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 .
DETAILED ACTION
The Action is responsive to the preliminary amendments filed on 12/10/2024 and the Application filed on 11/26/2024. Claims 1-16 and 20-23 are pending claims. Claims 1, 12, and 16 are written in independent form. Claims 17-19 were cancelled in the preliminary amendments.
Priority
Applicant's claim for benefit of prior-filed provisional application 63/602,637, filed on 11/26/2023, under 35. U.S.C. 119(e) or under 35 U.S.C. 120, 121, or 365(c) is acknowledged.
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 1-16 and 20-23 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-patentable subject matter. The claimed invention is directed to one or more abstract ideas without significantly more. The judicial exception is not integrated into a practical application. The claims do not include additional elements that are sufficient to amount to significantly more than judicial exception. The eligibility analysis in support of these findings is provided below.
As per Claims 1, 12, and 16,
STEP 1:In accordance with Step 1 of the eligibility inquiry (as explained in MPEP 2106), the claimed method (claims 1-11), method, (claims 12-15), and system (claims 16 and 20-23) are directed to one of the eligible categories of subject matter and therefore satisfies Step 1.
STEP 2A Prong One:The independent claims 1, 12, and 16 recite the following limitations directed to an abstract idea:
Applying the risk-aware model to the user input;
The limitation recites a mathematical concept of executing a mathematical formula in the form of a risk-aware model that takes as input “the user input”.
Comparing the corresponding risk values to a threshold risk value; and
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the corresponding risk values and a threshold risk value, and based on the observation and evaluation, making a judgement and/or opinion of a comparison.
Regenerating the user input based on the comparing.
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the comparing and the user input, and based on the observation and evaluation, making a judgement and/or opinion of a regenerated user input.
Independent Claim 12 further recites the following limitation(s) directed to an abstract idea:
Prepending logical instructions to the user input;
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the user input and logical instructions, and based on the observation and evaluation, making a judgement and/or opinion to prepend, or add/associate, the logical instructions to the user input.
STEP 2A Prong Two:Claim 16 recites that the steps are performed using “a computer system”, “one or more processors”, and “memory”, which is a high-level recitation of generic computer components and represents mere instructions to apply on a computer as in MPEP 2106.05(f), which does not provide integration into a practical application.
The claims recite the following additional elements:
Obtaining a risk-aware model and a user input;
The limitation recites an insignificant extra solution activity as sending/receiving data (ie. Mere data gathering) such as ‘obtaining information’ as identified in MPEP 2106.05(g) and does not provide integration into a practical application.
Receiving, based on the applying, model output and corresponding risk values;
The limitation recites an insignificant extra solution activity as sending/receiving data (ie. Mere data gathering) such as ‘obtaining information’ as identified in MPEP 2106.05(g) and does not provide integration into a practical application.
Viewing the additional limitations together and the claim as a whole, nothing provides integration into a practical application.
STEP 2B:
The conclusions for the mere implementation using a computer are carried over and does not provide significantly more.
With respect to “Obtaining a risk-aware model and a user input;” identified as insignificant extra-solution activity above this is also WURC as court-identified see MPEP 2106.05(d)(II)(i).
With respect to “Receiving, based on the applying, model output and corresponding risk values;” identified as insignificant extra-solution activity above this is also WURC as court-identified see MPEP 2106.05(d)(II)(i).
Looking at the claim as a whole does not change this conclusion and the claim is ineligible.
As per Dependent Claims 2-11, 13-15, and 20-23,
STEP 1:In accordance with Step 1 of the eligibility inquiry (as explained in MPEP 2106), the claimed method (claims 1-11), method, (claims 12-15), and system (claims 16 and 20-23) are directed to one of the eligible categories of subject matter and therefore satisfies Step 1.
STEP 2A Prong One:The dependent claims 2-11, 13-15, and 20-23 recite the following limitations directed to an abstract idea:
The limitation(s) of Dependent Claims 2, 13, and 20 include the step(s) of:
Iteratively performing the applying, the receiving, the comparing, and the regenerating using the regenerated user input until one or more processing conditions are met.
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating one or more processing conditions, and based on the observation and evaluation, making a judgement and/or opinion to iteratively repeat the steps using the regenerated user input until a judgment and/or opinion is made that the conditions are met.
The limitation(s) of Dependent Claims 3, 13, and 21 include the step(s) of:
In response to the one or more processing conditions being met, analyzing the outputted corresponding risk values against one or more risk criteria;
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the processing conditions being met, the outputted corresponding risk values, and one or more risk criteria.
Selecting, based on the analyzing, a user input having an outputted corresponding risk value that satisfies the one or more risk criteria; and
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the processing conditions being met, the outputted corresponding risk values, and one or more risk criteria, and based on the observation and evaluation, making a judgement and/or opinion of a selected user input having an outputted corresponding risk value that satisfies the one or more risk criteria.
The limitation(s) of Dependent Claim 4 includes the step(s) of:
Determining that the one or more processing conditions are met based on the corresponding risk values being less than the threshold risk value.
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the corresponding risk values and the threshold risk value, and based on the observation and evaluation, making a judgement and/or opinion that the corresponding risk values are less than the threshold risk value resulting in the processing conditions being met.
The limitation(s) of Dependent Claims 5, 14, and 22 include the step(s) of:
Wherein the user input is regenerated in response to determining that the corresponding risk values are greater than the threshold risk value.
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the corresponding risk values and the threshold risk value, and based on the observation and evaluation, making a judgement and/or opinion that the corresponding risk values are greater than the threshold risk value resulting in the user input being regenerated.
The limitation(s) of Dependent Claim 10 includes the step(s) of:
Determining that the one or more processing conditions are met based on the model output being a final output of the risk-aware model.
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the model output of the risk-aware model, and based on the observation and evaluation, making a judgement and/or opinion that it is “a final output” and based on the judgement and/or opinion that it is a “final output, making another judgement and/or opinion that the one or more processing conditions are met.
The limitation(s) of Dependent Claims 11 and 15 include the step(s) of:
Aggregating the corresponding risk values to generate an aggregated risk value;
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the corresponding risk values in aggregate, and based on the observation and evaluation, making a judgement and or opinion of “an aggregated risk value”.
Determining whether the aggregated risk value is less than the threshold risk value; and
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the aggregated risk value and the threshold risk value, and based on the observation and evaluation, making a judgement and/or opinion of whether the aggregated risk value is less than the threshold risk value.
In response to determining that the aggregated risk value is greater than the threshold risk value, performing the regenerating.
The limitation recites a mental process of observation, evaluation, judgement, and/or opinion capable of being performed by the human mind, or by a human using a pen and paper, by observing and evaluating the determination that the aggregated risk value is greater than the threshold risk value, and based on the observation and evaluation, making a judgement and/or opinion to perform the regenerating.
STEP 2A Prong Two:The claim(s) recite the following additional elements:
The limitation(s) of Dependent Claims 3 and 21 include the step(s) of:
Returning the selected user input to a user device.
The limitation recites an insignificant extra solution activity as sending/receiving data (ie. Mere data gathering) such as ‘obtaining information’ as identified in MPEP 2106.05(g) and does not provide integration into a practical application.
The limitation(s) of Dependent Claims 6 and 23 include the step(s) of:
Wherein the model output comprises one or more sequences in a train-of-thought (TOT) of the risk-aware model.
The limitation recites an insignificant extra-solution activity as selecting a particular type of data being used to represent the model output as identified in MPEP 2106.05(g) and does not provide integration into a practical application.
The limitation(s) of Dependent Claim 7 includes the step(s) of:
Wherein the corresponding risk values comprise an aleatoric uncertainty value corresponding to each sequence.
The limitation recites an insignificant extra-solution activity as selecting a particular type of data being used to represent the corresponding risk values as identified in MPEP 2106.05(g) and does not provide integration into a practical application.
The limitation(s) of Dependent Claim 8 includes the step(s) of:
Wherein the corresponding risk values comprise an epistemic uncertainty value corresponding to each sequence.
The limitation recites an insignificant extra-solution activity as selecting a particular type of data being used to represent the corresponding risk values as identified in MPEP 2106.05(g) and does not provide integration into a practical application.
The limitation(s) of Dependent Claim 9 includes the step(s) of:
Wherein the corresponding risk values comprise a bias value corresponding to each sequence.
The limitation recites an insignificant extra-solution activity as selecting a particular type of data being used to represent the corresponding risk values as identified in MPEP 2106.05(g) and does not provide integration into a practical application.
Viewing the additional limitations together and the claim as a whole, nothing provides integration into a practical application.
STEP 2B:
The conclusions for the mere implementation using a computer are carried over and does not provide significantly more.
With respect to Claims 3 and 21 reciting “Returning the selected user input to a user device.” identified as insignificant extra-solution activity above this is also WURC as court-identified see MPEP 2106.05(d)(II)(i).
With respect to Claims 6 and 23 reciting “Wherein the model output comprises one or more sequences in a train-of-thought (TOT) of the risk-aware model.” identified as insignificant extra-solution activity above this is also WURC as court-identified see MPEP 2106.05(d)(II)(iv).
With respect to Claim 7 reciting “Wherein the corresponding risk values comprise an aleatoric uncertainty value corresponding to each sequence.” identified as insignificant extra-solution activity above this is also WURC as court-identified see MPEP 2106.05(d)(II)(iv).
With respect to Claim 8 reciting “Wherein the corresponding risk values comprise an epistemic uncertainty value corresponding to each sequence.” identified as insignificant extra-solution activity above this is also WURC as court-identified see MPEP 2106.05(d)(II)(iv).
With respect to Claim 9 reciting “Wherein the corresponding risk values comprise a bias value corresponding to each sequence.” identified as insignificant extra-solution activity above this is also WURC as court-identified see MPEP 2106.05(d)(II)(iv).
Looking at the claim as a whole does not change this conclusion and the claim is ineligible.
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.
Claim(s) 1-6, 10, 12-14, 16, and 20-23 are rejected under 35 U.S.C. 103 as being unpatentable over Lafon et al. (U.S. Pre-Grant Publication No. 2025/0068741, hereinafter referred to as Lafon) and further in view of Clement et al. (U.S. Pre-Grant Publication No. 2024/0419917, hereinafter referred to as Clement).
Regarding Claim 1:
Suzuki teaches a method for improving accuracy of model output generation, the method comprising:
Obtaining a risk-aware model and a user input;
Lafon teaches obtaining a trained input filter 502 and receiving a user query 506 (Fig. 5)
Applying the risk-aware model to the user input;
Lafon teaches “At 506 the user query 122 is received from the user device 120. it is passed to the input filter 102 at 508 wherein an input risk score 132 is obtained for the user query 122.” (Para. [0036]).
Receiving, based on the applying, model output and corresponding risk values;
Lafon teaches “At 506 the user query 122 is received from the user device 120. it is passed to the input filter 102 at 508 wherein an input risk score 132 is obtained for the user query 122.” (Para. [0036]).
Comparing the corresponding risk values to a threshold risk value; and
Lafon teaches “The input risk score 132 is compared at 608 to an input risk threshold 232 by the threat indicator 244. At 610, it is determined if the input risk score 132 is greater than the input risk threshold 232. If yes, the method proceeds to 612 wherein it is determined that the user query 122 includes harmful content and is to be blocked.” (Para. [0038]).
Lafon explicitly teaches all of the elements of the claimed invention as recited above except:
Regenerating the user input based on the comparing.
However, in the related field of endeavor of analyzing queries/requests, Clement teaches:
Regenerating the user input based on the comparing.
Clement teaches “The service creates an additional prompt to alleviate any issues detected…the additional prompt includes the previously-transmitted prompts.” (Para. [0044])
Thus, it would have been obvious to one of ordinary skill in the art, having the teachings of Clement and Lafon at the time that the claimed invention was effectively filed, to have modified the systems and methods for detecting the risk of a user query, as taught by Lafon, with the repeating nature to review and revise input request/prompts, as taught by Clement.
One would have been motivated to make such combination because while Lafon teaches detecting risks with the user query (Fig. 5 Elements 508-510), Clement teaches “the service creates an additional prompt to alleviate any issues detected” (Para. [0044]) and “Upon detection of a software vulnerability by the static analyzer 132, the large language model 104 is used to generate source code to repair the vulnerability. The large language model 104 is given examples of previous fixes made to fix the same type of software vulnerability in the prompt” (Para. [0034]) and it would have been obvious to a person having ordinary skill in the art that fixing the detected issues/risks would improve the input and thus improve the user experience by being able to process the input once it is at an acceptable risk/vulnerability level.
Regarding Claim 2:
Clement and Lafon further teach:
Iteratively performing the applying, the receiving, the comparing, and the regenerating using the regenerated user input until one or more processing conditions are met.
Clement teaches repeating the same steps “until the vulnerability fix compiles successfully and passes the associated unit tests” (Para. [0065]) thereby teaching repeating the steps iteratively until the conditions are met.
Regarding Claim 3:
Clement and Lafon further teach:
In response to the one or more processing conditions being met, analyzing the outputted corresponding risk values against one or more risk criteria;
Clement teaches repeating the same steps “until the vulnerability fix compiles successfully and passes the associated unit tests” (Para. [0065]) thereby teaching repeating the steps iteratively until the conditions are met.
Lafon further teaches “The input filter and the output filter may each include a pre-trained language model as a base with additional layers trained to estimate corresponding risk scores for textual inputs provided to them. In an example, the user query and the model query response may include textual data. Accordingly, the input filter generates an input risk score for the user query. Based on a comparison of the input risk score to an input risk threshold, the user query may be forwarded to the generative AI model or suppressed from being forwarded to the generative AI model. The input risk score can be a numerical value indicative of harmful content if any included in the user query. The input filter is trained via supervised training to generate the input risk scores for user queries. The labeled training data used to train the input filter may include different user queries, tokens generated from the user queries, and corresponding input risk scores.” (Par. [0014]).
Selecting, based on the analyzing, a user input having an outputted corresponding risk value that satisfies the one or more risk criteria; and
Lafon teaches “Based on the input risk score 132, it is determined at 510 if the user query 122 should be provided to the generative AI model 150… If it is determined at 510 that the user query 122 should be passed to the generative AI model 150, then the user query 122 is provided to the generative AI model 150 at 514. The user query 122 can be passed to the generative AI model 150 if it is identified as benign, i.e., the user query 122 does not include any harmful content.” (Para. [0036]).
Returning the selected user input to a user device.
Lafon teaches “If it is determined at 520 that the model query response 124 does not include any restricted or forbidden content, it is forwarded to the user device 120 at 522.” (Para. [0037]).
Regarding Claim 4:
Clement and Lafon further teach:
Determining that the one or more processing conditions are met based on the corresponding risk values being less than the threshold risk value.
Lafon teaches “If at 610, it is determined that the input risk score 132 is less than the input risk threshold 232 the method proceeds to 614 wherein it is determined that the user query 122 does not include harmful content and hence may be forwarded to the generative AI model 150.” (Para. [0038]).
Regarding Claim 5:
Clement and Lafon further teach:
Wherein the user input is regenerated in response to determining that the corresponding risk values are greater than the threshold risk value.
Lafon in combination with Clement teaches “it is determined if the input risk score 132 is greater than the input risk threshold 232. If yes, the method proceeds to 612 wherein it is determined that the user query 122 includes harmful content and is to be blocked” (Lafon - Para. [0038]) where “The service creates an additional prompt to alleviate any issues detected…the additional prompt includes the previously-transmitted prompts.” (Clement - Para. [0044])
Regarding Claim 6:
Clement and Lafon further teach:
Wherein the model output comprises one or more sequences in a train-of-thought (TOT) of the risk-aware model.
Lafon teaches “The user query 122 can be passed to the generative AI model 150 if it is identified as benign, i.e., the user query 122 does not include any harmful content.” (Para. [0036] & Fig. 6 Element 614) thereby teaching at least one sequence in the train of thought of the risk-aware model being based on whether the query is identified as benign.
Regarding Claim 10:
Clement and Lafon further teach:
Determining that the one or more processing conditions are met based on the model output being a final output of the risk-aware model.
Clement teaches “If the response from the large language model is adequate (block 214—no), the service returns the response to the client (block 216). The client may continue the conversion by issuing further requests (block 218—yes) which are processed until there are no further requests” (Para. [0045]).
Regarding Claim 12:
Some of the limitations herein are similar to some or all of the limitations as recited in Claim 1.
Clement and Lafon further teach:
Prepending logical instructions to the user input;
Clement teaches “The request includes a query, context and intent 112” (Para. [0041]).
Regarding Claim 13:
All of the limitations herein are similar to some or all of the limitations as recited in Claims 2 and 3.
Regarding Claim 14:
All of the limitations herein are similar to some or all of the limitations as recited in Claim 5.
Regarding Claim 15:
All of the limitations herein are similar to some or all of the limitations as recited in Claim 11.
Regarding Claim 16:
Some of the limitations herein are similar to some or all of the limitations as recited in Claim 1.
Clement and Lafon further teach:
A computer system comprising one or more processors and memory storing instructions that, when executed by the one or more processors, cause the computer system to perform a process (Lafon – Paras. [0049]-[0050])
Regarding Claim 20:
All of the limitations herein are similar to some or all of the limitations as recited in Claim 2.
Regarding Claim 21:
All of the limitations herein are similar to some or all of the limitations as recited in Claim 3.
Regarding Claim 22:
All of the limitations herein are similar to some or all of the limitations as recited in Claim 5.
Regarding Claim 23:
All of the limitations herein are similar to some or all of the limitations as recited in Claim 6.
Claim(s) 7-9 are rejected under 35 U.S.C. 103 as being unpatentable over Clement and Lafon, and further in view of Suzuki (U.S. Pre-Grant Publication No. 2023/0115987).
Regarding Claim 7:
Clement and Lafon explicitly teach all of the elements of the claimed invention as recited above except:
Wherein the corresponding risk values comprise an aleatoric uncertainty value corresponding to each sequence.
However, in the related field of endeavor of data adjustment, Suzuki teaches:
Wherein the corresponding risk values comprise an aleatoric uncertainty value corresponding to each sequence.
Suzuki teaches “There are two types of uncertainty in deep learning. Uncertainty in deep learning can be divided into accidental uncertainty (Aleatoric uncertainty) and uncertainty in recognition (Epistemic uncertainty). The former Aleatoric uncertainty is due to observation noise and is not due to lack of data. For example, a case such as a hidden and invisible image (occlusion) corresponds to (matches) this (Aleatoric uncertainty). Since the mouth of the face of the masked person is originally hidden by the mask, it cannot be observed as data. On the other hand, the latter Epistemic uncertainty refers to the uncertainty due to the lack of data. Epistemic uncertainty can be improved if sufficient data is present. However, in general, it has been considered difficult to clarify epistemic uncertainties in the imaging field.” (Para. [0111].
Thus, it would have been obvious to one of ordinary skill in the art, having the teachings of Suzuki, Clement, and Lafon at the time that the claimed invention was effectively filed, to have modified the systems and methods for detecting the risk of a user query, as taught by Lafon, and the repeating nature to review and revise input request/prompts, as taught by Clement, with the assessed types of uncertainty when performing useful/harmful data determination, as taught by Suzuki.
One would have been motivated to make such combination because while Clement teaches fixing regenerating the prompt, Suzuki teaches “The data adjustment system 1 learns again using the adjusted data to generate a network. The data adjustment system 1 performs a test to find out the cause of the remaining erroneous determination. The data adjustment system 1 repeats a loop of automatically adjusting and relearning data so as to improve accuracy.” (Para. [0087]) and it would have been obvious toa person having ordinary skill in the art that identifying the cause of the erroneous determination would improve the ability to fix the erroneous determination.
Regarding Claim 8:
Suzuki, Clement, and Lafon further teach:
Wherein the corresponding risk values comprise an epistemic uncertainty value corresponding to each sequence.
Suzuki teaches “There are two types of uncertainty in deep learning. Uncertainty in deep learning can be divided into accidental uncertainty (Aleatoric uncertainty) and uncertainty in recognition (Epistemic uncertainty). The former Aleatoric uncertainty is due to observation noise and is not due to lack of data. For example, a case such as a hidden and invisible image (occlusion) corresponds to (matches) this (Aleatoric uncertainty). Since the mouth of the face of the masked person is originally hidden by the mask, it cannot be observed as data. On the other hand, the latter Epistemic uncertainty refers to the uncertainty due to the lack of data. Epistemic uncertainty can be improved if sufficient data is present. However, in general, it has been considered difficult to clarify epistemic uncertainties in the imaging field.” (Para. [0111].
Regarding Claim 9:
Bishof, Clement, and Lafon further teach:
Wherein the corresponding risk values comprise a bias value corresponding to each sequence.
Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over Clement and Lafon, and further in view of Iavania et al. (U.S. Pre-Grant Publication No. 2022/0043735, hereinafter referred to as Iavania).
Regarding Claim 11:
Clement and Lafon explicitly teach all of the elements of the claimed invention as recited above except:
Aggregating the corresponding risk values to generate an aggregated risk value;
Determining whether the aggregated risk value is less than the threshold risk value; and
In response to determining that the aggregated risk value is greater than the threshold risk value, performing the regenerating.
However, in the related field of endeavor of analyzing risk for a user’s received input, Iavania teaches:
Aggregating the corresponding risk values to generate an aggregated risk value;
Iavania teaches “In this embodiment, the program 104 calculates a risk score by assigning values for multiple factors for a single user's received input and aggregates the values for the factors that are present in the analysis.” (Para. [0017])
Determining whether the aggregated risk value is less than the threshold risk value; and
Iavania teaches “the program 104 determines the priority order by ranking the calculated risk scores for multiple users from high risk to low risk. In this embodiment, the program 104 defines low risk for the calculated risk score as less than or equal to 3. In this embodiment, the program 104 defines high risk for the calculated risk score as greater than 3” (Para. [0020])
In response to determining that the aggregated risk value is greater than the threshold risk value, performing the regenerating.
Iavania teaches “In step 206, the program 104 dynamically optimizes the risk analysis within a user interface. In this embodiment, the program 104 dynamically optimizes the risk analysis within a user interface by modifying the risk analysis to reflect any changes to received user input” (Para. [0021]).
Thus, it would have been obvious to one of ordinary skill in the art, having the teachings of Iavania, Clement, and Lafon at the time that the claimed invention was effectively filed, to have modified the systems and methods for detecting the risk of a user query, as taught by Lafon, and the repeating nature to review and revise input request/prompts, as taught by Clement, with the , as taught by Suzuki.
One would have been motivated to make such modification because Iavania teaches “the program 104 calculates a risk score by assigning values for multiple factors for a single user's received input and aggregates the values for the factors that are present in the analysis.” (Para. [0017]) and it would have been obvious to a person having ordinary skill in the art that analyzing multiple factors with scores instead of generating a single score would improve the visibility into the spread of the factors and their individual contribution to the overall score.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Russell (U.S. Pre-Grant Publication No. 2025/0094904) teaches a computing system is provided for risk management. The computing system includes processing circuitry configured to receive input of a control opportunity score, a numerical status score, and one or a plurality of risk impact values for a respective plurality of target objectives for a given risk, calculate a residual risk value for the given risk based on the control opportunity score and an inherent risk value, calculate a relative risk value for the given risk based on the residual risk value, the numerical status score, and the one or the plurality of risk impact values, generate a prompt including the relative risk value and a description of the given risk, input the prompt into a generative model to generate a recommendation for mitigating the given risk, and output the recommendation generated by the generative model.
Bishof et al. (U.S. Pre-Grant Publication No. 2025/0036982) teaches a method that includes receiving user input associated with identifying a threshold point associated with a classification task, identifying, a machine learning model, in the first set of visual data, a machine placement candidate point associated with identifying the threshold point, and identifying, based on the machine placement candidate point, a set of baseline confidence values via a baseline uncertainty model. The method includes training the machine learning model based on a determined state space by identifying, subsequent sets of visual data additional threshold points, receiving user feedback indicating an accuracy, comparing the baseline confidence values with locations associated with the additional threshold points, generating reward values based on an identified amount of error, and configuring the machine learning model based on the reward values. The method includes identifying in a second set of visual data, via the trained machine learning model, a visual feature associated with the classification task.The reference further teaches “Defining uncertainty as the probability that a model is incorrect (or confidence as the probability that a model is correct) is useful for evaluating trust in a model or metering the cognitive load and human interaction. However, this definition does have limitations when compared to other discussions in literature. One limitation is that it fails to distinguish between aleatoric uncertainty and epistemic uncertainty (Hullermeier and Waegeman, 2021; Hora, 1996). Aleatoric uncertainty involves the distribution of noise and other randomness within the data, while epistemic uncertainty addresses the lack of knowledge within the ML model. Aleatoric uncertainty is difficult to measure (Wang et al., 2019), especially under concept drift (Lu et al., 2020). Other studies more aligned with our approach have defined uncertainty as a measurement of what is not known at the time of classification (Kläs and Vollmer, 2018). Though this definition allows considerably more leeway, we choose a probabilistic interpretation that allows us to validate models for uncertainty experimentally within an IML paradigm.” (Para. [0047]).
Marlin et al. (U.S. Pre-Grant Publication No. 2021/0158221) teaches facilitating analysis of a model. Accordingly, the method may include receiving, using a communication device, a model data associated with a model from a user device, assessing, using a processing device, the model data, identifying, using the processing device, a field associated with the model based on the assessing, analyzing, using the processing device, the field based on the identifying of the field, identifying, using the processing device, a related field associated with the field based on the analyzing of the field, analyzing, using the processing device, the related field based on the model, generating, using the processing device, a notification based on the analyzing of the related field, transmitting, using the communication device, the notification to the user device, and storing, using a storage device, the model data and the model.
Zhao et al. (U.S. Pre-Grant Publication No. 2023/0207083) teaches a method (100) for generating and presenting a patient risk score, comprising: (i) receiving (104) a plurality of features about the patient comprising at least a plurality of vital signs obtained for the patient; (ii) characterizing (106), using a trained risk model, an importance of each of the received plurality of features to a risk score analysis; (iii) calculating (108) an initial risk score; (iv) identifying (110) one or more missing features; (v) calculating (110) a risk score confidence interval comprising an effect of the identified one or more missing features on a confidence range of the initial risk score; (vi) calculating (112), from the initial risk score and the calculated risk score confidence interval, a risk score range; and (vii) presenting (118) the risk score range comprising the initial the score plus and minus the calculated risk score confidence interval.The reference further teaches “Many ML-based tools comprising risk scores will either output a risk score regardless of certainty or not output score at all. Alternative methods exist to quantify uncertainty within a machine-learning model, but they focus on delineating model (epistemic) uncertainty and data (aleatoric) uncertainty, and do not relate to feature importance interpretation.” (Para. [0004]).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to ROBERT F MAY whose telephone number is (571)272-3195. The examiner can normally be reached Monday-Friday 9:30am to 6pm.
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, Boris Gorney can be reached on 571-270-5626. 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.
/BORIS GORNEY/Supervisory Patent Examiner, Art Unit 2154
/ROBERT F MAY/Examiner, Art Unit 2154 12/13/2025