DETAILED ACTION
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Continued Examination
2. A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 19 March 2026 [hereinafter Response] has been entered, where:
Claims 1, 7, 8, and 14 have been amended.
Claims 2-6, 9-13, and 15-23 have been cancelled.
New claims 24, 25, 26, and 27 are presented for examination.
Claims 1, 7, 8, 14, and 24-27 are pending.
Claims 1, 7, 8, 14, and 24-27 are rejected.
Claim Rejections - 35 U.S.C. § 101
3. 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.
4. Claims 1, 7, 8, 14 and 24-27 are rejected under 35 U.S.C. § 101 because the claimed invention is directed to an abstract idea without significantly more.
Claim 1 recites a method, which is a process, and thus one of the statutory categories of patentable subject matter. (35 U.S.C. § 101).
However, under Step 2A Prong One, the claim recites the limitations of “[(a)] selecting . . . in a multi-objective optimization, a plurality of metrics for which hyperparameters of a machine-learning model are to be tuned, each metric of the plurality of metrics being configured for a specific objective and associated with a plurality of specification parameters including a target score of the metric, a penalty factor of the metric, and a bonus factor of the metric,” “[(b)] assigning, by the digital assistant builder platform, a first weight to each metric of the plurality of metrics,” “[(c)] configuring the sets of plurality of specification parameters for each metric of the plurality of metrics in accordance with at least one criterion,” “[(d)] evaluating the machine-learning model using one or more validation datasets to obtain a metric score for each metric of the plurality of metrics,” “[(e)] formulating, by the digital assistant builder platform, a weighted loss function using second weights for the plurality of metrics in an asymmetric loss technique based on a difference between the metric score and the target score of each metric, and the penalty factor or the bonus factor,” “[(f)] establishing, by the digital assistant builder platform, one or more constraints for the machine-learning model based on one or more hyperparameters,” and “[(g)] iteratively tuning, by the digital assistant builder platform, the hyperparameters associated with the machine-learning model to obtain a minimum value of the weighted loss function across the plurality of metrics.” These activities of “[(a)] selecting,” [(b)] assigning,” “[(c)] configuring,” “[(d)] evaluating” “[(e)] formulating,” “[(f)] establishing . . . constraints,” and “[(g)] tuning” contain limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, recite a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP § 2106.04(a)(2)).
The claim provides more details or specifics to the abstract idea of “[(a)] selecting a plurality of metrics,” “[(a.1)] wherein the plurality of metrics is selected to include a regression error of the machine-learning model and at least one from among a size of the machine-learning model, a training time of the machine-learning model, an accuracy of the machine-learning model, a stability of the machine-learning model, and a confidence score of the machine-learning model,” and accordingly, is merely more specific to the abstract idea.
The claim recites more details or specifics to the abstract idea of “[(b)] assigning,” “[(b.1)] wherein the first weight specifies an importance of the metric in evaluating a performance of the machine-learning model,” and “[(c)] configuring,” in that “[(c.1)] the configuring further comprises: assigning , for a corresponding set of the plurality of specification parameters associated with the metric for the regression error, a first target score, a first penalty factor value that is a relatively high value, and a first bonus factor value that is a relatively low value lower than the first penalty factor value, for minimizing the regression error of the machine-learning model,” and “[(e)] formulating” “[(e.1)] wherein the penalty factor and the bonus factor are related to a second weight associated with each metric, and, in the asymmetric loss technique, the penalty factor is used as a higher value for the second weight associated with the metric failing the target value and the bonus factor is used as a lower value for the second weight associated with the metric exceeding the target value,” and accordingly, are merely more specific to the abstract idea.
The claim also recites more details or specifics to the abstract idea of “[(f)] establishing . . . constraints,” “[(f.1)] wherein the one or more constraints include at least one from among a maximum model size, a maximum inference latency, and a maximum training time,” and accordingly, is merely more specific to the abstract idea.
The claim also recites more details or specifics to the abstract idea of “[(g)] iteratively tuning the hyperparameters,” “comprises: [(g)](a) selecting initial values for the hyperparameters, the machine-learning model being configured based on the initial values; [(g)](b) evaluating the weighted loss function using the initial values of the hyperparameters; [(g)](c) determining whether the weighted loss function is optimized based on the evaluating; and [(g)](d) in response to the weighted loss function not being optimized, searching for new values for the hyperparameters, reconfiguring the machine-learning model with the new values, and repeating steps (b) and (c) using the reconfigured machine-learning model), and accordingly, are merely more specific to the abstract idea.
The claim also recites the limitations of “[(k.1)] in response to determining, for the regression error, that the metric score is greater than the first target score, assigning to the difference between the metric score and the first target score, the first bonus factor value as a weight associated with the metric, and [(k.2)] in response to determining that the metric score is lower than the first target score, assigning to the difference between the first target score and the metric score, the first penalty factor value as the weight associated with the metric,” which contain limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, are a mental process, (MPEP § 2106.04(a)(2) sub III), being one of the groupings of abstract ideas. (MPEP § 2106.04(a)(2)). Thus, claim 1 recites an abstract idea.
Under Step 2A Prong Two, the claim as a whole is not integrated into a practical application, because the additional elements recited in the claim beyond the identified abstract idea include “a machine learning model,” “a digital assistant builder platform,” and “a digital assistant chatbot system (DACS),” and a “network,” which are recited at a high-level of generality, and accordingly, are generic computer components in which the abstract ideas are applied, (MPEP § 2106.05(f)), and does not integrate the abstract idea into a practical application.
The claim also recites the limitation of “[(h)] in response to the weighted loss function being optimized and each constraint being satisfied, generating, by the digital assistant builder platform, a validated machine-learning model comprising the tuned hyperparameters, with values for the hyperparameters obtained in a last iteration,” which is the use of the generic computer component (by the digital assistant builder platform) to implement the abstract idea, (MPEP § 2106.05(f)), that does not serve to integrate the abstract idea into a practical application.
The claim further recites the limitations of “[(i)] communicating, via a network, by the digital assistant builder platform to a digital assistant chatbot system (DACS) that is configured for a set of predetermined intents, the validated machine-learning model,” and “[(j)] providing by the DACS as an input to the validated machine-learning model, an input utterance received from a user,” where the activities of “[(i)] communicating” and [(j)] providing” are insignificant extra-solution activities of mere receiving of data, (MPEP § 2106.05(g)), that does not serve to integrate the abstract idea into a practical application. The claim recites more details or specifics of the additional element of “[(h), (i)] providing,” “[(h.1)] thereby causing the DACS to incorporate the validated machine-learning model,” and “[(i.1)] wherein, based on the input utterance, the validated machine-learning model is configured to output a prediction of an intent from the set of predetermined intents,” and accordingly, are merely more specific to the additional elements. Therefore, claim 1 is directed to the abstract idea. Therefore, claim 1 is directed to the abstract idea.
Finally, under Step 2B, the additional elements, taken alone or in combination, do not represent significantly more than the abstract idea itself. The claim recites additional elements beyond the identified abstract idea include “a machine learning model,” and “a digital assistant chatbot system (DACS),” which are recited at a high-level of generality, and accordingly, are generic computer components in which the abstract ideas are applied, (MPEP § 2106.05(f)), and does not amount to significantly more than the abstract idea. The claim also recites the limitation of “[(h)] in response to the weighted loss function being optimized and each constraint being satisfied, generating, by the digital assistant builder platform, a validated machine-learning model comprising the tuned hyperparameters, with values for the hyperparameters obtained in a last iteration,” which is the use of the generic computer component (by the digital assistant builder platform) to implement the abstract idea, (MPEP § 2106.05(f)), that does not amount to significantly more than the abstract idea.
The claim further recites the limitations of “[(i)] communicating, via a network, by the digital assistant builder platform to a digital assistant chatbot system (DACS) that is configured for a set of predetermined intents, the validated machine-learning model,” and “[(j)] providing by the DACS as an input to the validated machine-learning model, an input utterance received from a user,” where the activities of “[ (i)] communicating,” and [(j)] providing” are well-understood, routine, and conventional activities of receiving or transmitting data over a network, (MPEP § 2106.05(d) sub II.i), that does not amount to significantly more than the abstract idea. The claim recites more details or specifics of the additional element of “[(h), (i)] providing,” “[(h.1)] thereby causing the DACS to incorporate the validated machine-learning model,” and “[(i.1)] wherein, based on the input utterance, the validated machine-learning model is configured to output a prediction of an intent from the set of predetermined intents,” and accordingly, are merely more specific to the additional elements. Thus, claim 1 is subject-matter ineligible.
Claim 8 recites a system, which is a machine, and thus one of the statutory categories of patentable subject matter. (35 U.S.C. § 101).
However, under Step 2A Prong One, the claim recites the limitations of “[(a)] selecting . . . in a multi-objective optimization, a plurality of metrics for which hyperparameters of a machine-learning model are to be tuned, each metric of the plurality of metrics being configured for a specific objective and associated with a plurality of specification parameters including a target score of the metric, a penalty factor of the metric, and a bonus factor of the metric,” “[(b)] assigning, by the digital assistant builder platform, a first weight to each metric of the plurality of metrics,” “[(c)] configuring the sets of plurality of specification parameters for each metric of the plurality of metrics in accordance with at least one criterion,” “[(d)] evaluating the machine-learning model using one or more validation datasets to obtain a metric score for each metric of the plurality of metrics,” “[(e)] formulating, by the digital assistant builder platform, a weighted loss function using second weights for the plurality of metrics in an asymmetric loss technique based on a difference between the metric score and the target score of each metric, and the penalty factor or the bonus factor,” “[(f)] establishing, by the digital assistant builder platform, one or more constraints for the machine-learning model based on one or more hyperparameters,” and “[(g)] iteratively tuning, by the digital assistant builder platform, the hyperparameters associated with the machine-learning model to obtain a minimum value of the weighted loss function across the plurality of metrics.” These activities of “[(a)] selecting,” [(b)] assigning,” “[(c)] configuring,” “[(d)] evaluating” “[(e)] formulating,” “[(f)] establishing . . . constraints,” and “[(g)] tuning” contain limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, recite a mental process, (MPEP § 2106.04(a)(2) sub III), which is one of the groupings of abstract ideas. (MPEP § 2106.04(a)(2)).
The claim provides more details or specifics to the abstract idea of “[(a)] selecting a plurality of metrics,” “[(a.1)] wherein the plurality of metrics is selected to include a regression error of the machine-learning model and at least one from among a size of the machine-learning model, a training time of the machine-learning model, an accuracy of the machine-learning model, a stability of the machine-learning model, and a confidence score of the machine-learning model,” and accordingly, is merely more specific to the abstract idea.
The claim recites more details or specifics to the abstract idea of “[(b)] assigning,” “[(b.1)] wherein the first weight specifies an importance of the metric in evaluating a performance of the machine-learning model,” and “[(c)] configuring,” in that “[(c.1)] the configuring further comprises: assigning , for a corresponding set of the plurality of specification parameters associated with the metric for the regression error, a first target score, a first penalty factor value that is a relatively high value, and a first bonus factor value that is a relatively low value lower than the first penalty factor value, for minimizing the regression error of the machine-learning model,” and “[(e)] formulating” “[(e.1)] wherein the penalty factor and the bonus factor are related to a second weight associated with each metric, and, in the asymmetric loss technique, the penalty factor is used as a higher value for the second weight associated with the metric failing the target value and the bonus factor is used as a lower value for the second weight associated with the metric exceeding the target value,” and accordingly, are merely more specific to the abstract idea.
The claim also recites more details or specifics to the abstract idea of “[(f)] establishing . . . constraints,” “[(f.1)] wherein the one or more constraints include at least one from among a maximum model size, a maximum inference latency, and a maximum training time,” and accordingly, is merely more specific to the abstract idea.
The claim also recites more details or specifics to the abstract idea of “[(g)] iteratively tuning the hyperparameters,” “comprises: [(g)](a) selecting initial values for the hyperparameters, the machine-learning model being configured based on the initial values; [(g)](b) evaluating the weighted loss function using the initial values of the hyperparameters; [(g)](c) determining whether the weighted loss function is optimized based on the evaluating; and [(g)](d) in response to the weighted loss function not being optimized, searching for new values for the hyperparameters, reconfiguring the machine-learning model with the new values, and repeating steps (b) and (c) using the reconfigured machine-learning model), and accordingly, are merely more specific to the abstract idea.
The claim also recites the limitations of “[(k.1)] in response to determining, for the regression error, that the metric score is greater than the first target score, assigning to the difference between the metric score and the first target score, the first bonus factor value as a weight associated with the metric, and [(k.2)] in response to determining that the metric score is lower than the first target score, assigning to the difference between the first target score and the metric score, the first penalty factor value as the weight associated with the metric,” which contain limitations that can practically be performed in the human mind, including, for example, observations, evaluations, judgments, and opinions, and accordingly, are a mental process, (MPEP § 2106.04(a)(2) sub III), being one of the groupings of abstract ideas. (MPEP § 2106.04(a)(2)). Thus, claim 1 recites an abstract idea.
Under Step 2A Prong Two, the claim as a whole is not integrated into a practical application, because the additional elements recited in the claim beyond the identified abstract idea include “a processor,” “[a digital assistant builder platform] configured to communicate with the DACS via network and comprising a processor and a memory including instructions that, when executed with the processor, cause the computing device to,” which are generic computer components used to execute computer instructions to apply the abstract idea, (MPEP § 2106.05(f)), and do not integrate the abstract idea into a practical application. The claim also recites the additional element of “a machine learning model,” a “digital assistant builder platform,” a “processor,” and “a digital assistant chatbot system (DACS),” which are recited at a high-level of generality, and accordingly, are generic computer components in which the abstract ideas are applied, (MPEP § 2106.05(f)), and does not integrate the abstract idea into a practical application.
The claim also recites the limitation of “[(h)] in response to the weighted loss function being optimized and each constraint being satisfied, generating, by the digital assistant builder platform, a validated machine-learning model comprising the tuned hyperparameters, with values for the hyperparameters obtained in a last iteration” which is the use of the generic computer component (by the digital assistant builder platform) to implement the abstract idea, (MPEP § 2106.05(f)), that does not serve to integrate the abstract idea into a practical application.
The claim further recites the limitations of “[(i)] providing via the network to the DACS the validated machine-learning model, thereby causing the DACS to incorporate the validated machine-learning model,,” and “[(j) wherein the DACS is configured to perform operations including:] providing, as an input to the validated machine-learning model an input utterance received from a user, ” where the activities of “[(i), (j)] providing” are insignificant extra-solution activities of mere transmitting and receiving of data, (MPEP § 2106.05(g)), that does not serve to integrate the abstract idea into a practical application.
The claim recites more details or specifics of the additional element of “[(i), (j)] providing,” “[(i.1)] thereby causing the DACS to incorporate the validated machine-learning model,” “[(j.1) wherein, based on the input utterance, the validated machine-learning model is configured to output a prediction of an intent for the input utterance from the set of predetermined intents,” and “[(j.2)] wherein the at least one criterion includes a criterion for minimizing the regression error,” and accordingly, are merely more specific to the additional elements. Therefore, claim 8 is directed to the abstract idea.
Finally, under Step 2B, the additional elements, taken alone or in combination, do not represent significantly more than the abstract idea itself. The claim recites additional elements beyond the identified abstract idea include “a processor,” “[a digital assistant builder platform] configured to communicate with the DACS via network and comprising a processor and a memory including instructions that, when executed with the processor, cause the computing device to,” which are generic computer components used to execute instructions to apply the abstract ideas, (MPEP § 2106.05(f)), and does not amount to significantly more than the abstract idea. The claim also recites the additional element of “a machine learning model,” and “a digital assistant chatbot system (DACS),” which are recited at a high-level of generality, and accordingly, are generic computer components in which the abstract ideas are applied, (MPEP § 2106.05(f)), and does not amount to significantly more than the abstract idea.
The claim also recites the limitation of “[(h)] in response to the weighted loss function being optimized and each constraint being satisfied, generating, by the digital assistant builder platform, a validated machine-learning model comprising the tuned hyperparameters, with values for the hyperparameters obtained in a last iteration,” which is the use of the generic computer component (by the digital assistant builder platform) to implement the abstract idea, (MPEP § 2106.05(f)), that does not amount to significantly more than the abstract idea.
The claim further recites the limitations of ““[(i)] providing via the network to the DACS the validated machine-learning model, thereby causing the DACS to incorporate the validated machine-learning model,,” and “[(j) wherein the DACS is configured to perform operations including:] providing, as an input to the validated machine-learning model an input utterance received from a user, ”where the activities of “[(h), (i)] providing” are well-understood, routine, and conventional activities of receiving or transmitting data over a network, (MPEP § 2106.05(d) sub II.i), that does not amount to significantly more than the abstract idea.
The claim recites more details or specifics of the additional element of “[(i), (j)] providing,” “[(i.1)] thereby causing the DACS to incorporate the validated machine-learning model,”
“[(j.1) wherein, based on the input utterance, the validated machine-learning model is configured to output a prediction of an intent for the input utterance from the set of predetermined intents,” and “[(j.2)] wherein the at least one criterion includes a criterion for minimizing the regression error,” and accordingly, are merely more specific to the additional elements. Thus, claim 8 is subject-matter ineligible.
Claim 7 depends from claim 1. Claim 14 depends from claim 8. The claims recite more details or specifics of the additional element of the “machine-learning model,” (claims 7 and 14: wherein the machine-learning model is a neural network model, and the hyperparameters associated with the machine-learning model include at least two from among a number of layers of the machine-learning model, a learning rate of the machine-learning model, a number of hidden units in each layer of the machine-learning model, a learning algorithm utilized to train the machine-learning model), and accordingly, are merely more specific to the additional element. The additional elements of the claim does not serve to integrate the abstract idea into integrated into a practical application, (see MPEP § 2106.04(d)), nor do the additional elements amount to significantly more than the abstract idea, (MPEP § 2106.05 sub I; see also MPEP § 2106.05(a) – (h)), and thus, the claim recites no more than the abstract idea. Accordingly, claims 7 and 14 are subject-matter ineligible.
Claim 24 depends directly or indirectly from claim 1. Claim 26 depends directly or indirectly from claim 8. The claims provide more details or specifics to the abstract idea of “[(f)] establishing,” (claims 24 and 26: “wherein: the establishing includes [(f.2)] establishing a plurality of constraints that include the one or more constraints, [(f.3)] each constraint of the plurality of constraints is assigned a priority value indicating a level of importance,” and accordingly, is merely more specific to the abstract idea.
The claims also provide more details or specifics to the abstract idea of “[(g)] iteratively tuning,” (claims 24 and 26: “wherein . . . [(g)] the iterative tuning further comprises [(g.1)] optimizing the machine-learning model to satisfy the plurality of constraints in an order of respective priority values, such that a constraint assigned a higher priority value is required to be satisfied before a constraint assigned a lower priority value is considered,” and accordingly, is merely more specific to the abstract idea. Therefore, claims 24 and 26 are subject-matter ineligible.
Claim 25 depends directly or indirectly from claim 1. Claim 27 depends directly or indirectly from claim 8. The claims provide more details or specifics to the additional element of “[(h)] generating a validated model learning model,” (claims 25 and 27: “wherein [(h)] the machine-learning model is validated across a plurality of validation datasets, [(h.1)] each of the plurality of validation datasets corresponding to a different user population or domain), and accordingly, is merely more specific to the additional element.
Also, the claims provide more details or specifics to the additional element of [(g)] iteratively tuning,” (claims 25 and 27: “wherein . . . [(g)] the iterative tuning is configured [(g)(e)] to adjust the hyperparameters until the machine-learning model satisfies dataset-specific constraints or target values for each of the plurality of validation datasets”), and accordingly, is merely more specific to the additional element. Therefore, claims 25 and 27 are subject-matter ineligible.
Response to Arguments
5. Examiner has fully considered Applicant’s arguments, and responds below accordingly.
Claim Rejections - 35 U.S.C. § 101
6. Applicant submits that “typical machine-learning training methods optimize only a single metric (such as accuracy), failing to account for other factors, e.g., regression error, model size, stability, or latency. This often results in models that do not meet operational requirements. Claimed solution improves upon conventional techniques by
(1) selecting and weighting multiple metrics - including regression error and at least one of accuracy, size, training time, stability, or confidence score.
(2) formulating a weighted, asymmetric loss function using penalty and bonus factors. This ensures that regression errors (i.e., cases where the new model performs worse than previous versions) are heavily penalized, while improvements are rewarded
(3) utilizing iterative tuning process that dynamically searches for parameter values that optimize all relevant metrics simultaneously.
(4) The tuning also uses explicit constraints (e.g., one of maximum training time, maximum model size, or maximum latency) that are enforced throughout the process.
(5) tuning continues until both the loss function is minimized and constraints are satisfied.
Tuning produces a model that is not just accurate, but also robust, stable, and optimized across multiple objectives. It avoids regressions and ensures the model is ready for practical use.” (Response at pp. 12-13).
Applicant also submits that “Claimed solution improves upon conventional techniques by (1) establishing explicit deployment constraints (maximum model size, maximum latency, maximum training time) during training - not after. (2) Only models that satisfy the constraints are validated and deployed. (3) The process automates deployment readiness, e.g., the resulting model is guaranteed to fit within memory, run fast enough, and be accurate enough for its intended environment (See Specification, "Constraint and Target Based Hyperparameter Tuning," "Bot System").” (Response at p. 13).
Also, Applicant submits that “[t]he claimed process integrates abstract mathematical optimization into a practical workflow that produces an AI model fit for a real-world use - not just a theoretical or mathematical outcome.
- The explicit steps and constraints ensure meaningful limits: the invention does not preempt all hyperparameter tuning, but solves the concrete technical problem of producing deployable AI models for chatbot systems.
- These improvements are detailed and supported throughout the specification (see
"Constraint and Target Based Hyperparameter Tuning," FIG. 6, FIG. 7).” (Response at pp. 13-14).
Examiner’s Response:
Examiner considers that, under leg one of the “improvements” per MPEP § 2106.04(d)(1), the Applicant’s disclosure provides “sufficient details such that one of ordinary skill in the art would recognize the claimed invention as providing an improvement.”
Under Step 2A Prong Two, “integration” may be based on the improvements in the functioning of a computer or an improvement to any other technology or technical field. (MPEP § 2106.04(d)(1)). The evaluation requires, [i]n sum, that (1) the specification should be evaluated to determine if the disclosure provides sufficient details such that one of ordinary skill in the art would recognize the claimed invention as providing an improvement. Next, (2) if the specification sets forth such an improvement, the claim must be evaluated to ensure that the claim itself reflects the disclosed improvement.
Moreover, as explained in MPEP § 2106.04(d), subsection III, the Step 2A, “Prong Two analysis considers the claim as a whole. That is, the limitations containing the [abstract idea] as well as the additional elements in the claim besides the [abstract idea] need to be evaluated together to determine whether the claim integrates the [abstract idea] into a practical application.” (2024 SME Guidance, 89 Fed. Reg. 137 at p. 58136 (17 July 2024)).
Also, the revised MPEP provides examples that may indicate an improvement in computer functionality, such as:
xiii. An improved way of training a machine learning model that protected the model’s knowledge about previous tasks while allowing it to effectively learn new tasks; Ex Parte Desjardins, Appeal No. 2024-000567 (PTAB September 26, 2025, Appeals Review Panel Decision) (precedential); and
xiv. Improvements to computer component or system performance based upon adjustments to parameters of a machine learning model associated with tasks or workstreams; Ex Parte Desjardins, Appeal No. 2024-000567 (PTAB September 26, 2025, Appeals Review Panel Decision) (precedential).
(Advance notice of change to the MPEP in light of Ex Parte Desjardins at p. 4 (05 December 2025)).
“[a] claim that integrates [an abstract idea] into a practical application of the exception will apply, rely on, or use the [abstract idea] in a manner that imposes a meaningful limit on the [abstract idea], such that the claim is more than a drafting effort designed to monopolize or preempt the [abstract idea].” (2024 SME Guidance, 89 Fed. Reg. 137 at p. 58136 (17 July 2024)).
The Applicant’s disclosure provides a digital assistant builder platform 102 that provides a chatbot model to a digital assistant / chatbot system 106:
PNG
media_image1.png
651
737
media_image1.png
Greyscale
(Applicant’s Figure 1). The accompanying disclosure recites that the DABP 102 creates digital assistants for different enterprises:
For example, DABP 102 can be used by a bank to create one or more digital assistants for use by the bank's customers. The same DABP 102 platform can be used by multiple enterprises to create digital assistants. As another example, an owner of a restaurant (e.g., a pizza shop) may use DABP 102 to create and deploy a digital assistant that enables customers of the restaurant to order food (e.g., order pizza).
(Specification ¶ 0031). In relation to skills sets, an enterprise may add these to a chatbot via the DABP 102:
There are various ways in which a skill or skill bot can be associated or added to a digital assistant. In some instances, a skill bot can be developed by an enterprise and then added to a digital assistant using DABP 102, e.g., through a user interface provided by DABP 102 for registering the skill bot with the digital assistant. In other instances, a skill bot can be developed and created using DABP 102 and then added to a digital assistant created using DABP 102.
(Specification ¶ 0042). That is, Applicant’s invention is directed to the provisioning of a chatbot based on an a set of enterprise criteria.
Considering the claims as a whole, the claims recite additional elements that are recited at a high-level of generality (e.g., of “a machine learning model,” a “digital assistant builder platform,” a “processor,” and “a digital assistant chatbot system (DACS)) that are used to implement the abstract ideas as identified at Step 2A Prong One as set out above in detail. In general, under Step 2A Prong Two, a claim that integrates an abstract idea into a practical application of the exception will apply, rely on, or use the abstract idea in a manner that imposes a meaningful limit on the abstract idea, such that the claim is more than a drafting effort designed to monopolize or preempt the abstract idea. (2024 SME Guidance, 89 Fed. Reg. 137 at p. 58136 (17 July 2024)). The generic computer components, for example, do not serve to impose a meaningful limit on the abstract idea.
Applicant submits that the specification provides an improved way of training a hyperparameter for a machine learning model (multi-objective optimization, (see, claim 1, line 2), and for a DACS configured to provide an intent prediction. (see claim 1, lines 51-58):
Digital assistant 106 is configured to apply natural language understanding (NLU) techniques to the utterance to understand the meaning of the user input. As part of the NLU processing for an utterance, digital assistant 106 is configured to perform processing to understand the meaning of the utterance, which involves identifying one or more intents and one or more entities corresponding to the utterance.
(Specification ¶ 0035). The improved accuracy, however, is directed to the improvement of the abstract idea of tuning a hyperparameter, (see claim 1, lines 31-.50), an improved abstract idea (even if novel and nonobvious) is still an abstract idea.
Also, under Step 2A Prong Two and Step 2B, the additional elements include generic computer components that are used to implement the abstract idea. The disclosure confirms that these components are not specialized for the purposes set out in the claims:
Processing subsystem 1004 controls the operation of computer system 1000 and may comprise one or more processors, application specific integrated circuits (ASICs), or field programmable gate arrays (FPGAs). The processors may include be single core or multicore processors. The processing resources of computer system 1000 may be organized into one or more processing units 1032, 1034, etc. A processing unit may include one or more processors, one or more cores from the same or different processors, a combination of cores and processors, or other combinations of cores and processors. In some examples, processing subsystem 1004 may include one or more special purpose co-processors such as graphics processors, digital signal processors (DSPs), or the like. In some examples, some or all of the processing units of processing subsystem 804 may be implemented using customized circuits, such as application specific integrated circuits (ASICs), or field programmable gate arrays (FPGAs).
(Specification ¶ 0164). Moreover, the use of Natural Language Understanding in chatbots is well-known, understood, and conventional activities. For example,
Natural language understanding (NLU) is a central technique to implement natural user interfaces such as chatbot, mobile secretary, and smart speakers. The goal of NLU is to extract meanings from natural language and infer user intention.
NLU typically involves two tasks: identifying user intent and extracting domain specific entities, which is often referred to as slot-filling. Intent detection is typically formulated as sentence classification in which single or multiple intent labels are predicted for each sentence. Domain entity extraction, usually referred to as slot-filling problem, is formulated as sequential tagging problem where parts of sentence are extracted and tagged with domain entities.
(Sangkeun Jung, “Semantic Vector Learning for Natural Language Understanding,” Science Direct (2019), at p. 1).
In view of the rejection set out in detail hereinabove, pending claims 1,7-8,14 and 24-27 are subject-matter ineligible.
Conclusion
7. The prior art made of record and not relied upon is considered pertinent to Applicant's disclosure:
(Bortchkarev, "Performance Metrics (error measures) in Machine Learning Regression, Forecasting, and Prognostics: Properties and Typology," arXiv (2019)) teaches a developed typology that can inform metric selection process decision making by structuring performance metrics considerations (point distance, normalization and aggregation phases) and focusing on the key properties of the components chosen. For example, if the business or research need is to emphasize outliers, squared error and arithmetic mean should be used. However, if the business requirement is to isolate outliers, then selection of absolute error and geometric mean is desirable. In other words, the use of this typology turns selection of a metric from a browsing exercise over dozens of metrics into a straightforward process of identifying point distance, normalization and aggregation methods that fit the purpose of the task.
(Adamopoulou et al., "An Overview of Chatbot Technology," IFIP (May 2020)) teaches that Natural Language Understanding (NLU) is at the core of any NLP task. It is a technique to implement natural user interfaces such as a chatbot. NLU aims to extract context and meanings from natural language user inputs, which may be unstructured and respond appropriately according to user intention [32]. It identifies user intent and extracts domain-specific entities. More specifically, an intent represents a mapping between what a user says and what action should be taken by the chatbot. Intent detection is typically formulated as sentence classification in which single or multiple intent labels are predicted for each sentence [32].
8. Any inquiry concerning this communication or earlier communications from the Examiner should be directed to KEVIN L. SMITH whose telephone number is (571) 272-5964. Normally, the Examiner is available on Monday-Thursday 0730-1730.
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, KAKALI CHAKI can be reached on 571-272-3719. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/K.L.S./
Examiner, Art Unit 2122
/KAKALI CHAKI/Supervisory Patent Examiner, Art Unit 2122
Prosecution Insights