DETAILED ACTION
This Office Action is in response to Applicant's Communication received on 02/15/2024 for application number 18/442,410.
Claims 1-17 are presented for examination. Claims 1, 6, and 13 are independent claims.
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 .
Claim Objections
Claim 13 is objected to because of the following informalities: Claim 13 recites “the new set of guidelines”, which has no antecedent basis; “the set of changes to made to the one or more features sets” should be “the set of changes to be made to the one or more features sets”. Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-17 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 1 recites “pre-contextually embedding the first plurality of words and the second plurality of words.” ‘Pre-contextual’ embedding is not defined in the specification; specification [0034] and [0052] merely repeat the phrase. It is unclear whether the term ‘pre-contextual embedding’ requires an embedding performed before some later contextual processing step, an embedding of a particular context, or something else, and one of ordinary skill in the art would not be reasonably apprised of the scope of the limitation. For the purposes of examination, the examiner will interpret this limitation as: embedding the first plurality of words and the second plurality of words without regard to the surrounding context in which the word appears.
Claims 1 and 6 recite “visually classifying and categorizing features, within the set of vectorized features, that are closely related to each other.” The term “closely related” is a relative term which renders the claim indefinite. The term “closely related” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention. For the purposes of examination, the examiner will interpret this limitation as: grouping the vectorized features based on any measure of similarity and producing a map or other representation.
Claims 1, 6, and 13 recite “simulating substantially all possible parameter and feature combinations.” The term “substantially all” is a relative term which renders the claim indefinite. The term “substantially all” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention. For the purposes of examination, the examiner will interpret this limitation as: simulating a plurality of combinations of parameters and features.
Claims 2-5, 7-12, and 14-17, which depend from independent claim 1, 6, and 13 respectively, are also rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for depending on an indefinite parent claim.
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 1-17 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Step 1
Claims 1-12 are directed to a method and Claims 13-17 are directed to a system and fall within one of the statutory categories (process, machine) and are eligible under Step 1.
Step 2A Prong 1
Independent Claims
Claim 1 recites:
A method for continuous system adaption, the method comprising: at a data processing stage: tokenizing the structure map and the set of new guidelines into a first plurality of words directed to the structure map and a second plurality of words directed to the set of new guidelines; creating a vocabulary from the tokenized structure map and the tokenized set of new guidelines, said vocabulary comprising the first plurality of words and the second plurality of words; visually classifying and categorizing features, within the set of vectorized features, that are closely related to each other; categorizing objects within the structure map as impacted or unimpacted; based on the simulating, grading a complexity of a set of changes to be made to the one or more sets of features; based on the simulating, identifying a degree of change compelled by the in-use computing environment to comply with the set of changes; splitting the combinations into multiple states of existence using a plurality of rules; processing each of the one or more mutually exclusive states - these limitations recite mental processes, i.e., observation, evaluation, judgement, and opinion that a person can practically perform in mind or with pen and paper: a person can read a description of a computing environment and a new set of guidelines, break each into words and list them, group items that are closely related, judge which features of the environment are and are not affected by the guidelines, grade the complexity of the needed changes (degree of change) and count how many features or parameters must change, sort candidate options into categories by applying rules, and evaluate each category to select the final recommendation.
pre-contextually embedding the first plurality of words and the second plurality of words included in the vocabulary; padding and assigning sequence length to the first plurality of words and the second plurality of words included in the vocabulary; extracting a set of vectorized features; simulating substantially all possible parameter and feature combinations; and based on the simulating, identifying resource consumption of each of the combinations - these limitations recite mathematical concepts, namely mapping words to numerical vectors, normalizing each word sequence to a fixed length, enumerating substantially all possible combinations of values assignable to parameters and features and calculating resource consumption (cost and time) for each combination (i.e., mathematical calculations and relationships).
Claim 6 recites:
A method for continuous system adaption, the method comprising: at a data processing stage: tokenizing the structure map into a first plurality of grams;
tokenizing the set of new guidelines into a second plurality of grams; creating a vocabulary from the first plurality of grams and the second plurality of grams; visually classifying and categorizing features, within the set of vectorized features, that are closely related to each other; categorizing each feature, included in the one or more sets of features, within the structure map as impacted or unimpacted; splitting the combinations of features into multiple states of existence using a plurality of rules; processing each of the one or more mutually exclusive states; said final recommendation comprising an assignable parameter for each feature - these limitations recite mental processes, i.e., observation, evaluation, judgement, and opinion that a person can practically perform in mind or with pen and paper: a person can read a description of a computing environment and a new set of guidelines, break each into words or grams and list them, group items that are closely related, judge which features of the environment are and are not affected by the guidelines, grade the complexity of the needed changes (degree of change) and count how many features or parameters must change, sort candidate options into categories by applying rules, and evaluate each category to select the final recommendation, including a parameter value for each feature as the recommendation.
padding and assigning sequence length to the first plurality of grams within the vocabulary; padding and assigning sequence length to the second plurality of grams within the vocabulary; extracting a set of vectorized features from the vocabulary; simulating substantially all combinations of features included in the one or more sets of features and all combinations of one or more parameters assignable to features included in the one or more sets of features;
identifying resource consumption of each of the combinations of features - these limitations recite mathematical concepts, namely mapping words to numerical vectors, normalizing each sequence of grams to a fixed length, enumerating substantially all possible combinations of values assignable to parameters and features and calculating resource consumption (cost and time) for each combination (i.e., mathematical calculations and relationships).
Claim 13 recites:
tokenize the structure map and the set of new guidelines into a first plurality of words directed to the structure map and a second plurality of words directed to the set of new guidelines; create a vocabulary from the first plurality of words and the second plurality of words; categorize features included in the one or more feature sets as impacted or unimpacted; identify a set of changes to be made to the one or more features sets based on the simulation; grade a complexity of the set of changes to made to the one or more features sets; re-identify a set of changes to be made to the one or more features sets based on the grade; identify a degree of change compelled by the in-use computing environment to comply with the set of changes; re-identify a set of changes to be made to the one or more features sets based on the degree of change - these limitations recite mental processes, i.e., observation, evaluation, judgement, and opinion that a person can practically perform in mind or with pen and paper: a person can read a description of a computing environment and a new set of guidelines, break each into words and list them, group items that are closely related, judge which features of the environment are and are not affected by the guidelines, identify the changes needed, grade the complexity of the needed changes (degree of change) and count how many features or parameters must change, and revise the changes again.
pad and assign sequence length to the first plurality of words and second plurality of words included in the vocabulary; extract a set of vectorized features from the vocabulary; simulate substantially all possible parameter and feature combinations to stabilize the impacted features when evaluated alongside the new set of guidelines - these limitations recite mathematical concepts, namely mapping words to numerical vectors, normalizing vector length, enumerating substantially all possible combinations of values assignable to parameters and features and calculating resource consumption (cost and time) for each combination (i.e., mathematical calculations and relationships).
Accordingly, the claims recites an abstract idea that falls under “mental process” and “mathematical concepts” grouping.
Step 2A Prong 2
Independent Claims
Additional elements
Claim 1:
at a data gathering stage: receiving a structure map of an in-use computing environment, the structure map of the in-use computing environment involving:
one or more sets of features; and one or more dependencies between features included in the one or more sets of features; receiving a set of new guidelines, the set of new guidelines for implementation in the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering (see MPEP § 2106.05(g)).
using a large language model; inputting the vocabulary into an artificially-intelligent engine; at a machine learning stage, at the artificially-intelligent engine: at an autoencoder located within the artificially-intelligent engine; at a two-class classifier located within the artificially-intelligent engine; at a quantum engine; at a quantum simulator within the quantum engine; a complexity gradient simulator within the quantum engine; at a diversity analyzer simulator within the quantum engine; at a controller within the quantum engine; at a state processing engine within the quantum engine - these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
outputting one or more mutually exclusive states from the controller; outputting a final recommendation from the state processing engine within the quantum engine; inputting the final recommendation to a transformer neural network; processing the final recommendation at the transformer neural network; outputting, from the transformer neural network , a final set of changes to the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering and outputting (see MPEP § 2106.05(g)), using generic computer components or generic class of machine learning models (see MPEP § 2106.05(f)). These limitations can also be viewed as generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
Claim 6:
at a data gathering stage: receiving a structure map of an in-use computing environment, the structure map of the in-use computing environment involving:
one or more sets of features; and one or more dependencies between features included in the one or more sets of features; receiving a set of new guidelines, the set of new guidelines for implementation in the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering (see MPEP § 2106.05(g)).
using a large language model; inputting the vocabulary into an artificially-intelligent engine; at a machine learning stage, at the artificially-intelligent engine: at an autoencoder located within the artificially-intelligent engine; at a two-class classifier located within the artificially-intelligent engine; at a quantum engine; at a controller within the quantum engine; at a state processing engine within the quantum engine - these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
outputting one or more mutually exclusive states from the controller; outputting a final recommendation from the state processing engine within the quantum engine; inputting the final recommendation to a transformer neural network;
processing the final recommendation at the transformer neural network; and
outputting, from the transformer neural network, a final set of changes to the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering and outputting (see MPEP § 2106.05(g)), using generic computer components or generic class of machine learning models (see MPEP § 2106.05(f)). These limitations can also be viewed as generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
Claim 13:
A system for continuous system adaption, the system comprising: a receiver; a data processor; using a large language model; an artificially-intelligent engine; using an autoencoder; using a quantum simulator; at a second quantum simulator; at a third quantum simulator - these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
receive: a structure map of an in-use computing environment, the structure map comprising: details relating to one or more feature sets; details relating to one or more dependencies between features included in the one or more feature sets; and a set of new guidelines, the new set of guidelines for implementation in the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering (see MPEP § 2106.05(g)).
output the set of changes as a final set of changes to be made to the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering and outputting (see MPEP § 2106.05(g)).
Accordingly, these additional elements do not integrate the judicial exception into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claim is directed to the abstract idea.
Step 2B
Independent Claims
Additional elements
Claim 1:
at a data gathering stage: receiving a structure map of an in-use computing environment, the structure map of the in-use computing environment involving:
one or more sets of features; and one or more dependencies between features included in the one or more sets of features; receiving a set of new guidelines, the set of new guidelines for implementation in the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering, which is well-understood, routine, and conventional activity (see MPEP § 2106.05(d), “receiving/ transmitting data”).
using a large language model; inputting the vocabulary into an artificially-intelligent engine; at a machine learning stage, at the artificially-intelligent engine: at an autoencoder located within the artificially-intelligent engine; at a two-class classifier located within the artificially-intelligent engine; at a quantum engine; at a quantum simulator within the quantum engine; a complexity gradient simulator within the quantum engine; at a diversity analyzer simulator within the quantum engine; at a controller within the quantum engine; at a state processing engine within the quantum engine - these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
outputting one or more mutually exclusive states from the controller; outputting a final recommendation from the state processing engine within the quantum engine; inputting the final recommendation to a transformer neural network; processing the final recommendation at the transformer neural network; outputting, from the transformer neural network , a final set of changes to the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering and outputting, which is well-understood, routine, and conventional activity (see MPEP § 2106.05(d), “receiving/ transmitting data”, “presenting offers”), performed by generic computer functions (see MPEP § 2106.05(f)). These limitations can also be viewed as generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
Claim 6:
at a data gathering stage: receiving a structure map of an in-use computing environment, the structure map of the in-use computing environment involving:
one or more sets of features; and one or more dependencies between features included in the one or more sets of features; receiving a set of new guidelines, the set of new guidelines for implementation in the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering, which is well-understood, routine, and conventional activity (see MPEP § 2106.05(d), “receiving/ transmitting data”).
using a large language model; inputting the vocabulary into an artificially-intelligent engine; at a machine learning stage, at the artificially-intelligent engine: at an autoencoder located within the artificially-intelligent engine; at a two-class classifier located within the artificially-intelligent engine; at a quantum engine; at a controller within the quantum engine; at a state processing engine within the quantum engine - these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
outputting one or more mutually exclusive states from the controller; outputting a final recommendation from the state processing engine within the quantum engine; inputting the final recommendation to a transformer neural network;
processing the final recommendation at the transformer neural network; and
outputting, from the transformer neural network, a final set of changes to the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering and outputting, which is well-understood, routine, and conventional activity (see MPEP § 2106.05(d), “receiving/ transmitting data”, “presenting offers”), performed by generic computer functions (see MPEP § 2106.05(f)). These limitations can also be viewed as generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
Claim 13:
A system for continuous system adaption, the system comprising: a receiver; a data processor; using a large language model; an artificially-intelligent engine; using an autoencoder; using a quantum simulator; at a second quantum simulator; at a third quantum simulator - these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
receive: a structure map of an in-use computing environment, the structure map comprising: details relating to one or more feature sets; details relating to one or more dependencies between features included in the one or more feature sets; and a set of new guidelines, the new set of guidelines for implementation in the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering, which is well-understood, routine, and conventional activity (see MPEP § 2106.05(d), “receiving/ transmitting data”).
output the set of changes as a final set of changes to be made to the in-use computing environment - these limitations amount to insignificant extra-solution activity of mere data gathering and outputting, which is well-understood, routine, and conventional activity (see MPEP § 2106.05(d), “receiving/ transmitting data”, “presenting offers”).
Accordingly, these additional elements do not amount to significantly more than the judicial exception. As such, the claim is patent ineligible.
Step 2A Prong 1
Dependent Claims
Claims 3, 10, 15:
iteratively scales down each inputted element, included in a set of input elements, to a lowest possible dimension (referred to as, a “bottleneck layer”) and recreates each input element included in the set of input elements until the set of input elements is reproduced; and extracts, from the bottleneck layer, the set of vectorized features - these limitations recite mathematical relationships and calculations in extracting the vectorized features.
Claims 5, 12, 17:
a t-distributed stochastic neighbor embedding model executes the visually classifying and categorizing features - these limitations recite specific mathematical algorithm used in visually classifying and categorizing features.
Claim 7:
grades a complexity of a set of changes to be made to the one or more sets of features - these limitations encompass mental processes.
Claim 8:
identifies a degree of change compelled by the in-use computing environment to comply with each of the one or more mutually exclusive states - these limitations encompass mental processes.
Thus, the claims recite the abstract idea.
Step 2A Prong 2
Dependent Claims
Additional elements
Claims 2, 9, 14:
implementing the final set of changes within the in-use computing environment – this limitation merely recites an idea of a solution or outcome with no restriction on how the result is accomplished and no description of the mechanism for accomplishing the result (see MPEP § 2106.05(f)).
Claims 3, 10, 15:
at an output layer of the autoencoder - these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
Claims 4, 11, 16:
the autoencoder is a stacked convolutional autoencoder - these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
Claims 5, 12, 17:
within the artificially-intelligent engine- these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
Accordingly, these additional elements do not integrate the judicial exception into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claims are directed to the abstract idea.
Step 2B
Dependent Claims
Additional elements
Claims 2, 9, 14:
implementing the final set of changes within the in-use computing environment – this limitation merely recites an idea of a solution or outcome with no restriction on how the result is accomplished and no description of the mechanism for accomplishing the result (see MPEP § 2106.05(f)).
Claims 3, 10, 15:
at an output layer of the autoencoder - these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
Claims 4, 11, 16:
the autoencoder is a stacked convolutional autoencoder - these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
Claims 5, 12, 17:
within the artificially-intelligent engine- these limitations are recited at a high-level of generality such that it amounts to no more than merely using generic computer components or generic class of machine learning models as a tool to apply the judicial exception (see MPEP § 2106.05(f)) and generally linking the use of a judicial exception to the field of machine learning models (see MPEP § 2106.05(h)).
Accordingly, these additional elements do not amount to significantly more than the judicial exception. As such, the claims are patent 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.
Claims 1-17 are rejected under 35 U.S.C. 103 as being unpatentable over Milind et al. (US 2023/0306134 A1 hereinafter Milind) in view of Wolf et al. (“Transformers: State-of-the-Art Natural Language Processing”, 2020, hereinafter Wolf), further in view of Sherif et al. (“Unsupervised clustering of SARS-CoV-2 using deep convolutional autoencoder”, 2022, hereinafter Sherif) and Ulyanov (US 2013/0096698 A1).
Regarding Claim 1, Milind teaches a method for continuous system adaption ([0004] computing systems must be managed to ensure that the data controls that have been implemented are properly maintained and/or are up to date to further ensure these certain types of data are handled in a manner that is secure from experiencing data-related incidents; [0005] updates/ changes are often made to regulatory frameworks, such as, defined within regulatory standards, regulations, and/or laws that are designed to aid entities in ensuring that the proper data controls are in place to address these technological advances; regulatory frameworks can define appropriate technical and organizational measures/data controls needed to secure certain types of data; [0047] if the change to the regulatory framework involves an addition of a data control to the framework or a modification to a data control in the framework, then the change detection computing system coordinate having one or more actions performed that may involve automatically setting up a task to have the data control implemented or implementation thereof modified on the affected computing system), the method comprising:
at a data gathering stage: receiving a structure map of an in-use computing environment ([0044] the subscription profile may include more detailed information on the entity such as the various computing systems of the entity that are involved in handling certain types of data, data controls and control actions thereof defined by the applicable regulatory frameworks that have been implemented within the various computing systems of the entity - thus, receiving structure map/ subscription profile describing the entity's in-use computing systems and the data controls implemented on them), the structure map of the in-use computing environment involving:
one or more sets of features ([0044] the subscription profile may identify the types of data (e.g., PII) handled by the entity; the subscription profile may include more detailed information on the entity such as the various computing systems of the entity that are involved in handling certain types of data, data controls and control actions thereof defined by the applicable regulatory frameworks that have been implemented within the various computing systems of the entity; [0056] the subscription profile identity the computing system(s) for the entity and what target data is being handled by the computing system(s). The target data handled by the computing systems and the data controls and control actions implemented on it are construed as set of features); and
one or more dependencies between features included in the one or more sets of features ([0007] the data control is used in handling target data; [0053] a data control for the computing system that involves a control action of encrypting the credit card data that is stored in the database (i.e., the control depends on the target data it governs); [0005] adding or modifying a control action for implementing a data control (i.e., control actions depend on their data control));
receiving a set of new guidelines, the set of new guidelines for implementation in the in-use computing environment ([0061] detect that a document on a particular regulatory framework has been updated (e.g., a new version of the document exists), a new bulletin on a particular regulatory framework has been posted, a new article on a particular regulatory framework has been published; [0069] system retrieves and/or receives new (updated) content on a particular regulatory framework; [0005] an update to a regulatory framework may involve adding or modifying a control action for implementing a data control);
at a data processing stage: tokenizing the structure map and the set of new guidelines into a first plurality of words directed to the structure map and a second plurality of words directed to the set of new guidelines ([0083] the detect change module detects changes made to a regulatory framework by comparing a new (updated) version of a dataset containing content on the regulatory framework with a previous (existing) version of the dataset; [0086] the detect change module remove stop words from the text of the content, perform stemming on the text of the content, perform normalization on the text of the content, and/or the like; such preprocessing aid the detect change module in performing the comparison of the two versions of the dataset to identify differences between the versions of the dataset; [0091] the process change module performs keyword extraction on the text found in at least a portion of the dataset discussing (describing) the change to the regulatory framework to generate one or more tags representing characteristics of the change to the regulatory framework; [0093] the process change module processes the portion of the of the dataset using a featurization technique, such as bag of words (i.e., splitting the text into individual words), TFiDF, Word2Vec, and/or the like, to generate a feature representation of the change to the regulatory framework (e.g., change to the data control); [0045] the change detection computing system may compare the applicable domain(s) and/or jurisdiction(s) for the change to the regulatory framework to the domain(s) and/or jurisdiction(s) identified in the entity's subscription profile to determine whether the change to the regulatory framework affects any of the computing systems of the entity handling the target data (e.g., to determine a match between an applicable domain and/or jurisdiction to a domain or jurisdiction found in the subscription profile) - thus, tokenizing into words directed to the structure map and the updated/ new guidelines);
pre-contextually embedding the first plurality of words and the second plurality of words included in the vocabulary ([0093] the process change module processes the portion of the of the dataset using a featurization technique, such as bag of words, TFiDF, Word2Vec, and/or the like, to generate a feature representation of the change to the regulatory framework (e.g., change to the data control); the feature representation may be a vector representation that includes a plurality of values in which each value represents an attribute of the text found in the portion of the dataset discussing the change - thus, the words are embedded with Word2Vec/pre-contextual word embedding);
inputting the vocabulary into an artificially-intelligent engine ([0093] the process change module processes the portion of the of the dataset using a featurization technique, such as bag of words, TFiDF, Word2Vec, and/or the like, to generate a feature representation of the change to the regulatory framework (e.g., change to the data control); [0094] the process change module processes the feature representation using a first machine-learning model to generate a plurality of tags representing characteristics of the change to the regulatory framework; [0064] using multiple machine-learning models in identifying the characteristics of the change);
at a machine learning stage, at the artificially-intelligent engine ([0094] the process change module processes the feature representation using a first machine-learning model to generate a plurality of tags representing characteristics of the change to the regulatory framework; [0064] using multiple machine-learning models in identifying the characteristics of the change):
extracting a set of vectorized features ([0093] the process change module processes the portion of the of the dataset using a featurization technique, such as bag of words, TFiDF, Word2Vec, and/or the like, to generate a feature representation of the change to the regulatory framework (e.g., change to the data control); the feature representation may be a vector representation that includes a plurality of values in which each value represents an attribute of the text found in the portion of the dataset discussing the change; [0094] the process change module processes the feature representation using a first machine-learning model to generate a plurality of tags representing characteristics of the change to the regulatory framework);
at a two-class classifier located within the artificially-intelligent engine, categorizing objects within the structure map as impacted or unimpacted ([0094] the first machine-learning model may be configured with a one-versus-rest classifier such as a support vector machine, logistic regression, and/or the like (i.e., a set of binary/ two-class classifiers); [0097] the first feature representation(i.e., output of the model) contain a prediction (e.g., value) as to whether the change to the regulatory framework is applicable to the corresponding domain; [0101] perform action module determines whether the change to the regulatory framework affects (may affect) one or more computing systems 180 of the entity (i.e., impacted/ unimpacted));
grading a complexity of a set of changes to be made to the one or more sets of features ([0108] the perform action module use various characteristics, and/or combinations thereof, in setting a severity level for the change; use various indicators in representing the severity level such as degrees (e.g., high, medium, low,) numerical values (e.g., a range from 1 to 10); [0110] the perform action module base the priority level on what actions are to be performed (e.g., the amount of time required to perform the actions), what types of entity computing system(s) 180 are affected by the change, and/or the like); and
identifying a degree of change compelled by the in-use computing environment to comply with the set of changes ([0058] the severity of the change satisfies a certain threshold (e.g., high, or medium, as opposed to low) with respect to its effect on a computing system of the entity; [0095] one or more tags may identify one or more types for a data control associated with the change to the regulatory framework; one or more of the tags may identify characteristics of a computing system, or components thereof, to which the change to the regulatory framework applies; a particular tag may identify that the change is applicable to a database implementation, a Web server, a particular software application, and/or the like - thus, the tags identify the system components that must change); and
outputting a final recommendation ([0103] determines one or more recommendations to provide to the entity in addressing the change to the regulatory framework; [0121] the rules-based model to generate the one or more control actions that need to be performed to implement the change to the data control; [0124] the identify recommendation module 150 provide the identified control action(s) as the recommendation(s) for addressing the change to the regulatory standard);
outputting a final set of changes to the in-use computing environment ([0047] if the change to the regulatory framework involves an addition of a data control to the framework or a modification to a data control in the framework, then the change detection computing system coordinate having one or more actions performed that may involve automatically setting up a task to have the data control implemented or implementation thereof modified on the affected computing system; [0103] determines one or more recommendations to provide to the entity in addressing the change to the regulatory framework).
However, Milind does not expressly teach wherein creating, using a large language model, a vocabulary, from the tokenized structure map and the tokenized set of new guidelines, said vocabulary comprising the first plurality of words and the second plurality of words; padding and assigning sequence length to the first plurality of words and the second plurality of words included in the vocabulary; inputting the final recommendation to a transformer neural network; processing the final recommendation at the transformer neural network; and outputting, from the transformer neural network, a final set of changes.
In the same field of endeavor, Wolf teaches wherein creating, using a large language model, a vocabulary, from the tokenized structure map and the tokenized set of new guidelines, said vocabulary comprising the first plurality of words and the second plurality of words (page 38, Abstract - natural language processing has been driven by advances in both model architecture and model pretraining; the library consists of carefully engineered state of-the art Transformer architectures under a unified API; backing this library is a cu rated collection of pretrained models; page 41, section: Tokenizers - Tokenizer classes can be instantiated from a corresponding pretrained model; these classes store the vocabulary token-to-index map for their corresponding model; add additional token mappings, special tokens, or otherwise resize the vocabulary); padding and assigning sequence length to the first plurality of words and the second plurality of words included in the vocabulary (page 41, section: Tokenizers - Tokenizers implement additional useful features such as token type indices in the case of sequence classification to maximum length sequence truncating taking into account the added model-specific special tokens; most pretrained Transformer models have a maxi mum sequence length; page 41, section: Transformers - transformer architecture handle positional representations, masking, padding, and the use of sequence-to-sequence design); inputting text to a transformer neural network; processing the text at the transformer neural network; and outputting, from the transformer neural network, the results (page 39, section: Library Design - every model in the library includes… a tokenizer, which converts raw text to sparse index encodings; a transformer, which transforms sparse indices to contextual embeddings (i.e., processing at the transformer neural network); page 42, fig. 3 - Users can enter arbitrary text and a full version of the model is deployed on the fly to produce a summary; page 4, section: Transformers - the models target… understanding, generation, and conditional generation; page 41, section: Heads - some heads, such as conditional generation, support extra functionality like sampling and beam search. See page 40, figure 2 - Conditional generation input and output).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein creating, using a large language model, a vocabulary, from the tokenized structure map and the tokenized set of new guidelines, said vocabulary comprising the first plurality of words and the second plurality of words; padding and assigning sequence length to the first plurality of words and the second plurality of words included in the vocabulary; inputting text to a transformer neural network; processing text at the transformer neural network; and outputting, from the transformer neural network, the results, as suggested in Wolf into Milind. Doing so would be desirable because it would allow for effectively utilizing the capacity of transformer for a wide variety of tasks (Wolf, page 38, Abstract) and would provide useful features for the users (Wolf, page 41, section: Tokenizers).
However, Milind and Wolf do not expressly teach wherein at an autoencoder located within the artificially-intelligent engine, extracting a set of vectorized features; visually classifying and categorizing features, within the set of vectorized features, that are closely related to each other.
In the same field of endeavor, Sherif teaches wherein at an autoencoder located within the artificially-intelligent engine, extracting a set of vectorized features (page 11, section: Convolutional Autoencoder - the convolutional autoencoder performs… automated feature extraction and dimensionality reduction; page 12, section: Results - the last layer in the encoder is referred to as the bottleneck layer; page 13, section: Results - lower-dimensional feature vector produced from the encoder module hE has 45 features (i.e., set of vectorized features)); visually classifying and categorizing features, within the set of vectorized features, that are closely related to each other (page 13, section: Results - we utilized the lower-dimensional feature space (coded vector) in K-means clustering; page 14, section: The effect of number of clusters - we used T-distributed stochastic neighbor embedding (t-SNE) to show the outcomes of the deep clustering of features; page 14, section: The effect of number of clusters - we used T-distributed stochastic neighbor embedding (t-SNE) to show the outcomes of the deep clustering; the strains were isolated correctly between clusters in the t-SNE plot; comparative visualization of SARS-CoV-2 grouping into four, five, and six clusters was shown in Fig. 3; the visual results of clustered SARS-CoV-2 demonstrate that the best number of clusters was six as shown in the silhouette score (i.e., closely related items are visually grouped into categories)).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein at an autoencoder located within the artificially-intelligent engine, extracting a set of vectorized features; visually classifying and categorizing features, within the set of vectorized features, that are closely related to each other, as suggested in Sherif into Milind and Wolf. Doing so would be desirable because it would enable a better understanding of the major features (Sherif, page 1, section: Abstract).
However, Milind, Wolf, and Sherif do not expressly teach wherein at a quantum engine: simulating substantially all possible parameter and feature combinations; and based on the simulating, identifying resource consumption of each of the combinations at a quantum simulator within the quantum engine; based on the simulating, grading at a complexity gradient simulator within the quantum engine; based on the simulating, identifying a change at a diversity analyzer simulator within the quantum engine; splitting the combinations into multiple states of existence using a plurality of rules at a controller within the quantum engine; outputting one or more mutually exclusive states from the controller; processing each of the one or more mutually exclusive states at a state processing engine within the quantum engine; outputting a final recommendation from the state processing engine within the quantum engine.
In the same field of endeavor, Ulyanov teaches wherein at a quantum engine: simulating substantially all possible parameter and feature combinations ([0050] a self-organizing controller can include in knowledge bases of candidate controllers and a quantum inference unit; [0051] processing the n sets of controller gains by using quantum computing operations; the quantum state of a controller gain is a superposition of a real controller state and a virtual controller state (i.e., all candidate parameter/ gain combinations are represented simultaneously); [0052] the quantum inference unit can combine features of the each independent knowledge base; [0055] quantum computing techniques can be implemented on a quantum computer or simulated, using classical efficient simulation methods of quantum algorithms); and based on the simulating, identifying resource consumption of each of the combinations at a quantum simulator within the quantum engine ([0141] the genetic algorithm unit can include a simulator configured to estimate the fitness of certain solutions; [0143] simulations, rather than real experiments, can be convenient for running a large number of experiments; [0144] the fitness requirements of an algorithm executed by the genetic algorithm unit evaluated on the fitness function unit by the following equation: ; [0145] simulated solutions evaluated against restrictions on control object output parameters such that a control force and generalized entropy are minimum (i.e. resource consumption)); based on the simulating, grading at a complexity gradient simulator within the quantum engine ([0141] the genetic algorithm unit can include a simulator configured to estimate the fitness of certain solutions; [0143] simulations… for running a large number of experiments); based on the simulating, identifying a change at a diversity analyzer simulator within the quantum engine ([0141] the genetic algorithm unit can include a simulator configured to estimate the fitness of certain solutions; [0143] simulations… for running a large number of experiments; [0144] the fitness requirements of an algorithm executed by the genetic algorithm unit evaluated on the fitness function unit by the following equation: ; [0145] simulated solutions evaluated against restrictions on control object output parameters such that a control force and generalized entropy are minimum; [0062] when the information risk increment is large, it can be desirable to change the control structure and/or controller gains; where the information risk increment is small, it can be desirable to utilize the controller); splitting the combinations into multiple states of existence using a plurality of rules at a controller within the quantum engine ([0091] intelligent quantum tuning unit include a robust knowledge base and a quantum inference unit; [0033] the quantum inference unit can be a quantum fuzzy inference unit configured for providing one or more controller gains to a fuzzy controller (i.e., controller within the quantum engine); [0051] processing the n sets of controller gains by using quantum computing operations; the quantum state of a controller gain is a superposition of a real controller state and a virtual controller state; [0094] the knowledge bases can include information risk production rules; [0093] knowledge bases can be developed using fuzzy, neural, fuzzy-neural, or the like techniques; knowledge bases… to produce reliable sets of controller gains (i.e., combinations); [0125] it can be sufficient to obtain real-time reactions of production rules of the n knowledge bases; [0111] the real controller gains can be associated with their real quantum states; a probability can be associated with each real state; [0112] the probability of the virtual state can be calculated based on the probability of the real state; virtual state corresponding to a virtual gain that can subsequently be extracted from the knowledge base; [0113] the virtual gains can be determined by evaluating a function that approximates the inverse map; [0114] a quantum state can be generated based the real state and the virtual state and their corresponding probabilities - thus, each candidate is encoded into quantum states built from a real controller state and a virtual controller state (i.e., split into multiple states of existence according to the rule-based knowledge bases and the probability map)); outputting one or more mutually exclusive states from the controller ([0051] the quantum inference unit can correlate two or more of the encoded quantum states to create a superposition signal of correlated states (i.e., output); [0100] the quantum encoder (controller) can encode the n sets of controller gains into corresponding quantum states; the quantum encoder can include a probability database; the quantum states can be provided to the quantum correlator; [0051] processing the n sets of controller gains by using quantum computing operations; the quantum state of a controller gain is a superposition of a real controller state and a virtual controller state; [0112] the probability of the virtual state can be calculated based on the probability of the real state; pi,j 0+pi,j 1=1 (i.e., real and virtual states are complementary alternatives, whose probabilities sum to 1 and therefore, mutually exclusive)); processing each of the one or more mutually exclusive states at a state processing engine within the quantum engine ([0051] the quantum interference unit can perform an interference operation, which selects a correlated state; the correlated state can be selected based on having the highest probability; the selected correlated quantum state can be decoded into a controller gain (i.e., inference operation processes correlated states)); outputting a final recommendation from the state processing engine within the quantum engine ([0051] the selected correlated quantum state can be decoded into a controller gain (i.e., robust parameter set output/ final recommendation).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein at a quantum engine: simulating substantially all possible parameter and feature combinations; and based on the simulating, identifying resource consumption of each of the combinations at a quantum simulator within the quantum engine; based on the simulating, grading at a complexity gradient simulator within the quantum engine; based on the simulating, identifying a change at a diversity analyzer simulator within the quantum engine; splitting the combinations into multiple states of existence using a plurality of rules at a controller within the quantum engine; outputting one or more mutually exclusive states from the controller; processing each of the one or more mutually exclusive states at a state processing engine within the quantum engine; outputting a final recommendation from the state processing engine within the quantum engine, as suggested in Ulyanov into Milind, Wolf, and Sherif. Doing so would be desirable because simulations, rather than real experiments, can be convenient for running a large number of experiments (Ulyanov [0143]) and that the quantum inference unit can combine features of each independent knowledge base and/or generate a robust controller gain for a situation not addressed by any individual knowledge base by extracting information from the collection of knowledge bases (Ulyanov [0052]).
As to dependent Claim 2, Milind, Wolf, Sherif, and Ulyanov teach all the limitations of claim 1. Milind further teaches wherein implementing the final set of changes within the in-use computing environment ([0047] if the change to the regulatory framework involves an addition of a data control to the framework or a modification to a data control in the framework, then the change detection computing system coordinate having one or more actions performed that may involve automatically setting up a task to have the data control implemented or implementation thereof modified on the affected computing system; [0112] computing system associated with the entity that then automatically performs the action to address the change to the regulatory framework).
As to dependent Claim 3, Milind, Wolf, Sherif, and Ulyanov teach all the limitations of claim 1. Sherif further teaches wherein the autoencoder: iteratively scales down each inputted element, included in a set of input elements, to a lowest possible dimension (referred to as, a “bottleneck layer”) and recreates each input element included in the set of input elements until the set of input elements is reproduced at an output layer of the autoencoder; and extracts, from the bottleneck layer, the set of vectorized features (page 11, section: convolutional autoencoder - progressively decreasing the bottleneck dimension during pre-training until the reconstruction loss begins to noticeably increase (i.e., bottle neck is reduced step by step); page 12, section: Results - the decoder module is trained to recover the original input from the encoder module; we repeat the procedure of fine-tuning until the reconstruction error improves by no more than 6.4045e-04 (a set amount) (i.e., until the input is reproduces at the output layer); the last layer in the encoder is referred to as the bottleneck layer; page 13 - lower-dimensional feature vector produced from the encoder module (i.e., vectorized features)).
As to dependent Claim 4, Milind, Wolf, Sherif, and Ulyanov teach all the limitations of claim 3. Sherif further teaches wherein the autoencoder is a stacked convolutional autoencoder (page 10 - Convolutional Autoencoder (CAE) can have multiple convolutions and pooling layers; page 12 - the encoder component hE is composed of two convolutional layers and two max-pooling layers (i.e., stacked layers); the last layer in the encoder is referred to as the bottleneck layer; the decoder module is trained to recover the original input from the encoder module).
As to dependent Claim 5, Milind, Wolf, Sherif, and Ulyanov teach all the limitations of claim 1. Sherif further teaches wherein a t-distributed stochastic neighbor embedding model within the artificially-intelligent engine executes the visually classifying and categorizing features (page 14, section: The effect of number of clusters - we used T-distributed stochastic neighbor embedding (t-SNE) to show the outcomes of the deep clustering of feature; the strains were isolated correctly between clusters in the t-SNE plot; comparative visualization of SARS-CoV-2 grouping into four, five, and six clusters was shown in Fig. 3).
Regarding Claim 6, Milind teaches a method for continuous system adaption ([0004] computing systems must be managed to ensure that the data controls that have been implemented are properly maintained and/or are up to date to further ensure these certain types of data are handled in a manner that is secure from experiencing data-related incidents; [0005] updates/ changes are often made to regulatory frameworks, such as, defined within regulatory standards, regulations, and/or laws that are designed to aid entities in ensuring that the proper data controls are in place to address these technological advances; regulatory frameworks can define appropriate technical and organizational measures/data controls needed to secure certain types of data; [0047] if the change to the regulatory framework involves an addition of a data control to the framework or a modification to a data control in the framework, then the change detection computing system coordinate having one or more actions performed that may involve automatically setting up a task to have the data control implemented or implementation thereof modified on the affected computing system), the method comprising:
at a data gathering stage: receiving a structure map of an in-use computing environment ([0044] the subscription profile may include more detailed information on the entity such as the various computing systems of the entity that are involved in handling certain types of data, data controls and control actions thereof defined by the applicable regulatory frameworks that have been implemented within the various computing systems of the entity - thus, receiving structure map/ subscription profile describing the entity's in-use computing systems and the data controls implemented on them), the structure map of the in-use computing environment involving:
one or more sets of features ([0044] the subscription profile may identify the types of data (e.g., PII) handled by the entity; the subscription profile may include more detailed information on the entity such as the various computing systems of the entity that are involved in handling certain types of data, data controls and control actions thereof defined by the applicable regulatory frameworks that have been implemented within the various computing systems of the entity; [0056] the subscription profile identity the computing system(s) for the entity and what target data is being handled by the computing system(s). The target data handled by the computing systems and the data controls and control actions implemented on it are construed as set of features); and
one or more dependencies between features included in the one or more sets of features ([0007] the data control is used in handling target data; [0053] a data control for the computing system that involves a control action of encrypting the credit card data that is stored in the database (i.e., the control depends on the target data it governs); [0005] adding or modifying a control action for implementing a data control (i.e., control actions depend on their data control));
receiving a set of new guidelines, the set of new guidelines for implementation in the in-use computing environment ([0061] detect that a document on a particular regulatory framework has been updated (e.g., a new version of the document exists), a new bulletin on a particular regulatory framework has been posted, a new article on a particular regulatory framework has been published; [0069] system retrieves and/or receives new (updated) content on a particular regulatory framework; [0005] an update to a regulatory framework may involve adding or modifying a control action for implementing a data control);
at a data processing stage: tokenizing the structure map into a first plurality of grams; tokenizing the set of new guidelines into a second plurality of grams ([0083] the detect change module detects changes made to a regulatory framework by comparing a new (updated) version of a dataset containing content on the regulatory framework with a previous (existing) version of the dataset; [0086] the detect change module remove stop words from the text of the content, perform stemming on the text of the content, perform normalization on the text of the content, and/or the like; such preprocessing aid the detect change module in performing the comparison of the two versions of the dataset to identify differences between the versions of the dataset; [0091] the process change module performs keyword extraction on the text found in at least a portion of the dataset discussing (describing) the change to the regulatory framework to generate one or more tags representing characteristics of the change to the regulatory framework; [0093] the process change module processes the portion of the of the dataset using a featurization technique, such as bag of words (i.e., splitting the text into individual words/ grams), TFiDF, Word2Vec, and/or the like, to generate a feature representation of the change to the regulatory framework (e.g., change to the data control); [0045] the change detection computing system may compare the applicable domain(s) and/or jurisdiction(s) for the change to the regulatory framework to the domain(s) and/or jurisdiction(s) identified in the entity's subscription profile to determine whether the change to the regulatory framework affects any of the computing systems of the entity handling the target data (e.g., to determine a match between an applicable domain and/or jurisdiction to a domain or jurisdiction found in the subscription profile) - thus, tokenizing into words/ grams directed to the structure map and the updated/ new guidelines);
inputting the vocabulary into an artificially-intelligent engine ([0093] the process change module processes the portion of the of the dataset using a featurization technique, such as bag of words, TFiDF, Word2Vec, and/or the like, to generate a feature representation of the change to the regulatory framework (e.g., change to the data control); [0094] the process change module processes the feature representation using a first machine-learning model to generate a plurality of tags representing characteristics of the change to the regulatory framework; [0064] using multiple machine-learning models in identifying the characteristics of the change);
at a machine learning stage, at the artificially-intelligent engine ([0094] the process change module processes the feature representation using a first machine-learning model to generate a plurality of tags representing characteristics of the change to the regulatory framework; [0064] using multiple machine-learning models in identifying the characteristics of the change):
extracting a set of vectorized features from the vocabulary ([0093] the process change module processes the portion of the of the dataset using a featurization technique, such as bag of words, TFiDF, Word2Vec, and/or the like, to generate a feature representation of the change to the regulatory framework (e.g., change to the data control); the feature representation may be a vector representation that includes a plurality of values in which each value represents an attribute of the text found in the portion of the dataset discussing the change; [0094] the process change module processes the feature representation using a first machine-learning model to generate a plurality of tags representing characteristics of the change to the regulatory framework);
at a two-class classifier located within the artificially-intelligent engine, categorizing each feature, included in the one or more sets of features, within the structure map as impacted or unimpacted ([0094] the first machine-learning model may be configured with a one-versus-rest classifier such as a support vector machine, logistic regression, and/or the like (i.e., a set of binary/ two-class classifiers); [0097] the first feature representation(i.e., output of the model) contain a prediction (e.g., value) as to whether the change to the regulatory framework is applicable to the corresponding domain; [0101] perform action module determines whether the change to the regulatory framework affects (may affect) one or more computing systems 180 of the entity; [0102] determining whether the entity has a particular computing system 180 that is affected by the change (i.e., impacted/ unimpacted));
outputting a final recommendation, said final recommendation comprising an assignable parameter for each feature ([0103] determines one or more recommendations to provide to the entity in addressing the change to the regulatory framework; [0121] the rules-based model to generate the one or more control actions that need to be performed to implement the change to the data control; [0124] the identify recommendation module 150 provide the identified control action(s) as the recommendation(s) for addressing the change to the regulatory standard);
outputting a final set of changes to the in-use computing environment ([0047] if the change to the regulatory framework involves an addition of a data control to the framework or a modification to a data control in the framework, then the change detection computing system coordinate having one or more actions performed that may involve automatically setting up a task to have the data control implemented or implementation thereof modified on the affected computing system; [0103] determines one or more recommendations to provide to the entity in addressing the change to the regulatory framework).
However, Milind does not expressly teach wherein creating, using a large language model, a vocabulary, from the first plurality of grams and the second plurality of grams; padding and assigning sequence length to the first plurality of grams within the vocabulary; padding and assigning sequence length to the second plurality of grams within the vocabulary; inputting the final recommendation to a transformer neural network; processing the final recommendation at the transformer neural network; and outputting, from the transformer neural network, a final set of changes.
In the same field of endeavor, Wolf teaches wherein creating, using a large language model, a vocabulary, from the first plurality of grams and the second plurality of grams (page 38, Abstract - natural language processing has been driven by advances in both model architecture and model pretraining; the library consists of carefully engineered state of-the art Transformer architectures under a unified API; backing this library is a cu rated collection of pretrained models; page 41, section: Tokenizers - Tokenizer classes can be instantiated from a corresponding pretrained model; these classes store the vocabulary token-to-index map for their corresponding model; add additional token mappings, special tokens, or otherwise resize the vocabulary); padding and assigning sequence length to the first plurality of grams within the vocabulary and padding and assigning sequence length to the second plurality of grams within the vocabulary (page 41, section: Tokenizers - Tokenizers implement additional useful features such as token type indices in the case of sequence classification to maximum length sequence truncating taking into account the added model-specific special tokens; most pretrained Transformer models have a maxi mum sequence length; page 41, section: Transformers - transformer architecture handle positional representations, masking, padding, and the use of sequence-to-sequence design); inputting text to a transformer neural network; processing the text at the transformer neural network; and outputting, from the transformer neural network, the results (page 39, section: Library Design - every model in the library includes… a tokenizer, which converts raw text to sparse index encodings; a transformer, which transforms sparse indices to contextual embeddings (i.e., processing at the transformer neural network); page 42, fig. 3 - Users can enter arbitrary text and a full version of the model is deployed on the fly to produce a summary; page 4, section: Transformers - the models target… understanding, generation, and conditional generation; page 41, section: Heads - some heads, such as conditional generation, support extra functionality like sampling and beam search. See page 40, figure 2 - Conditional generation input and output).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein creating, using a large language model, a vocabulary, from the first plurality of grams and the second plurality of grams; padding and assigning sequence length to the first plurality of grams within the vocabulary; padding and assigning sequence length to the second plurality of grams within the vocabulary; inputting text to a transformer neural network; processing the text at the transformer neural network; and outputting, from the transformer neural network, the results, as suggested in Wolf into Milind. Doing so would be desirable because it would allow for effectively utilizing the capacity of transformer for a wide variety of tasks (Wolf, page 38, Abstract) and would provide useful features for the users (Wolf, page 41, section: Tokenizers).
However, Milind and Wolf do not expressly teach wherein at an autoencoder located within the artificially-intelligent engine, extracting a set of vectorized features; visually classifying and categorizing features, within the set of vectorized features, that are closely related to each other.
In the same field of endeavor, Sherif teaches wherein at an autoencoder located within the artificially-intelligent engine, extracting a set of vectorized features (page 11, section: Convolutional Autoencoder - the convolutional autoencoder performs… automated feature extraction and dimensionality reduction; page 12, section: Results - the last layer in the encoder is referred to as the bottleneck layer; page 13, section: Results - lower-dimensional feature vector produced from the encoder module hE has 45 features (i.e., set of vectorized features)); visually classifying and categorizing features, within the set of vectorized features, that are closely related to each other (page 13, section: Results - we utilized the lower-dimensional feature space (coded vector) in K-means clustering; page 14, section: The effect of number of clusters - we used T-distributed stochastic neighbor embedding (t-SNE) to show the outcomes of the deep clustering of features; page 14, section: The effect of number of clusters - we used T-distributed stochastic neighbor embedding (t-SNE) to show the outcomes of the deep clustering; the strains were isolated correctly between clusters in the t-SNE plot; comparative visualization of SARS-CoV-2 grouping into four, five, and six clusters was shown in Fig. 3; the visual results of clustered SARS-CoV-2 demonstrate that the best number of clusters was six as shown in the silhouette score (i.e., closely related items are visually grouped into categories)).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein at an autoencoder located within the artificially-intelligent engine, extracting a set of vectorized features; visually classifying and categorizing features, within the set of vectorized features, that are closely related to each other, as suggested in Sherif into Milind and Wolf. Doing so would be desirable because it would enable a better understanding of the major features (Sherif, page 1, section: Abstract).
However, Milind, Wolf, and Sherif do not expressly teach wherein at a quantum engine: simulating substantially all combinations of features included in the one or more sets of features and all combinations of one or more parameters assignable to features included in the one or more sets of features; identifying resource consumption of each of the combinations of features; splitting the combinations of features into multiple states of existence using a plurality of rules at a controller within the quantum engine; outputting one or more mutually exclusive states from the controller; processing each of the one or more mutually exclusive states at a state processing engine within the quantum engine; outputting a final recommendation from the state processing engine within the quantum engine.
In the same field of endeavor, Ulyanov teaches wherein at a quantum engine: simulating substantially all combinations of features included in the one or more sets of features and all combinations of one or more parameters assignable to features included in the one or more sets of features ([0050] a self-organizing controller can include in knowledge bases of candidate controllers and a quantum inference unit; [0051] processing the n sets of controller gains by using quantum computing operations; the quantum state of a controller gain is a superposition of a real controller state and a virtual controller state (i.e., all candidate parameter/ gain combinations are represented simultaneously); [0052] the quantum inference unit can combine features of the each independent knowledge base; [0055] quantum computing techniques can be implemented on a quantum computer or simulated, using classical efficient simulation methods of quantum algorithms); identifying resource consumption of each of the combinations of features ([0141] the genetic algorithm unit can include a simulator configured to estimate the fitness of certain solutions; [0143] simulations, rather than real experiments, can be convenient for running a large number of experiments; [0144] the fitness requirements of an algorithm executed by the genetic algorithm unit evaluated on the fitness function unit by the following equation: ; [0145] simulated solutions evaluated against restrictions on control object output parameters such that a control force and generalized entropy are minimum (i.e. resource consumption)); splitting the combinations of features into multiple states of existence using a plurality of rules at a controller within the quantum engine ([0091] intelligent quantum tuning unit include a robust knowledge base and a quantum inference unit; [0033] the quantum inference unit can be a quantum fuzzy inference unit configured for providing one or more controller gains to a fuzzy controller (i.e., controller within the quantum engine); [0051] processing the n sets of controller gains by using quantum computing operations; the quantum state of a controller gain is a superposition of a real controller state and a virtual controller state; [0094] the knowledge bases can include information risk production rules; [0093] knowledge bases can be developed using fuzzy, neural, fuzzy-neural, or the like techniques; knowledge bases… to produce reliable sets of controller gains (i.e., combinations); [0125] it can be sufficient to obtain real-time reactions of production rules of the n knowledge bases; [0111] the real controller gains can be associated with their real quantum states; a probability can be associated with each real state; [0112] the probability of the virtual state can be calculated based on the probability of the real state; virtual state corresponding to a virtual gain that can subsequently be extracted from the knowledge base; [0113] the virtual gains can be determined by evaluating a function that approximates the inverse map; [0114] a quantum state can be generated based the real state and the virtual state and their corresponding probabilities - thus, each candidate is encoded into quantum states built from a real controller state and a virtual controller state (i.e., split into multiple states of existence according to the rule-based knowledge bases and the probability map)); outputting one or more mutually exclusive states from the controller ([0051] the quantum inference unit can correlate two or more of the encoded quantum states to create a superposition signal of correlated states (i.e., output); [0100] the quantum encoder (controller) can encode the n sets of controller gains into corresponding quantum states; the quantum encoder can include a probability database; the quantum states can be provided to the quantum correlator; [0051] processing the n sets of controller gains by using quantum computing operations; the quantum state of a controller gain is a superposition of a real controller state and a virtual controller state; [0112] the probability of the virtual state can be calculated based on the probability of the real state; pi,j 0+pi,j 1=1 (i.e., real and virtual states are complementary alternatives, whose probabilities sum to 1 and therefore, mutually exclusive)); processing each of the one or more mutually exclusive states at a state processing engine within the quantum engine ([0051] the quantum interference unit can perform an interference operation, which selects a correlated state; the correlated state can be selected based on having the highest probability; the selected correlated quantum state can be decoded into a controller gain (i.e., inference operation processes correlated states)); outputting a final recommendation from the state processing engine within the quantum engine ([0051] the selected correlated quantum state can be decoded into a controller gain (i.e., robust parameter set output/ final recommendation); [0106] certain quantum states of the superposition signal can be selected to form the robust quantum state based on selection criteria; the robust quantum states can be optimal quantum states; [0144] the results of the simulation can be evaluated to determine whether the fitness requirement is satisfied; [0146] when the fitness requirements are satisfied, the correlation type corresponding to the closed loop simulation executed at block 908 can be selected as the selected correlation type).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein at a quantum engine: simulating substantially all combinations of features included in the one or more sets of features and all combinations of one or more parameters assignable to features included in the one or more sets of features; identifying resource consumption of each of the combinations of features; splitting the combinations of features into multiple states of existence using a plurality of rules at a controller within the quantum engine; outputting one or more mutually exclusive states from the controller; processing each of the one or more mutually exclusive states at a state processing engine within the quantum engine; outputting a final recommendation from the state processing engine within the quantum engine, as suggested in Ulyanov into Milind, Wolf, and Sherif. Doing so would be desirable because simulations, rather than real experiments, can be convenient for running a large number of experiments (Ulyanov [0143]) and that the quantum inference unit can combine features of the each independent knowledge base and/or generate a robust controller gain for a situation not addressed by any individual knowledge base by extracting information from the collection of knowledge bases (Ulyanov [0052]).
As to dependent Claim 7, Milind, Wolf, Sherif, and Ulyanov teach all the limitations of claim 6. Milind further taches wherein grades a complexity of a set of changes to be made to the one or more sets of features ([0108] the perform action module use various characteristics, and/or combinations thereof, in setting a severity level for the change; use various indicators in representing the severity level such as degrees (e.g., high, medium, low,) numerical values (e.g., a range from 1 to 10); [0110] the perform action module base the priority level on what actions are to be performed (e.g., the amount of time required to perform the actions), what types of entity computing system(s) 180 are affected by the change, and/or the like). Ulyanov further teaches wherein the quantum engine grades ([0141] the genetic algorithm unit can include a simulator configured to estimate the fitness of certain solutions; [0143] simulations… for running a large number of experiments).
As to dependent Claim 8, Milind, Wolf, Sherif, and Ulyanov teach all the limitations of claim 6. Milind further taches wherein identifies a degree of change compelled by the in-use computing environment to comply with states ([0058] the severity of the change satisfies a certain threshold (e.g., high, or medium, as opposed to low) with respect to its effect on a computing system of the entity; [0095] one or more tags may identify one or more types for a data control associated with the change to the regulatory framework; one or more of the tags may identify characteristics of a computing system, or components thereof, to which the change to the regulatory framework applies; a particular tag may identify that the change is applicable to a database implementation, a Web server, a particular software application, and/or the like - thus, the tags identify the system components that must change). Ulyanov further teaches wherein the quantum engine identifies change with each of the one or more mutually exclusive states([0141] the genetic algorithm unit can include a simulator configured to estimate the fitness of certain solutions; [0143] simulations… for running a large number of experiments; [0144] the fitness requirements of an algorithm executed by the genetic algorithm unit evaluated on the fitness function unit by the following equation: ; [0145] simulated solutions evaluated against restrictions on control object output parameters such that a control force and generalized entropy are minimum;[0146] when the fitness requirements are satisfied, the correlation type corresponding to the closed loop simulation executed at block 908 can be selected as the selected correlation type; otherwise, a different correlation type can be chosen (i.e., evaluation performed at each state); [0100] the quantum encoder can encode the n sets of controller gains into corresponding quantum states; the quantum states 420 can be provided to the quantum correlator (i.e., outputting); [0051] processing the n sets of controller gains by using quantum computing operations; the quantum state of a controller gain is a superposition of a real controller state and a virtual controller state; [0112] the probability of the virtual state can be calculated based on the probability of the real state; pi,j 0+pi,j 1=1 (i.e., real and virtual states are complementary alternatives, whose probabilities sum to 1 and therefore, mutually exclusive); [0062] when the information risk increment is large, it can be desirable to change the control structure and/or controller gains; where the information risk increment is small, it can be desirable to utilize the controller).
Claims 9-12 recite similar limitations as claims 2-5 above and are likewise rejected.
Regarding Claim 13, Milind teaches a system for continuous system adaption ([0011] a system is provided comprising a non-transitory computer-readable medium storing instructions and a processing device; [0004] computing systems must be managed to ensure that the data controls that have been implemented are properly maintained and/or are up to date to further ensure these certain types of data are handled in a manner that is secure from experiencing data-related incidents; [0005] updates/ changes are often made to regulatory frameworks, such as, defined within regulatory standards, regulations, and/or laws that are designed to aid entities in ensuring that the proper data controls are in place to address these technological advances; regulatory frameworks can define appropriate technical and organizational measures/data controls needed to secure certain types of data; [0047] if the change to the regulatory framework involves an addition of a data control to the framework or a modification to a data control in the framework, then the change detection computing system coordinate having one or more actions performed that may involve automatically setting up a task to have the data control implemented or implementation thereof modified on the affected computing system; ), the system comprising:
a receiver operable to receive: a structure map of an in-use computing environment ([0044] the subscription profile may include more detailed information on the entity such as the various computing systems of the entity that are involved in handling certain types of data, data controls and control actions thereof defined by the applicable regulatory frameworks that have been implemented within the various computing systems of the entity - thus, receiving structure map/ subscription profile describing the entity's in-use computing systems and the data controls implemented on them), the structure map comprising:
details relating to one or more feature sets ([0044] the subscription profile may identify the types of data (e.g., PII) handled by the entity; the subscription profile may include more detailed information on the entity such as the various computing systems of the entity that are involved in handling certain types of data, data controls and control actions thereof defined by the applicable regulatory frameworks that have been implemented within the various computing systems of the entity; [0056] the subscription profile identity the computing system(s) for the entity and what target data is being handled by the computing system(s). The target data handled by the computing systems and the data controls and control actions implemented on it are construed as details of features);
details relating to one or more dependencies between features included in the one or more feature sets ([0007] the data control is used in handling target data; [0053] a data control for the computing system that involves a control action of encrypting the credit card data that is stored in the database (i.e., the control depends on the target data it governs); [0005] adding or modifying a control action for implementing a data control (i.e., control actions depend on their data control)); and
a set of new guidelines, the new set of guidelines for implementation in the in-use computing environment ([0061] detect that a document on a particular regulatory framework has been updated (e.g., a new version of the document exists), a new bulletin on a particular regulatory framework has been posted, a new article on a particular regulatory framework has been published; [0069] system retrieves and/or receives new (updated) content on a particular regulatory framework; [0005] an update to a regulatory framework may involve adding or modifying a control action for implementing a data control);
a data processor operable to: tokenize the structure map and the set of new guidelines into a first plurality of words directed to the structure map and a second plurality of words directed to the set of new guidelines ([0083] the detect change module detects changes made to a regulatory framework by comparing a new (updated) version of a dataset containing content on the regulatory framework with a previous (existing) version of the dataset; [0086] the detect change module remove stop words from the text of the content, perform stemming on the text of the content, perform normalization on the text of the content, and/or the like; such preprocessing aid the detect change module in performing the comparison of the two versions of the dataset to identify differences between the versions of the dataset; [0091] the process change module performs keyword extraction on the text found in at least a portion of the dataset discussing (describing) the change to the regulatory framework to generate one or more tags representing characteristics of the change to the regulatory framework; [0093] the process change module processes the portion of the of the dataset using a featurization technique, such as bag of words (i.e., splitting the text into individual words), TFiDF, Word2Vec, and/or the like, to generate a feature representation of the change to the regulatory framework (e.g., change to the data control); [0045] the change detection computing system may compare the applicable domain(s) and/or jurisdiction(s) for the change to the regulatory framework to the domain(s) and/or jurisdiction(s) identified in the entity's subscription profile to determine whether the change to the regulatory framework affects any of the computing systems of the entity handling the target data (e.g., to determine a match between an applicable domain and/or jurisdiction to a domain or jurisdiction found in the subscription profile) - thus, tokenizing into words directed to the structure map and the updated/ new guidelines);
an artificially-intelligent engine operable to: extract, using an autoencoder, a set of vectorized features from the vocabulary ([0094] the process change module processes the feature representation using a first machine-learning model to generate a plurality of tags representing characteristics of the change to the regulatory framework; [0064] using multiple machine-learning models in identifying the characteristics of the change; ([0093] the process change module processes the portion of the of the dataset using a featurization technique, such as bag of words, TFiDF, Word2Vec, and/or the like, to generate a feature representation of the change to the regulatory framework (e.g., change to the data control); the feature representation may be a vector representation that includes a plurality of values in which each value represents an attribute of the text found in the portion of the dataset discussing the change; [0094] the process change module processes the feature representation using a first machine-learning model to generate a plurality of tags representing characteristics of the change to the regulatory framework);
categorize features included in the one or more feature sets as impacted or unimpacted ([0094] the first machine-learning model may be configured with a one-versus-rest classifier such as a support vector machine, logistic regression, and/or the like ; [0097] the first feature representation contain a prediction (e.g., value) as to whether the change to the regulatory framework is applicable to the corresponding domain; [0101] perform action module determines whether the change to the regulatory framework affects (may affect) one or more computing systems 180 of the entity (i.e., impacted/ unimpacted));
identify a set of changes to be made to the one or more features sets ([0108] the perform action module use various characteristics, and/or combinations thereof, in setting a severity level for the change; use various indicators in representing the severity level such as degrees (e.g., high, medium, low,) numerical values (e.g., a range from 1 to 10); [0110] the perform action module base the priority level on what actions are to be performed (e.g., the amount of time required to perform the actions), what types of entity computing system(s) 180 are affected by the change, and/or the like);
grade a complexity of the set of changes to made to the one or more features sets ([0108] the perform action module use various characteristics, and/or combinations thereof, in setting a severity level for the change; use various indicators in representing the severity level such as degrees (e.g., high, medium, low,) numerical values (e.g., a range from 1 to 10); [0110] the perform action module base the priority level on what actions are to be performed (e.g., the amount of time required to perform the actions), what types of entity computing system(s) 180 are affected by the change, and/or the like);
re-identify a set of changes to be made to the one or more features sets based on the grade ([0108] the perform action module use various characteristics, and/or combinations thereof, in setting a severity level for the change; use various indicators in representing the severity level such as degrees (e.g., high, medium, low,) numerical values (e.g., a range from 1 to 10); [0110] the perform action module base the priority level on what actions are to be performed (e.g., the amount of time required to perform the actions), what types of entity computing system(s) 180 are affected by the change, and/or the like; [0107] the severity level can be used by the perform action module in determining what actions need to be performed to address the change (i.e., re-identify what actions to be performed based on grade));
identify a degree of change compelled by the in-use computing environment to comply with the set of changes ([0058] the severity of the change satisfies a certain threshold (e.g., high, or medium, as opposed to low) with respect to its effect on a computing system of the entity; [0095] one or more tags may identify one or more types for a data control associated with the change to the regulatory framework; one or more of the tags may identify characteristics of a computing system, or components thereof, to which the change to the regulatory framework applies; a particular tag may identify that the change is applicable to a database implementation, a Web server, a particular software application, and/or the like - thus, the tags identify the system components that must change);
re-identify a set of changes to be made to the one or more features sets based on the degree of change ([0095] one or more tags may identify one or more types for a data control associated with the change to the regulatory framework; one or more of the tags may identify characteristics of a computing system, or components thereof, to which the change to the regulatory framework applies; a particular tag may identify that the change is applicable to a database implementation, a Web server, a particular software application, and/or the like;[0058] the severity of the change satisfies a certain threshold (e.g., high, or medium, as opposed to low) with respect to its effect on a computing system of the entity; [0108] setting a severity level for the change; use various indicators in representing the severity level such as degrees (e.g., high, medium, low,) numerical values (e.g., a range from 1 to 10); [0107] determine a severity (severity level) of the change based on various characteristics of the change and/or computing system(s); the severity level can be used by the perform action module in determining what actions need to be performed to address the change (i.e., re-identify changes to be made)); and
output the set of changes as a final set of changes to be made to the in-use computing environment ([0047] if the change to the regulatory framework involves an addition of a data control to the framework or a modification to a data control in the framework, then the change detection computing system coordinate having one or more actions performed that may involve automatically setting up a task to have the data control implemented or implementation thereof modified on the affected computing system; [0103] determines one or more recommendations to provide to the entity in addressing the change to the regulatory framework).
However, Milind does not expressly teach wherein create, using a large language model, a vocabulary from the first plurality of words and the second plurality of words; pad and assign sequence length to the first plurality of words and second plurality of words included in the vocabulary.
In the same field of endeavor, Wolf teaches wherein create, using a large language model, a vocabulary from the first plurality of words and the second plurality of words (page 38, Abstract - natural language processing has been driven by advances in both model architecture and model pretraining; the library consists of carefully engineered state of-the art Transformer architectures under a unified API; backing this library is a cu rated collection of pretrained models; page 41, section: Tokenizers - Tokenizer classes can be instantiated from a corresponding pretrained model; these classes store the vocabulary token-to-index map for their corresponding model; add additional token mappings, special tokens, or otherwise resize the vocabulary); pad and assign sequence length to the first plurality of words and second plurality of words included in the vocabulary (page 41, section: Tokenizers - Tokenizers implement additional useful features such as token type indices in the case of sequence classification to maximum length sequence truncating taking into account the added model-specific special tokens; most pretrained Transformer models have a maxi mum sequence length; page 41, section: Transformers - transformer architecture handle positional representations, masking, padding, and the use of sequence-to-sequence design).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein create, using a large language model, a vocabulary from the first plurality of words and the second plurality of words; pad and assign sequence length to the first plurality of words and second plurality of words included in the vocabulary, as suggested in Wolf into Milind. Doing so would be desirable because it would allow for effectively utilizing the capacity of transformer for a wide variety of tasks (Wolf, page 38, Abstract) and would provide useful features for the users (Wolf, page 41, section: Tokenizers).
However, Milind and Wolf do not expressly teach wherein extract, using an autoencoder, a set of vectorized features.
In the same field of endeavor, Sherif teaches wherein extract, using an autoencoder, a set of vectorized features (page 11, section: Convolutional Autoencoder - the convolutional autoencoder performs… automated feature extraction and dimensionality reduction; page 12, section: Results - the last layer in the encoder is referred to as the bottleneck layer; page 13, section: Results - lower-dimensional feature vector produced from the encoder module hE has 45 features (i.e., set of vectorized features)).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein extract, using an autoencoder, a set of vectorized features, as suggested in Sherif into Milind and Wolf. Doing so would be desirable because it would enable a better understanding of the major features (Sherif, page 1, section: Abstract).
However, Milind, Wolf, and Sherif do not expressly teach wherein simulate, using a quantum simulator, substantially all possible parameter and feature combinations to stabilize the impacted features when evaluated alongside the new set of guidelines; identify a set of changes based on the simulation; grade, at a second quantum simulator; identify, at a third quantum simulator, a degree of change.
In the same field of endeavor, Ulyanov teaches wherein simulate, using a quantum simulator, substantially all possible parameter and feature combinations to stabilize the impacted features when evaluated alongside the new set of guidelines ([0050] a self-organizing controller can include in knowledge bases of candidate controllers and a quantum inference unit; [0051] processing the n sets of controller gains by using quantum computing operations; the quantum state of a controller gain is a superposition of a real controller state and a virtual controller state (i.e., all candidate parameter/ gain combinations are represented simultaneously); [0052] the quantum inference unit can combine features of the each independent knowledge base; [0055] quantum computing techniques can be implemented on a quantum computer or simulated, using classical efficient simulation methods of quantum algorithms); and based on the simulating, identifying resource consumption of each of the combinations at a quantum simulator within the quantum engine ([0141] the genetic algorithm unit can include a simulator configured to estimate the fitness of certain solutions; [0143] simulations, rather than real experiments, can be convenient for running a large number of experiments; [0144] the fitness requirements of an algorithm executed by the genetic algorithm unit evaluated on the fitness function unit by the following equation: ; [0145] simulated solutions evaluated against restrictions on control object output parameters such that a control force and generalized entropy are minimum (i.e. resource consumption)); identify a set of changes based on the simulation ([0141] the genetic algorithm unit can include a simulator configured to estimate the fitness of certain solutions; [0143] simulations… for running a large number of experiments; [0144] the fitness requirements of an algorithm executed by the genetic algorithm unit evaluated on the fitness function unit by the following equation: ; [0145] simulated solutions evaluated against restrictions on control object output parameters such that a control force and generalized entropy are minimum; [0062] when the information risk increment is large, it can be desirable to change the control structure and/or controller gains; where the information risk increment is small, it can be desirable to utilize the controller); grade, at a second quantum simulator ([0141] the genetic algorithm unit can include a simulator configured to estimate the fitness of certain solutions; [0143] simulations… for running a large number of experiments); identify, at a third quantum simulator, a degree of change ([0141] the genetic algorithm unit can include a simulator configured to estimate the fitness of certain solutions; [0143] simulations… for running a large number of experiments; [0144] the fitness requirements of an algorithm executed by the genetic algorithm unit evaluated on the fitness function unit by the following equation: ; [0145] simulated solutions evaluated against restrictions on control object output parameters such that a control force and generalized entropy are minimum; [0062] when the information risk increment is large, it can be desirable to change the control structure and/or controller gains; where the information risk increment is small, it can be desirable to utilize the controller).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to have incorporated wherein simulate, using a quantum simulator, substantially all possible parameter and feature combinations to stabilize the impacted features when evaluated alongside the new set of guidelines; identify a set of changes based on the simulation; grade, at a second quantum simulator; identify, at a third quantum simulator, a degree of change, as suggested in Ulyanov into Milind, Wolf, and Sherif. Doing so would be desirable because simulations, rather than real experiments, can be convenient for running a large number of experiments (Ulyanov [0143]) and that the quantum inference unit can combine features of the each independent knowledge base and/or generate a robust controller gain for a situation not addressed by any individual knowledge base by extracting information from the collection of knowledge bases (Ulyanov [0052]).
Claims 14-17 recite similar limitations as claims 2-5 above and are likewise rejected.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Applicant is required under 37 CFR § 1.111(c) to consider these references fully when responding to this action.
Adam et al. (US 2022/0357954 A1) teaches: A computing application can be subject to one or more compliance standards, where a compliance standard can be a rule or regulation that governs the computing application; the compliance component 116 can iterate through each node in the dependency topology; for each node, the compliance component 116 can electronically compare the configuration parameters assigned to that node with the set of compliance standards (e.g., with a subset of the set of compliance standards, which subset can specifically apply to the node under consideration); based on such comparison, the compliance component 116 can label the node under consideration as either individually compliant or individually non-compliant (see [0002], [0073]).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to REJI KARTHOLY whose telephone number is (571)272-3432. The examiner can normally be reached on Monday - Thursday from 7:30 am to 3:30 pm.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Jennifer Welch, can be reached at telephone number 571-272-7212. 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 Patent Center. Status information for published applications may be obtained from Patent Center. Status information for unpublished applications is available through Patent Center for authorized users only. Should you have questions about access to Patent Center, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
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) Form at https://www.uspto.gov/patents/uspto-automated- interview-request-air-form.
/REJI KARTHOLY/Primary Examiner, Art Unit 2143
Prosecution Insights