process data using the computing model to determine whether the prediction error is still an issue, whether the new model requirement can be detected, or the like. Beser at [0055].
The model update data may comprise training data (e.g., data to be ingested by a node to “train” the computing model to cure the prediction error, detect the new model requirement, etc.) and/or a model update (e.g., programmable code to merge with the computing model to cure the prediction error, detect the new model requirement, etc.). Beser at [0040].
receiving, by the second computing system, the request
In various embodiments, in response to detecting a model update event while using a computing model, processor 112 may be configured to generate model update data, as discussed further herein. Beser at [0040].
Node 110 detects a model update event (step 302). For example, node 110 may detect the model update event while operating the computing model, through a transmission (e.g., an update is transmitted to node 110), or the like. Beser at [0052].
“Detecting a model update event” is analogous to “receiving…the request.”
The “node 110” is analogous to the “second computing system.”
in response to receiving the request from the first computing system, protecting real data by generating, by the second computing system
Node 110 generates model update data (step 304) based on the model update event. The model update data may comprise training data (e.g., data to be ingested by a node to “train” the computing model to cure the prediction error, detect the new model requirement, etc.), a model update (e.g., programmable code to merge with the computing model to cure the prediction error, detect the new model requirement, etc.), and/or a new computing model. Beser at [0052].
In various embodiments, and with reference to FIG. 1B, an exemplary computing node 110 is depicted (e.g., node 110-1, 110-2, and/or 110-3). Node 110 may comprise any suitable combination of hardware, software, and/or database or memory components. For example, node 110 may comprise a processor 112, a memory 114, and/or a communication interface 116. Beser at [0037].
The “model update data” is analogous to the “data set” and is generated “based on the model update event” (analogous to “in response to receiving the response…”).
after generating the synthetic data set, transmitting, by the second computing system, the
Node 110 transmits the model update data to validation node 130 (step 306). In various embodiments, validation node 130 may transmit the model update data based on a distribution algorithm or the like. Beser at [0053].
Validation network 120 may comprise any suitable number of validation nodes 130, and each validation node 130 may be in electronic communication with one or more other validation nodes 130. In that respect, each validation node 130 may function as a blockchain node in the blockchain network. Beser at [0034].
storing, by the blockchain node in the blockchain network, the generated synthetic data set received from the second computing system;
In response to receiving the model update data, processor 132 may interact with model blockchain 140 to write the model update data, write an updated computing model, and/or the like, as discussed further herein. For example, processor 132 may run applications, application programming interfaces (APIs), software development kits (SDKs), blockchain oracles, or the like to interact with model blockchain 140, communicate with other nodes, perform crypto-operations, consent with other nodes to writes on model blockchain 140, and otherwise operate within system 100. Beser at [0045].
transmitting, by the blockchain node, a blockchain data value included in one block of a plurality of blocks of the blockchain, wherein the blockchain data value includes the generated synthetic data set previously generated by the second computing system;
A blockchain can be used to record and provide accepted computing models and/or model updates to distributed applications of the computing model. The use of this shared and secure record of computing models ensures that any of the distributed applications may easily validate the received model. Beser at [0019].
receiving, by a receiver of the first computing system, from the blockchain network, the blockchain data value that includes the generated synthetic data set previously generated by the second computing system;
For example, the respective validation node 130-1, 130-2 may write model updating data, an update computing model, or the like to blockchain 140, as discussed further herein. The respective node 110-1, 110-2, 110-3 may retrieve the response transaction from model blockchain 140… Beser at [0033].
receiving, by the receiver of the first computing system, an expected accuracy value;
Node 110 generates model update data (step 304) based on the model update event. The model update data may comprise training data (e.g., data to be ingested by a node to “train” the computing model to cure the prediction error, detect the new model requirement, etc.). Beser at [0052].
For example, the prediction error may relate to an incorrect identification using the computing model, or may be associated with any other suitable or desired error. Beser at [0052].
The “prediction error” is analogous to an “expected accuracy value” because the prediction error is an indication of the accuracy of the existing machine learning model. The “prediction error” is received with the “model update data” and/or from the node that detected the “model update event.” Upon applying the model update data (in the next step), the accuracy of the updated machine learning model can be tested to determine whether the prediction error is reduced or eliminated.
generating, by a processor of the first computing system, a result value via application of the
Validation nodes 130 may also be configured to validate the updated computing model by locally testing the updated computing model to determine whether the updated computing model cures the prediction error, detects the new model requirement, or the like. For example, validation nodes 130 may locally implement the updated computing model, and may process data using the computing model to determine whether the prediction error still results, to detect detects the new model requirement, or the like. Beser at [0058].
determining, by the processor of the first computing system, a validity measurement for the machine learning model based on a comparison of the generated result value and the expected accuracy value.
As a further example, validation nodes 130 may test the model update data to identity false positives, false negatives, or the like. Beser at [0055].
“False positives” and “false negatives” are types of “validity measurements” that can be compared to the “prediction errors” to determine whether the accuracy is improved after the model update.
Beser does not appear to disclose:
transmitting, by the first computing system, to a second computing system within the system, a request for a synthetic data set
receiving,
using a generative adversarial network [to generate] the synthetic data set such that it approximates real data for application to the machine learning model but without using real data;
receiving,
said first computing system and said second computing system are off- chain systems;
Lally, which is analogous art to the claimed invention, discloses:
transmitting, by the first computing system, to a second computing system within the system, a request for a
One or more of the requesting computers requests access to the generated AI information useable by the one or more requesting computers to facilitate machine learning, model operation, and/or data aggregation with other computers across the computer network. Lally at [0007].
The “requesting computer” is analogous to “the first computing system.”
receiving, by the second computing system, the request from the first computing system;
The requesting computer requests from the one providing computer access to the AI information controlled by the one providing computer. Authorization capabilities are issued by the providing computer and used by the requesting computer to access the AI information. Then, the requesting computer can operate on or with the requested AI information to generate a result useable to facilitate machine learning, model operation, and/or data aggregation and store the result from the operating in the decentralized database. Lally at [0008].
The “providing computer” is analogous to the “second computing system.”
said first computing system and said second computing system are off-chain systems;
Each independently-controlled set of client-controlled services 4 may connect to and exchange information with, under the direction of the providing/requesting computer, other client-controlled services controlled by third party computers (labeled third-party client-controlled services 7 in Figure 1). Each independently-controlled set of client-controlled services 4 may subscribe to one or more blockchain nodes 5, each blockchain node 5 being a service controlled and operated by a single computer, group of computers, or a service provider 5. In some embodiments, each blockchain node 5 is configured and operated according to a set of policies approved by a subscribing client computer(s). Such policies may relate to network boundary configuration, system monitoring, validation and consensus settings, access control, etc., and the subscribers to a particular blockchain node may align within a common trust boundary, i.e., a members of a company or government organization, or citizens of a nation subscribing to a blockchain node operated by or on behalf of that organization or nation. Lally at [0034]-[0035].
As illustrated in Figure 3, blockchain node 5 is subscribed to by the provider and the requestor, but “external” to “Providing/Requesting Computer A and B.”
Lally is analogous art to the claimed invention because both are related to utilizing a blockchain to exchange machine learning model information between two entities. It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to combine the validation process of Beser with the separate blockchain, provider, and requestor systems of Lally to result in a system whereby the blockchain network is external to both the provider and requestor, and whereby the blockchain network facilitates data transfer based on requests and permission grants of the transacting parties. Such a system is beneficial because neither party is tasked with maintaining a blockchain network and further, the trustworthiness of the system is improved because the control of the blockchain is independent from all other parties.
Lally does not appear to disclose:
using a generative adversarial network [to generate] the synthetic data set such that it approximates real data for application to the machine learning model but without using real data;
Trost, which is analogous art to the claimed invention, discloses:
using a generative adversarial network [to generate] the synthetic data set such that it approximates real data for application to the machine learning model but without using real data
When the discriminator model is unable to distinguish between the classified real data and the adjusted generated data, the generator model is used to generate mock data for an application being tested. Trost at [0006].
Generative Adversarial Networks (GANs) may be used for an automated system to learn from test data and generate production like mock test data used for the purposes of application testing. The generator model 101 generate new data instances while the discriminator model 107 evaluates them for authenticity. Trost at [0039].
Trost is analogous art to the claimed invention because both are related to generating data that approximates real data but does not include the real data. It would have been obvious to a person having ordinary skill in the art, before the effective filing date of the claimed invention, to combine the generative adversarial network of Trost with the machine learning model validation process of Beser to result in a system that allows for a node to generate testing data that does not include real data and provide the testing data to a validation node to validate the accuracy of a model. Motivation to combine includes allowing for the testing of a model while maintaining the security of actual data. Further, accuracy of testing as models is improved by allowing for mock data that mimics real data to be tested when sufficient real data is not available for adequate testing.
Claim 2
Beser discloses:
wherein receiving the expected accuracy value further includes receiving a cryptographic key, and the method further comprises:
In various embodiments, each node 110-1, 110-2, 110-3, and/or validation node 130-1, 130-2 may each be assigned cryptographic keys (e.g., asymmetric keys) used to digitally sign and/or encrypt transmissions in system 100. Beser at [0048].
validating, by the processor of the first computing system, the data set using the cryptographic key.
For example, each node 110-1, 110-2, 110-3 may use its associated public key to digitally sign transactions (e.g., transmissions of model update data) in system 100. Each node 110-1, 110-2, 110-3, and/or validation node 130-1, 130-2 may also be configured to verify digital signatures, maintain integrity of Merkle tree, verify proof of work, and/or the like, similar to typical blockchain technologies. Beser at [0049].
“Verify digital signatures” of “transmissions of model update data” is analogous to “validating…the data set.”
Claim 3
Beser discloses:
wherein the expected accuracy value is received from the blockchain data value.
In various embodiments, the validation node may also propagate the model update data write to the model blockchain to at least the second validation node in the neural network. Beser at [0007].
The model update data may comprise testing data and/or an updated model. The model update data may be generated by the first node of the neural network based on the detection of a model update event in the computing model. The model update event may comprise a prediction error, a new model requirement, or the like. Beser at [0007].
The “model update data,” analogous to the “data set” that is written to the blockchain as the “data value,” can include the “prediction error,” which is analogous to the “expected accuracy value.”
Claim 4
Beser discloses:
wherein the expected accuracy value is received from the second computing system.
In various embodiments, the validation node may also propagate the model update data write to the model blockchain to at least the second validation node in the neural network. Beser at [0007].
The model update data may be generated by the first node of the neural network based on the detection of a model update event in the computing model. The model update event may comprise a prediction error, a new model requirement, or the like. Beser at [0007].
The “validation node” (i.e., first validation node 130-1) is analogous to the “second computing system,” which provides the “model update data” (which includes the expected accuracy value in the form of a “prediction error”) to the “second validation node.”
Claim 5
Beser discloses:
wherein the machine learning model is received from the second computing system prior to storage in the memory of the first computing system.
In various embodiments, the validation node may also propagate the model update data write to the model blockchain to at least the second validation node in the neural network. Beser at [0007].
The model update data may comprise …a new computing model. Beser at [0052].
The “model update data” can include a “new computing model,” which is received from the “first validation node” (i.e., “second computing system”) to be tested by the “second validation node” (i.e., “first computing system).
Claim 6
Beser discloses:
receiving, by the receiver of the first computing system, one or more additional variables from a second computing system; and
The model update data may comprise training data (e.g., data to be ingested by a node to “train” the computing model to cure the prediction error, detect the new model requirement, etc.), a model update (e.g., programmable code to merge with the computing model to cure the prediction error, detect the new model requirement, etc.), and/or a new computing model. Beser at [0052].
refining the determined validity measurement based on the one or more additional variables.
For example, validation nodes 130 may locally implement the test data and/or model update, and may process data using the computing model to determine whether the prediction error is still an issue, whether the new model requirement can be detected, or the like. As a further example, validation nodes 130 may test the model update data to identity false positives, false negatives, or the like. Testing may also be performed by a human to identify new scenarios and/or to validate the model update data. Beser at [0055].
The “additional variables” can include training data and/or a model update, which can be utilized to change the locally stored machine learning model to an improved version. The improved version of the machine learning model can then be tested to determine whether the model output is improved (i.e., “refined”) over the previous machine learning model.
Claim 8
Trost discloses:
wherein the data set is generated using a generative adversarial network.
At this point the generator model is now generating mock test data indiscernible from real data 109. In step 219, the method 200 provides the generator to a test environment where it can be used to generate mock test data for the given application to be tested. Trost at [0066].
Claim 9
Beser discloses:
A system…comprising:
See FIG. 1A, element 100
With reference to FIG. 1A, a system 100 for consensus and updating of neural network computing models using blockchain is depicted according to various embodiments. Beser at [0026].
a first computing system (See FIG. 1C, element 130) including
a first memory…, (FIG. 1C, element 134)
a first transmitter (FIG. 1C, element 136)
a first receiver…, and (FIG. 1C, element 136)
a first processor (FIG. 1C, element 132)
a second computing system (See FIG. 1B, element 110) including
a second receiver (FIG. 1B, element 116)
a second processor (FIG. 1B, element 112)
a second transmitter (FIG. 1B, element 116)
Further, claim 9 recites the components performing the method of claim 1. As previously asserted, claim 1 is rejected under 35 U.S.C. 103 as being obvious over Beser in view of Lally and Trost. Accordingly, for at least the same reasons, claim 9 is rejected under 35 U.S.C. 103.
Claims 10-14 and 16
Claims 10-14 and 16 recite the system recited in claim 9 performing substantially the same limitations as recited in claims 2-6 and 8. For at least the same reasons as asserted regarding rejection of claims 2-6 and 8, claims 10-14 and 16 are rejected under 35 U.S.C. 103 as being obvious over Beser in view of Lally and Trost.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Communication
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JOSEPH MORRIS whose telephone number is (703)756-5735. The examiner can normally be reached M-F 8:30-5:00.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Ryan Pitaro can be reached at (571) 272-4071. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
JOSEPH MORRIS
Examiner
Art Unit 2188
/JOSEPH P MORRIS/Examiner, Art Unit 2188
/RYAN F PITARO/Supervisory Patent Examiner, Art Unit 2188
1 Of additional note regarding Applicant’s argument, “first and second provider systems” is not recited in the independent claims.
Prosecution Insights