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 .
Status of Claims
Claims 1-20 were previously pending and subject to a non-final Office Action having a notification date of December 15, 2025 (“non-final Office Action”). Following the non-final Office Action, Applicant filed an amendment on March 5, 2026 (the “Amendment”), amending claims 1, 6, 8, 10, 13, and 15.
The present Final Office Action addresses pending claims 1-20 in the Amendment.
Response to Arguments
Response to Applicant’s Arguments Regarding Claim Rejections Under 35 USC §112
These rejections are withdrawn in view of the Amendment.
Response to Applicant’s Arguments Regarding Claim Rejections Under 35 USC §101
On page 16 of the Amendment, Applicant takes the position that claim 1 does not recite an abstract idea because "[the] recited "generates, via named-entity recognition or natural language processing, a semantic hierarchy of semantic objects of the plain text service manual" requires automated NER/NLP applied over the entire manual corpus to produce a machine data structure - a "semantic hierarchy of semantic objects" - which is a particular internal representation in memory, not a human-mental classification of chapters or headings, and the claim does not recite any mathematical relationships, formulas, or calculations." The Examiner disagrees that claim 1 does not recite an abstract idea.
As set forth in the rejection below, a person could practically in their mind with pen and paper access/observe "plain text" documentation (e.g., a service manual) that includes information/instructions for solving malfunctions of a defined type of machine (e.g., medical imaging scanner) and use natural language processing to identify a semantic hierarchy of objects in the documentation (e.g., via analyzing a hierarchical manner in which natural language in the documentation is arranged, such as chapters in an index, headings, etc.; classifying/categorizing named entities in text such as objects, parts, etc.). For instance, a person could readily review how certain chapters/sections in the service manual are related to each other via reviewing the table of contents and/or index, reading/reviewing the information in each section, etc., and record their findings where each chapter/section/page/paragraph/etc. can be equated to a different respective "semantic object" or node and where objects can be related in the form of a knowledge graph or tree structure.
Applicant's suggestions that the recited "plain text service manual" contains "large volumes of scattered troubleshooting information" and that the NER/NLP is applied "over the entire manual corpus to produce a machine data structure" are irrelevant because such features are not actually recited in the claims.
Furthermore, the Examiner disagrees with Applicant's analogization of the present claims to the claim in USPTO Example 39 at pages 16-17 of the Amendment. Initially, the claim in Example 39 does not even recite an abstract idea in the first place. Conversely, the present claims recite various mental processes as noted herein. Furthermore, while the present claims now recite how the DLNN is trained "using the semantic hierarchy of the semantic objects" of a plain text service manual including "information for solving malfunctions of the defined type of medical imaging scanner" to generate the navigation sequences through the semantic objects to trouble the malfunctions, the Examiner fails to understand how the semantic hierarchy actually facilitates medical imaging scanner malfunction troubleshooting (e.g., in a manner similar to how the digital facial images, modified facial images, and non-facial images in the first set facilitate digital facial recognition in Example 39).
On page 17 of the Amendment, Applicant then takes the position that "[the] "trains, using the sematic hierarchy of the semantic objects, a deep learning neural network to generate navigation sequences through the semantic objects of the plain text service manual to troubleshoot the malfunctions of the defined type of machine" configures a deep learning model to output ordered sequences over the semantic hierarchy conditioned on malfunction-related inputs, which entails high-dimensional internal parameter updates and sequence modeling that cannot be performed in the human mind at the claimed scale and complexity." However, the Examiner is not asserting that this limitation is performable in the human mind but instead amounts to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished which is equivalent to the words “apply it” (see MPEP § 2106.05(f)).
Applicant then reiterates the remaining limitations of the claim (the "iterative retraining" limitation and details thereof) and asserts "Under its broadest reasonable interpretation, this recites (1) electronic capture of GUI interaction sequences by an "electronic tracking device," (2) computation of a loss over sequences output by a deep learning neural network, and (3) automated update of trainable internal parameters-all of which are technical machine operations analogous to the two-stage training and error-focused dataset construction in Example 39, and none of which can be practically performed mentally as "observations/evaluations/judgments/analyses" in the sense contemplated in the Office Action."
However, the Examiner is not asserting that such limitations are "practically performed mentally as observations/evaluations/judgments/analyses" as alleged by Applicant. Respectfully, Applicant is confusing the issues by asserting that the Examiner is asserting certain limitations to be practically performable in the human mind when the Examiner is not asserting such limitations to be practically performable in the human mind in the first place. Instead, the Examiner is asserting that these limitations amount to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished which is equivalent to the words “apply it” (see MPEP § 2106.05(f)). Requirements that a machine learning model be “iteratively trained” or dynamically adjusted do not represent a technological improvement because iterative training using selected training material and dynamic adjustments based on real-time changes are incident to the very nature of machine learning. Recentive Analytics, Inc. v. Fox Corp., Fox Broadcasting Company, LLC, Fox Sports Productions, LLC, Case No. 23-2437, (Fed. Cir. 2025), p. 12. “[T]he way machine learning works is the inputs are defined, the model is trained, and then the algorithm is actually updated and improved over time based on the input.” Id.
Other than reciting that the plain text service manual (from which the semantic object hierarchy is developed) includes "information for solving malfunctions of the defined type of medical imaging scanner," there are not details regarding how such information facilitates solving of malfunctions of the defined type of medical imaging scanner. Accordingly, training the DLNN on the "semantic hierarchy of semantic objects" in the first training stage just amounts to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished which is equivalent to the words “apply it” (see MPEP § 2106.05(f)). Furthermore, while the iterative retraining in the recited second training stage determines an error based on the generated navigation sequence and the troubleshooting trace and then updates at least one trainable internal parameter "based on the error," it never specifies the specific training data used in the second training stage (e.g., in a manner similar to how the digital facial images, modified facial images, non-facial images, and non-facial incorrectly classified as facial in the first training stage in the second set facilitate digital facial recognition in Example 39).
On pages 17-18 of the Amendment, Applicant then asserts that a person could not record a troubleshooting trace through the semantic hierarchy of a service technician, that is troubleshooting a malfunction of a deployed machine for a service complaint of the service complaints, sequentially navigating through the semantic objects of the plain text service manual because "Real-world field service appointments can last hours and involve numerous fine-grained GUI interactions (e.g., clicks, scrolls, jumps between sections) across many semantic objects. It is not practical, at each appointment, to assign a second person to stand over the technician's shoulder and attempt to manually record, with pen and paper, every navigation move, map each GUI interaction back to a specific semantic object in the hierarchy, and do so continuously and accurately for the entire session across many different service calls. Such a hypothetical observer-based process would be cost-prohibitive, error-prone, and inherently limited to coarse notes (such as "chapter 2, then chapter 7-8"), which cannot serve as the precise, step-by-step, machine-readable sequence data needed as training targets for a deep learning neural network."
However, where do the claims recite "real-world field service appointments [that] can last hours and involve numerous fine-grained GUI interactions (e.g., clicks, scrolls, jumps between sections) across many semantic objects" or that the troubleshooting trace is recorded "continuously and accurately for [an] entire session across many different service calls"? They do not and Applicant is again unnecessarily clouding the issues by implying that they do. In contrast, the claims encompass a service technician merely sequentially navigating through only a few "semantic objects" with only a few interactions. As noted herein and in contrast to Applicant's tenuous assertion that the Examiner has presented an "unrealistic thought experiment," a person could practically record (e.g., watch/observe and write down) how a trained service technician troubleshooting a malfunction of the machine/medical imaging scanner for a service complaint sequentially navigates through the objects of the hierarchy (which could merely include just a few objects over any period of time such as a few seconds or minutes, for example), thereby yielding a "troubleshooting trace" that corresponds to the service complaint (e.g., writing down how the technician first navigated to objects corresponding to chapter 2 and then navigated to objects corresponding to chapters 7-8). The recited electronic tracking device and GUI of the display device just amount to using a computer or other machinery as tools performing their typical functionality in conjunction with performing the above-noted at least one abstract idea (see MPEP § 2106.05(f)).
At pages 18-19 of the Amendment, Applicant next asserts that the present claims integrate "any such concept" (mental processes as asserted by the Examiner) into a practical application that improves trained DLNN functionality and a specific technological field of medical imaging scanner troubleshooting and then recites substantially the entirety of the independent claim limitations. However, the Examiner asserts that many of the limitations asserted by Applicant are part of the abstract idea while the remaining limitations do not provide a "practical application" of the abstract idea as already discussed herein.
Applicant's assertion of alleged similarity between the present claims and the claim in Example 39 at the bottom of page 19 has already been addressed herein.
Applicant's position at page 20 of the Amendment that the present claims do not amount to using ML in a new environment because they recite particulars such as navigation sequences over a semantic hierarchy, a service complaint, a troubleshooting trace, and a loss between the model's sequence and the trace actually support the Examiner's position that the present claims do just amount to using ML in a new environment. That is, training and retraining an ML model via updating trainable internal parameters based on error between certain "ground truth" and outputs from the ML model is central to almost all types of ML techniques. Accordingly, that Applicant recites performing the training with a "semantic hierarchy of semantic objects" and performing the retraining where the ground truth happens to be a "troubleshooting trace" through the hierarchy by a service technician does just amount to using ML in a new environment (i.e., the new environment being machine troubleshooting). There is nothing in the present specification supporting Applicant's continued assertion that above limitations are "engineered to correct a specific failure mode - erroneous navigation sequences when trained only on plain text service manuals - rather than simply updating a model as new generic data arrives." That is, the present invention is not directed to improving an existing ML technique but rather is just automating existing techniques for troubleshooting machine malfunctions via a user manually sifting through a service manual (e.g., [0002], [0025] of the application).
Applicant then asserts that the recited navigation sequences over a semantic hierarchy, service complaint, troubleshooting trace, and loss between the model's sequence and the trace "is a particular solution to a problem…that improves a "technical field" of troubleshoot of medical imaging scanners" allegedly consistent with the Kim memo. The Examiner disagrees and asserts that the "additional limitations" do not provide a "practical application" as already discussed herein.
The Examiner disagrees with Applicant's analogization to Ex parte Desjardins at page 20 of the Amendment. Desjardins finds that improvements in training the ML model itself can provide a practical application of the abstract idea recited in the claims rather than improvements in the abstract idea. For instance, page 9 of Desjardins notes how the specification describes improvements in the training of the ML model itself such as "effectively [learning] new tasks in succession whilst protecting knowledge about previous tasks" and "[allowing] artificial intelligence (AI) systems to 'us[e] less of their storage capacity' and [enabling] 'reduced system complexity'." It is then noted how at least "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" as recited in the independent claim reflects the above-noted improvement described in the specification. Id. Notably, the PTAB indicates they "are persuaded that [the above claim limitation] constitutes an improvement to how the machine learning model itself operates, and not, for example, the identified mathematical calculation [(abstract idea)]." (Emphasis added). Id.
Unlike in Desjardins, Applicant has provided no evidence that actual existing ML training techniques are being improved other than the attorney/agent argumentation that that above limitations are "engineered to correct a specific failure mode - erroneous navigation sequences when trained only on plain text service manuals - rather than simply updating a model as new generic data arrives" which is not even supported in the present application in the first place.
Applicant then asserts at page 21 of the Amendment that claim 1 integrates any alleged judicial exception into a concrete, real-world feedback loop that enhances the ML model's navigation sequence generation and the technical process of scanner troubleshooting. However, "scanner troubleshooting" is part of the mental process-type abstract idea and is thus not a "technical process" and the loss-based retraining/feedback loop is "the way machine learning works." Recentive Analytics, Inc. v. Fox Corp., Fox Broadcasting Company, LLC, Fox Sports Productions, LLC, Case No. 23-2437, (Fed. Cir. 2025), p. 12.
In relation to step 2B, Applicant reiterates the assertion that the present claims are engineered to overcome a "known deficiency" of DL models trained only on plain text manuals and is not simply the application of established methods of ML to a new data environment. The Examiner disagrees as already set forth herein.
On page 22 of the Amendment, Applicant appears to assert that the Recentive claims just generically recited iterative ML updating in the context of event scheduling while the present claims instead "confines the model's improvement mechanism" to a specialized domain representation, second-stage training data, and objective thereby ""changing the operation of the model" in a way that is both domain-specific and architecture-specific." Again, training and retraining an ML model via updating trainable internal parameters based on error between certain "ground truth" and outputs from the ML model is central to almost all types of ML techniques. Accordingly, that Applicant recites performing the training with a "semantic hierarchy of semantic objects" and performing the retraining where the ground truth happens to be a "troubleshooting trace" through the hierarchy by a service technician does just amount to using ML in a new environment (i.e., the new environment being machine troubleshooting). There is nothing in the present specification supporting Applicant's continued assertion that above limitations are "engineered to correct a specific failure mode - erroneous navigation sequences when trained only on plain text service manuals - rather than simply updating a model as new generic data arrives." That is, the present invention is not directed to improving an existing ML technique but rather is just automating existing techniques for troubleshooting machine malfunctions via a user manually sifting through a service manual (e.g., [0002], [0025] of the application).
Applicant then improperly analyzes the recording of the troubleshooting trace under step 2B even though the majority of such limitation was already indicated as being part of the abstract idea as discussed herein. That the trace is recorded with the electronic recording device of a display just amounts to using computers as tools to carry out the abstract idea under MPEP 2106.05(f).
Claims 1-20 are not "clearly patent eligible" as asserted by Applicant on page 23 of the Amendment and the 35 USC 101 rejection is maintained.
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 1-20 are rejected under 35 U.S.C. §101 because the claimed invention is directed to an abstract idea without significantly more:
Subject Matter Eligibility Criteria - Step 1:
Claims 1-7 are directed to a system (i.e., a machine), claims 8-14 are directed to a method (i.e., a process), and claims 15-20 are directed to a non-transitory computer-readable medium (i.e., a manufacture). Accordingly, claims 1-20 are all within at least one of the four statutory categories. 35 USC §101.
Subject Matter Eligibility Criteria - Alice/Mayo Test: Step 2A - Prong One:
Regarding Prong One of Step 2A of the Alice/Mayo test (which collectively includes the guidance in the January 7, 2019 Federal Register notice and the October 2019 and July 2024 updates issued by the USPTO as incorporated into the MPEP, as supported by relevant case law), the claim limitations are to be analyzed to determine whether, under their broadest reasonable interpretation, they “recite” a judicial exception or in other words whether a judicial exception is “set forth” or “described” in the claims. MPEP 2106.04(II)(A)(1). An “abstract idea” judicial exception is subject matter that falls within at least one of the following groupings: a) certain methods of organizing human activity, b) mental processes, and/or c) mathematical concepts. MPEP 2106.04(a).
Representative independent claim 15 includes limitations that recite at least one abstract idea. Specifically, independent claim 15 recites:
A non-transitory computer-readable medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:
access a plain text service manual associated with a defined type of machine;
generate, via named-entity recognition or natural language processing, a semantic hierarchy of semantic objects of the plain text service manual, wherein the one or more plain text documents comprise information for solving malfunctions of the defined type of medical imaging scanner;
train, using the sematic hierarchy of the semantic objects, a deep learning neural network to generate navigation sequences through the semantic objects of the plain text service manual to troubleshoot the malfunctions of the defined type of machine;
iteratively retrain the deep learning neural network to generate the navigation sequences through the semantic objects of the plain text service manual to troubleshoot the malfunctions of the defined type of machine, wherein the iterative retraining comprises, for service complaints from users of malfunctions of deployed machines of the defined type:
recording, using an electronic tracking device of a display device, a troubleshooting trace through the semantic hierarchy of a service technician, that is troubleshooting a malfunction of a deployed machine for a service complaint of the service complaints, sequentially navigating through the semantic objects of the plain text service manual displayed on a graphical user interface of the display device to resolve the malfunction, and
retraining the deep learning neural network using the service complaint and the troubleshooting trace, wherein the retraining comprises:
determining, using a loss function, an error based on the troubleshooting trace and an output navigation sequence of the deep learning neural network using the service complaint, and
updating one or more trainable internal parameters of the deep learning neural network based on the error.
The Examiner submits that the foregoing underlined limitations constitute “mental processes” because they are observations/evaluations/judgments/analyses that can, at the currently claimed high level of generality, be practically performed in the human mind (e.g., with pen and paper). As an example, a person could practically in their mind with pen and paper access/observe "plain text" documentation (e.g., a service manual) that includes information/instructions for solving malfunctions of a defined type of machine (e.g., medical imaging scanner) and use natural language processing to identify a semantic hierarchy of objects in the documentation (e.g., via analyzing a hierarchical manner in which natural language in the documentation is arranged, such as chapters in an index, headings, etc.; classifying/categorizing named entities in text such as objects, parts, etc.). For instance, a person could readily review how certain chapters/sections in the service manual are related to each other via reviewing the table of contents and/or index, reading/reviewing the information in each section, etc., and record their findings where each chapter/section/page/paragraph/etc. can be equated to a different respective "semantic object" or node and where objects can be related in the form of a knowledge graph or tree structure. The person could also readily record (e.g., watch/observe and write down) how a trained service technician troubleshooting a malfunction of the machine/medical imaging scanner for a service complaint sequentially navigates through the objects of the hierarchy, thereby yielding a "troubleshooting trace" that corresponds to the service complaint (e.g., writing down how the technician first navigated to objects corresponding to chapter 2 and then navigated to objects corresponding to chapters 7-8).
These recitations, under their broadest reasonable interpretation, are similar to the concepts of collecting information, analyzing it and displaying certain results of the collection and analysis in Electric Power Group, LLC, v. Alstom (830 F.3d 1350, 119 USPQe2d 1739 (Fed. Cir. 2016)). MPEP 2106.04(a)(2)(III).
Claims “directed to collection of information, comprehending the meaning of that collected information, and indication of the results, all on a generic computer network operating in its normal, expected manner,” fail step one of the Alice framework. In re Killian, 45 F.4th 1373, 1380 (Fed. Cir. 2022). Claims directed to “collecting, analyzing, manipulating, and displaying data’’ are abstract. Univ. of Fla. Research Found., Inc. v. General Elec. Co., 916 F.3d 1363, 1368 (Fed. Cir. 2019). Claims directed to organizing, storing, and transmitting information determined to be directed to an abstract idea. Cyberfone Sys., L.L.C. v. CNN Interactive Grp., Inc., 558 F. App’x 988, 992 (Fed. Cir. 2014).
Accordingly, the claim recites at least one abstract idea.
Furthermore, dependent claims 3, 7, 10, 14, 17, and 18 further define the at least one abstract idea (and thus fail to make the abstract idea any less abstract) as set forth below:
-Claims 3 and 10 calls for accessing an operational report of the scanner and updating the semantic hierarchy based on sematic objects of the report, where the recording includes recording how the technician navigates through the objects of the manual and the report to resolve the complaint. All of these limitations are practically performable in the human mind with pen and paper ("mental processes").
-Claims 7 and 14 recite how a node of the semantic hierarchy is based on a keyword located in the plain text service manual, the keyword being associated with an electronic action performable by the medical imaging scanner. These limitations just further define the semantic hierarchy that is practically identifiable in the human mind with pen and paper ("mental processes").
-Claims 17-18 recite how the plain text documents include a service manual of the defined type of machine or an operational report produced by the machine which just further define the documents that are mentally reviewable by a person ("mental processes").
Subject Matter Eligibility Criteria - Alice/Mayo Test: Step 2A - Prong Two:
Regarding Prong Two of Step 2A of the Alice/Mayo test, it must be determined whether the claim as a whole integrates the abstract idea into a practical application. As noted at MPEP §2106.04(II)(A)(2), it must be determined whether any additional elements in the claim beyond the abstract idea integrate the exception into a practical application in a manner that imposes a meaningful limit on the judicial exception. The courts have indicated that additional elements such as merely using a computer to implement an abstract idea, adding insignificant extra solution activity, or generally linking use of a judicial exception to a particular technological environment or field of use do not integrate a judicial exception into a “practical application.” MPEP §2106.05(I)(A).
In the present case, the additional limitations beyond the above-noted at least one abstract idea recited in the claim are as follows (where the bolded portions are the “additional limitations” while the underlined portions continue to represent the at least one “abstract idea”):
A non-transitory computer-readable medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to (using computers or machinery as mere tools to perform the abstract idea as noted below, see MPEP § 2106.05(f)):
access a plain text service manual associated with a defined type of machine;
generate, via named-entity recognition or natural language processing, a semantic hierarchy of semantic objects of the plain text service manual, wherein the one or more plain text documents comprise information for solving malfunctions of the defined type of medical imaging scanner;
train, using the sematic hierarchy of the semantic objects, a deep learning neural network to generate navigation sequences through the semantic objects of the plain text service manual to troubleshoot malfunctions of the defined type of machine (merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished, see MPEP § 2106.05(f));
iteratively retrain the deep learning neural network to generate the navigation sequences through the semantic objects of the plain text service manual to troubleshoot the malfunctions of the defined type of machine, wherein the iterative retraining comprises, for service complaints from users of malfunctions of deployed machines of the defined type (merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished, see MPEP § 2106.05(f)):
recording, using an electronic tracking device of a display device (using computers or machinery as mere tools to perform the abstract idea as noted below, see MPEP § 2106.05(f)), a troubleshooting trace through the semantic hierarchy of a service technician, that is troubleshooting a malfunction of a deployed machine for a service complaint of the service complaints, sequentially navigating through the semantic objects of the plain text service manual displayed on a graphical user interface of the display device (using computers or machinery as mere tools to perform the abstract idea as noted below, see MPEP § 2106.05(f)) to resolve the malfunction, and
retraining the deep learning neural network using the service complaint and the troubleshooting trace, wherein the retraining comprises (merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished, see MPEP § 2106.05(f)):
determining, using a loss function, an error based on the troubleshooting trace and an output navigation sequence of the deep learning neural network using the service complaint (merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished, see MPEP § 2106.05(f)), and
updating one or more trainable internal parameters of the deep learning neural network based on the error (merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished, see MPEP § 2106.05(f)).
For the following reasons, the Examiner submits that the above-identified additional limitations, when considered as a whole with the limitations reciting the at least one abstract idea, do not integrate the above-noted at least one abstract idea into a practical application.
Regarding the additional limitations of the non-transitory computer readable medium, instructions, processor, electronic tracking device of the display device, and GUI, the Examiner submits that these limitations amount to merely using a computer or other machinery as tools performing their typical functionality in conjunction with performing the above-noted at least one abstract idea (see MPEP § 2106.05(f)).
Regarding the additional limitations of training a deep learning NN with the semantic hierarchy to generate "navigation sequences" through the objects (e.g., starting at object #2 and proceeding to object #4 and then, if necessary, to object #10) to troubleshoot the malfunctions of the defined type of machine (medical scanner), and retraining the deep learning NN via determining an error between the troubleshooting trace of the technician and an output navigation sequence of the deep learning NN using an input service complaint and updating trainable internal parameters based on the error, the Examiner submits that these limitations amount to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished which is equivalent to the words “apply it” (see MPEP § 2106.05(f)).
Requirements that a machine learning model be “iteratively trained” or dynamically adjusted do not represent a technological improvement because iterative training using selected training material and dynamic adjustments based on real-time changes are incident to the very nature of machine learning. Recentive Analytics, Inc. v. Fox Corp., Fox Broadcasting Company, LLC, Fox Sports Productions, LLC, Case No. 23-2437, (Fed. Cir. 2025), p. 12. “[T]he way machine learning works is the inputs are defined, the model is trained, and then the algorithm is actually updated and improved over time based on the input.” Id.
Instead of disclosing “a specific implementation of a solution to a problem in the software arts,” Enfish, LLC v. Microsoft Corp., 822 F.3d 1327, 1339 (Fed. Cir. 2016), or “a specific means or method that solves a problem in an existing technological process,” Koninklijke, 942 F.3d at 1150, the only thing the claims disclose about the use of machine learning is that machine learning is used in a new environment. Recentive Analytics, Inc. v. Fox Corp., Fox Broadcasting Company, LLC, Fox Sports Productions, LLC, Case No. 23-2437, (Fed. Cir. 2025), p. 13. In Recentive, the new environment was event scheduling and the creation of network maps. Id., p. 13. Similarly, in the present case, the new environment is machine (e.g., medical scanner) troubleshooting and the creation of navigation sequences through plain text service manuals.
Claims that do no more than apply established methods of machine learning to a new data environment are not patent eligible. Id., pp. 10, 14. An abstract idea does not become nonabstract by limiting the invention to a particular field of use or technological environment. Id.
Thus, taken alone, the additional elements do not integrate the at least one abstract idea into a practical application. Furthermore, looking at the additional limitations as an ordered combination adds nothing that is not already present when looking at the elements taken individually. MPEP §2106.05(I)(A) and §2106.04(II)(A)(2).
For these reasons, representative independent claim 15 and analogous independent claims 1 and 8 do not recite additional elements that integrate the judicial exception into a practical application. Accordingly, representative independent claim 15 and analogous independent claims 1 and 8 are directed to at least one abstract idea.
The remaining dependent claim limitations not addressed above fail to integrate the abstract idea into a practical application as set forth below:
-Claims 2, 9, and 16 recite how the troubleshooting trace is employed during the retraining as ground truth for the malfunction troubleshooting which again amounts to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished which is equivalent to the words “apply it” (see MPEP § 2106.05(f)). Again, requirements that the machine learning model be “iteratively trained” or dynamically adjusted do not represent a technological improvement because iterative training using selected training material and dynamic adjustments based on real-time changes are incident to the very nature of machine learning. Recentive Analytics, Inc. v. Fox Corp., Fox Broadcasting Company, LLC, Fox Sports Productions, LLC, Case No. 23-2437, (Fed. Cir. 2025), p. 12.
-Claims 4, 11, and 19 recite how the electronic tracking device performs click-tracking, scroll-tracking, and/or eye-movement-tracking which does no more than generally link use of the abstract idea to a particular technological environment or field of use without adding an inventive concept to the abstract idea (see MPEP § 2106.05(h)).
-Claims 5 and 12 recite how the service complaint is employed during the retraining as a training input for the deep learning NN which again amounts to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished which is equivalent to the words “apply it” (see MPEP § 2106.05(f)). Again, requirements that the machine learning model be “iteratively trained” or dynamically adjusted do not represent a technological improvement because iterative training using selected training material and dynamic adjustments based on real-time changes are incident to the very nature of machine learning. Recentive Analytics, Inc. v. Fox Corp., Fox Broadcasting Company, LLC, Fox Sports Productions, LLC, Case No. 23-2437, (Fed. Cir. 2025), p. 12.
-Claims 6, 13, and 20 call for deploying the retrained deep learning NN as a third-party service which amounts to using computers as tools to perform an existing process at such high level of generality and merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished (see MPEP § 2106.05(f)). Furthermore, these claims call for receiving a complaint from a third-party computing device and transmitting an inferred trace to the third-party device which represents insignificant extra-solution activity (see MPEP § 2106.05(g)). Finally, these claims call for executing the deep learning neural network on the third-party service complaint to yield an inferred navigation sequence that represents a sequential reading path through the one or more plain text documents predicted to solve or address the third-party service complaint. These limitations again amount to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished (see MPEP § 2106.05(f)). These additional limitations provide only a result-oriented solution and lack details as to how execution of the deep learning NN actually occurs.
-Claims 7 and 14 recite how clicking on or invoking the node or the keyword causes the medical imaging scanner to automatically perform the electronic action which amounts to using computers as tools to perform an existing process at such high level of generality and merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished (see MPEP § 2106.05(f)).
When the above additional limitations are considered as a whole along with the limitations directed to the at least one abstract idea, the at least one abstract idea is not integrated into a practical application. Therefore, the claims are directed to at least one abstract idea.
Subject Matter Eligibility Criteria - Alice/Mayo Test: Step 2B:
Regarding Step 2B of the Alice/Mayo test, representative independent claim 15 does not include additional elements (considered both individually and as an ordered combination) that are sufficient to amount to significantly more than the judicial exception for reasons the same as those discussed above with respect to determining that the claim does not integrate the abstract idea into a practical application.
Regarding the additional limitations of the non-transitory computer readable medium, instructions, processor, electronic tracking device of a display device, and GUI, the Examiner submits that these limitations amount to merely using a computer or other machinery as tools performing their typical functionality in conjunction with performing the above-noted at least one abstract idea (see MPEP § 2106.05(f)).
Regarding the additional limitations of training a deep learning NN with the semantic hierarchy to generate "navigation sequences" through the objects (e.g., starting at object #2 and proceeding to object #4 and then, if necessary, to object #10) to troubleshoot the malfunctions of the defined type of machine (medical scanner), and retraining the deep learning NN via determining an error between the troubleshooting trace of the technician and an output navigation sequence of the deep learning NN using an input service complaint and updating trainable internal parameters based on the error, the Examiner submits that these limitations amount to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished which is equivalent to the words “apply it” (see MPEP § 2106.05(f)).
Requirements that the machine learning model be “iteratively trained” or dynamically adjusted do not represent a technological improvement because iterative training using selected training material and dynamic adjustments based on real-time changes are incident to the very nature of machine learning. Recentive Analytics, Inc. v. Fox Corp., Fox Broadcasting Company, LLC, Fox Sports Productions, LLC, Case No. 23-2437, (Fed. Cir. 2025), p. 12. “[T]he way machine learning works is the inputs are defined, the model is trained, and then the algorithm is actually updated and improved over time based on the input.” Id.
Instead of disclosing “a specific implementation of a solution to a problem in the software arts,” Enfish, LLC v. Microsoft Corp., 822 F.3d 1327, 1339 (Fed. Cir. 2016), or “a specific means or method that solves a problem in an existing technological process,” Koninklijke, 942 F.3d at 1150, the only thing the claims disclose about the use of machine learning is that machine learning is used in a new environment. Recentive Analytics, Inc. v. Fox Corp., Fox Broadcasting Company, LLC, Fox Sports Productions, LLC, Case No. 23-2437, (Fed. Cir. 2025), p. 13. In Recentive, the new environment was event scheduling and the creation of network maps. Id., p. 13. Similarly, in the present case, the new environment is machine (e.g., medical scanner) troubleshooting and the creation of navigation sequences through plain text service manuals.
Claims that do no more than apply established methods of machine learning to a new data environment are not patent eligible. Id., pp. 10, 14. An abstract idea does not become nonabstract by limiting the invention to a particular field of use or technological environment. Id.
The dependent claims also do not include additional elements (considered both individually and as an ordered combination) that are sufficient to amount to significantly more than the judicial exception for the same reasons to those discussed above with respect to determining that the dependent claims do not integrate the at least one abstract idea into a practical application.
-Claims 2, 9, and 16 recite how the troubleshooting trace is employed during the retraining as ground truth for the malfunction troubleshooting which again amounts to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished which is equivalent to the words “apply it” (see MPEP § 2106.05(f)). Again, requirements that the machine learning model be “iteratively trained” or dynamically adjusted do not represent a technological improvement because iterative training using selected training material and dynamic adjustments based on real-time changes are incident to the very nature of machine learning. Recentive Analytics, Inc. v. Fox Corp., Fox Broadcasting Company, LLC, Fox Sports Productions, LLC, Case No. 23-2437, (Fed. Cir. 2025), p. 12.
-Claims 4, 11, and 19 recite how the electronic tracking device performs click-tracking, scroll-tracking, and/or eye-movement-tracking which does no more than generally link use of the abstract idea to a particular technological environment or field of use without adding an inventive concept to the abstract idea (see MPEP § 2106.05(h)).
-Claims 5 and 12 recite how the service complaint is employed during the retraining as a training input for the deep learning NN which again amounts to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished which is equivalent to the words “apply it” (see MPEP § 2106.05(f)). Again, requirements that the machine learning model be “iteratively trained” or dynamically adjusted do not represent a technological improvement because iterative training using selected training material and dynamic adjustments based on real-time changes are incident to the very nature of machine learning. Recentive Analytics, Inc. v. Fox Corp., Fox Broadcasting Company, LLC, Fox Sports Productions, LLC, Case No. 23-2437, (Fed. Cir. 2025), p. 12.
-Claims 6, 13, and 20 call for deploying the retrained deep learning NN as a third-party service which amounts to using computers as tools to perform an existing process at such high level of generality and merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished (see MPEP § 2106.05(f)). Furthermore, these claims call for receiving a complaint from a third-party computing device and transmitting an inferred trace to the third-party device which represents insignificant extra-solution activity (see MPEP § 2106.05(g)). Still further, the Examiner has reevaluated these limitations and determined such limitations to not be unconventional as they merely consist of receiving/transmitting data over a network. See Intellectual Ventures I v. Symantec Corp., 838 F.3d 1307, 1321, 120 USPQ2d 1353, 1362 (Fed. Cir. 2016); See MPEP 2106.05(d)(II). Finally, these claims call for executing the deep learning neural network on the third-party service complaint to yield an inferred navigation sequence that represents a sequential reading path through the one or more plain text documents predicted to solve or address the third-party service complaint. These limitations again amount to merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished (see MPEP § 2106.05(f)). These additional limitations provide only a result-oriented solution and lack details as to how execution of the deep learning NN actually occurs.
-Claims 7 and 14 recite how clicking on or invoking the node or the keyword causes the medical imaging scanner to automatically perform the electronic action which amounts to using computers as tools to perform an existing process at such high level of generality and merely reciting the idea of a solution or outcome without reciting details of how a solution to a problem is accomplished (see MPEP § 2106.05(f)).
Therefore, claims 1-20 are ineligible under 35 USC §101.
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.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JONATHON A. SZUMNY whose telephone number is (303) 297-4376. The examiner can normally be reached Monday-Friday 7-5.
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, Jason Dunham, can be reached at 571-272-8109. 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.
/JONATHON A. SZUMNY/Primary Examiner, Art Unit 3686