DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
This action is in reply to the communications filed on 04/06/2026.
Claims 1-20 are currently pending and have been examined.
This action is made Final.
Examiner Request
The Applicant is requested to indicate where in the specification there is support for U.S.C. §112(a) paragraph issues that can arise when claims are amended without support in the specification. The Examiner thanks the Applicant in advance.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea of processing credit decisions and recommended credit values and limits; without significantly more.
Claim 1 is directed to a method, which is one of the statutory categories of invention; Claim 10 is directed to a system, which is one of the statutory categories of invention; and Claim 19 is directed to a non-transitory computer-readable storage medium, which is one of the statutory categories of invention. (Step 1: YES).
Claim 1 is directed to a method for managing one or more credit risks of one or more first users, the ML-based computing method comprising: receiving, by one or more hardware processors, one or more inputs from one or more electronic devices associated with one or more second users, wherein the one or more inputs comprise information related to at least one of: one or more entities associated with the one or more first users; retrieving, by the one or more hardware processors, one or more data associated with the one or more first users from one or more databases, based on the one or more inputs received from the one or more electronic devices associated with the one or more second users, wherein the one or more data comprise at least one of: one or more credit agency data, one or more accounts receivables data, one or more financial metrics, and one or more entity data, associated with the one or more first users; training, by the one or more hardware processors, one or more machine learning models, by obtaining, by the one or more hardware processors, one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; selecting, by the one or more hardware processors, one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; segmenting, by the one or more hardware processors, the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; and training, by the one or more hardware processors, the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decisions, and wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model; determining, by the one or more hardware processors, the one or more credit risks of the one or more entities associated with the one or more first users based on preprocessed one or more data, by the trained one or more machine learning models; generating, by the one or more hardware processors, one or more credit decisions for the one or more entities associated with the one or more first users based on the determined one or more credit risks of the one or more entities associated with the one or more first users, by the trained one or more machine learning models, wherein the one or more credit decisions comprise at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; generating, by the one or more hardware processors, one or more confidence scores for each credit decision of the one or more credit decisions to classify the one or more credit decisions, based on a correlation between the one or more data and the one or more credit decisions and based on the trained one or more machine learning models, wherein the classification of the one or more credit decisions comprises at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; determining, by the one or more hardware processors, at least one of: one or more recommended credit values, one or more recommended first credit limits, one or more recommended second credit limits, and one or more recommended third credit limits, based on the classification of at least one of: the one or more first credit decisions, the one or more second credit decisions, and the one or more third credit decisions; providing, by the one or more hardware processors, one or more automated approvals for the one or more credit decisions, based on one or more second pre-configured rules and parameters; and providing, by the one or more hardware processors, an output of at least one of: the one or more credit decisions, the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, to the one or more second users on one or more user interfaces associated with the one or more electronic devices. These series of steps describe the abstract idea of processing credit decisions and recommended credit values and limits (with the exception of the italicized and bolded terms above), which is mitigating risk of potential losses arising from credit risks by determining and managing the credit risks of entities associated with users based on preprocessed data; therefore, corresponding to a fundamental economic principle or practice (including mitigating risk). Hence, a fundamental economic principle or practice (mitigating risk) is a Certain Methods of Organizing Human Activity. The abstract idea is also processing and analyzing credit risks of users to assist lenders and financial entities in making credit decisions, which is a commercial interaction. Therefore, a commercial interaction is also a Certain Methods of Organizing Human Activity. The system limitations, e.g., one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces, do not necessarily restrict the claim from reciting an abstract idea. Thus, claim 1 recites an abstract idea (Step 2A-Prong 1: YES).
This judicial exception is not integrated into a practical application because the additional elements of one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces, are no more than simply applying the abstract idea using generic computer elements. The additional elements listed above are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Merely invoking one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces is similar to invoking software and software components. The presence of a generic computer arrangement is nothing more than to implement the claimed invention (MPEP 2106.05(f)). Therefore, the recitations of additional elements do not meaningfully apply the abstract idea and hence do not integrate the abstract idea into a practical application. Thus, claim 1 does not integrate the abstract idea into a practical application (Step 2A-Prong 2: NO).
Claim 1 does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements of one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces, are recited at a high level of generality in that it results in no more than simply applying the abstract idea using generic computer elements. The additional elements when considered separately and as an ordered combination do not amount to add significantly more as these limitations provide nothing more than to simply apply the exception in a generic computer environment (Step 2B: NO). Thus, claim 1 is not patent eligible.
Dependent claims 2-9 are directed to a method that recites a series steps that describe the abstract idea of processing credit decisions and recommended credit values and limits. These series of steps describe the abstract idea of processing credit decisions and recommended credit values and limits, which is mitigating risk of potential losses arising from credit risks by determining and managing the credit risks of entities associated with users based on preprocessed data; therefore, corresponding to a fundamental economic principle or practice (including mitigating risk). Hence, a fundamental economic principle or practice (mitigating risk) is a Certain Methods of Organizing Human Activity. The abstract idea is also processing and analyzing credit risks of users to assist lenders and financial entities in making credit decisions, which is a commercial interaction. Therefore, a commercial interaction is also a Certain Methods of Organizing Human Activity. Additionally, dependent claims 3-5 recite a series of steps that describe the abstract idea of determining a recommended credit limit in a credit upgrade decision using statistical parameters to measure an entity's credit limit relative to a median credit limit of all entities across accounts at a specific point in time; where, the median and standard deviation provide insights into a central tendency and variability of the data. Therefore, corresponding to a mathematical calculation, relationship, and/or equation. Hence, a mathematical calculation, relationship, and/or equation is a Mathematical Concept. The system limitations, e.g., one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces, do not necessarily restrict the claim from reciting an abstract idea. Thus, claims 2-9 recite an abstract idea. The additional elements listed above are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Specifically, the additional elements, one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces, are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Merely invoking one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces is similar to invoking software and software components. The presence of a generic computer arrangement is nothing more than to implement the claimed invention (MPEP 2106.05(f)). Therefore, the recitations of additional elements do not meaningfully apply the abstract idea, and hence do not integrate the abstract idea into a practical application. Furthermore, the additional elements: one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces do not amount to add significantly more as these limitations provide nothing more than to simply apply the exception in a generic computer environment.
Claim 10 is directed to a machine learning based (ML-based) computing system for managing one or more credit risks of one or more first users, the ML-based computing system comprising: one or more hardware processors; a memory coupled to the one or more hardware processors, wherein the memory comprises a plurality of subsystems in form of programmable instructions executable by the one or more hardware processors, and wherein the plurality of subsystems comprises: an input receiving subsystem configured to receive one or more inputs from one or more electronic devices associated with one or more second users, wherein the one or more inputs comprise information related to at least one of: one or more entities associated with the one or more first users; a data retrieval subsystem configured to retrieve one or more data associated with the one or more first users from one or more databases, based on the one or more inputs received from the one or more electronic devices associated with the one or more second users, wherein the one or more data comprise at least one of: one or more credit agency data, one or more accounts receivables data, one or more financial metrics, and one or more entity data, associated with the one or more first users; a training subsystem configured to train one or more machine learning models, wherein in training the one or more machine learning models, the training subsystem is configured to: obtain one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; select one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; segment the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; and train the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decision, wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model; a credit risk determining subsystem configured to determine the one or more credit risks of the one or more entities associated with the one or more first users based on preprocessed one or more data, by the trained one or more machine learning models; a credit decision generation subsystem configured to generate one or more credit decisions for the one or more entities associated with the one or more first users based on the determined one or more credit risks of the one or more entities associated with the one or more first users, by the trained one or more machine learning models, wherein the one or more credit decisions comprise at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; a confidence score generation subsystem configured to generate one or more confidence scores for each credit decision of the one or more credit decisions to classify the one or more credit decisions, based on a correlation between the one or more data and the one or more credit decisions and based on the trained one or more machine learning models, wherein the classification of the one or more credit decisions comprises at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions: a credit limit determining subsystem configured to determine at least one of: one or more recommended credit values, one or more recommended first credit limits, one or more recommended second credit limits, and one or more recommended third credit limits, based on the classification of at least one of: the one or more first credit decisions, the one or more second credit decisions, and the one or more third credit decisions; an auto-approval subsystem configured to provide one or more automated approvals for the one or more credit decisions, based on one or more second pre-configured rules and parameters; and an output subsystem configured to provide an output of at least one of: the one or more credit decisions, the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, to one or more second users on one or more user interfaces associated with the one or more electronic devices. These series of steps describe the abstract idea of processing credit decisions and recommended credit values and limits (with the exception of the italicized and bolded terms above), which is mitigating risk of potential losses arising from credit risks by determining and managing the credit risks of entities associated with users based on preprocessed data; therefore, corresponding to a fundamental economic principle or practice (including mitigating risk). Hence, a fundamental economic principle or practice (mitigating risk) is a Certain Methods of Organizing Human Activity. The abstract idea is also processing and analyzing credit risks of users to assist lenders and financial entities in making credit decisions, which is a commercial interaction. Therefore, a commercial interaction is also a Certain Methods of Organizing Human Activity. The system limitations, e.g., a machine learning based (ML-based) computing system, one or more hardware processors, memory, plurality of subsystems, programmable instructions, input receiving subsystem, one or more electronic devices, data retrieval subsystem, one or more databases, training subsystem, credit risk determining subsystem, one or more machine learning models, credit decision generation subsystem, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, confidence score generation subsystem, credit limit determining subsystem, auto-approval subsystem, output subsystem, and one or more user interfaces, do not necessarily restrict the claim from reciting an abstract idea. Thus, claim 10 recites an abstract idea (Step 2A-Prong 1: YES).
This judicial exception is not integrated into a practical application because the additional elements of a machine learning based (ML-based) computing system, one or more hardware processors, memory, plurality of subsystems, programmable instructions, input receiving subsystem, one or more electronic devices, data retrieval subsystem, one or more databases, training subsystem, credit risk determining subsystem, one or more machine learning models, credit decision generation subsystem, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, confidence score generation subsystem, credit limit determining subsystem, auto-approval subsystem, output subsystem, and one or more user interfaces, are no more than simply applying the abstract idea using generic computer elements. The additional elements listed above are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Specifically, the additional elements, a machine learning based (ML-based) computing system, one or more hardware processors, memory, plurality of subsystems, programmable instructions, input receiving subsystem, one or more electronic devices, data retrieval subsystem, one or more databases, training subsystem, credit risk determining subsystem, one or more machine learning models, credit decision generation subsystem, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, confidence score generation subsystem, credit limit determining subsystem, auto-approval subsystem, output subsystem, and one or more user interfaces, are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Merely invoking a machine learning based (ML-based) computing system, one or more hardware processors, memory, plurality of subsystems, programmable instructions, input receiving subsystem, one or more electronic devices, data retrieval subsystem, one or more databases, training subsystem, credit risk determining subsystem, one or more machine learning models, credit decision generation subsystem, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, confidence score generation subsystem, credit limit determining subsystem, auto-approval subsystem, output subsystem, and one or more user interfaces, is similar to invoking software and software components. The presence of a generic computer arrangement is nothing more than to implement the claimed invention (MPEP 2106.05(f)). Therefore, the recitations of additional elements do not meaningfully apply the abstract idea and hence do not integrate the abstract idea into a practical application. Thus, claim 10 does not integrate the abstract idea into a practical application (Step 2A-Prong 2: NO).
Claim 10 does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements of a machine learning based (ML-based) computing system, one or more hardware processors, memory, plurality of subsystems, programmable instructions, input receiving subsystem, one or more electronic devices, data retrieval subsystem, one or more databases, training subsystem, credit risk determining subsystem, one or more machine learning models, credit decision generation subsystem, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, confidence score generation subsystem, credit limit determining subsystem, auto-approval subsystem, output subsystem, and one or more user interfaces, are recited at a high level of generality in that it results in no more than simply applying the abstract idea using generic computer elements. The additional elements when considered separately and as an ordered combination do not amount to add significantly more as these limitations provide nothing more than to simply apply the exception in a generic computer environment (Step 2B: NO). Thus, claim 10 is not patent eligible.
Dependent claims 11-18 are directed to a system that performs a series steps that describe the abstract idea of processing credit decisions and recommended credit values and limits. These series of steps describe the abstract idea of processing credit decisions and recommended credit values and limits, which is mitigating risk of potential losses arising from credit risks by determining and managing the credit risks of entities associated with users based on preprocessed data; therefore, corresponding to a fundamental economic principle or practice (including mitigating risk). Hence, a fundamental economic principle or practice (mitigating risk) is a Certain Methods of Organizing Human Activity. The abstract idea is also processing and analyzing credit risks of users to assist lenders and financial entities in making credit decisions, which is a commercial interaction. Therefore, a commercial interaction is also a Certain Methods of Organizing Human Activity. Additionally, dependent claims 12-14 recite a series of steps that describe the abstract idea of determining a recommended credit limit in a credit upgrade decision using statistical parameters to measure an entity's credit limit relative to a median credit limit of all entities across accounts at a specific point in time; where, the median and standard deviation provide insights into a central tendency and variability of the data. Therefore, corresponding to a mathematical calculation, relationship, and/or equation. Hence, a mathematical calculation, relationship, and/or equation is a Mathematical Concept. The system limitations, e.g., a machine learning based (ML-based) computing system, one or more hardware processors, memory, plurality of subsystems, programmable instructions, input receiving subsystem, one or more electronic devices, data retrieval subsystem, one or more databases, training subsystem, credit risk determining subsystem, one or more machine learning models, credit decision generation subsystem, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, confidence score generation subsystem, credit limit determining subsystem, auto-approval subsystem, output subsystem, and one or more user interfaces, do not necessarily restrict the claim from reciting an abstract idea. Thus, claims 11-18 recite an abstract idea. The additional elements listed above are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Specifically, the additional elements, a machine learning based (ML-based) computing system, one or more hardware processors, memory, plurality of subsystems, programmable instructions, input receiving subsystem, one or more electronic devices, data retrieval subsystem, one or more databases, training subsystem, credit risk determining subsystem, one or more machine learning models, credit decision generation subsystem, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, confidence score generation subsystem, credit limit determining subsystem, auto-approval subsystem, output subsystem, and one or more user interfaces, are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Merely invoking a machine learning based (ML-based) computing system, one or more hardware processors, memory, plurality of subsystems, programmable instructions, input receiving subsystem, one or more electronic devices, data retrieval subsystem, one or more databases, training subsystem, credit risk determining subsystem, one or more machine learning models, credit decision generation subsystem, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, confidence score generation subsystem, credit limit determining subsystem, auto-approval subsystem, output subsystem, and one or more user interfaces, is similar to invoking software and software components. The presence of a generic computer arrangement is nothing more than to implement the claimed invention (MPEP 2106.05(f)). Therefore, the recitations of additional elements do not meaningfully apply the abstract idea, and hence do not integrate the abstract idea into a practical application. Furthermore, the additional elements: a machine learning based (ML-based) computing system, one or more hardware processors, memory, plurality of subsystems, programmable instructions, input receiving subsystem, one or more electronic devices, data retrieval subsystem, one or more databases, training subsystem, credit risk determining subsystem, one or more machine learning models, credit decision generation subsystem, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, confidence score generation subsystem, credit limit determining subsystem, auto-approval subsystem, output subsystem, and one or more user interfaces, do not amount to add significantly more as these limitations provide nothing more than to simply apply the exception in a generic computer environment.
Claim 19 is directed to a non-transitory computer-readable storage medium having instructions stored therein that when executed by one or more hardware processors, cause the one or more hardware processors to execute operations of: receiving one or more inputs from one or more electronic devices associated with one or more second users, wherein the one or more inputs comprise information related to at least one of: one or more entities associated with the one or more first users; retrieving one or more data associated with the one or more first users from one or more databases, based on the one or more inputs received from the one or more electronic devices associated with the one or more second users, wherein the one or more data comprise at least one of: one or more credit agency data, one or more accounts receivables data, one or more financial metrics, and one or more entity data, associated with the one or more first users; training one or more machine learning models, by: obtaining one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; selecting one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; segmenting the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; and training the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decisions, wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model; determining the one or more credit risks of the one or more entities associated with the one or more first users based on preprocessed one or more data, by the trained one or more machine learning models; generating one or more credit decisions for the one or more entities associated with the one or more first users based on the determined one or more credit risks of the one or more entities associated with the one or more first users, by the trained one or more machine learning models, wherein the one or more credit decisions comprise at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; generating one or more confidence scores for each credit decision of the one or more credit decisions to classify the one or more credit decisions, based on a correlation between the one or more data and the one or more credit decisions and the trained one or more machine learning models, wherein the classification of the one or more credit decisions comprises at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; determining at least one of: one or more recommended credit values, one or more recommended first credit limits, one or more recommended second credit limits, and one or more recommended third credit limits, based on the classification of at least one of: the one or more first credit decisions, the one or more second credit decisions, and the one or more third credit decisions; providing one or more automated approvals for the one or more credit decisions with at least one of: the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, based on one or more second preconfigured rules and parameters; and providing an output of at least one of: the one or more credit decisions, the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, to one or more second users on one or more user interfaces associated with the one or more electronic devices. These series of steps describe the abstract idea of processing credit decisions and recommended credit values and limits (with the exception of the italicized and bolded terms above), which is mitigating risk of potential losses arising from credit risks by determining and managing the credit risks of entities associated with users based on preprocessed data; therefore, corresponding to a fundamental economic principle or practice (including mitigating risk). Hence, a fundamental economic principle or practice (mitigating risk) is a Certain Methods of Organizing Human Activity. The abstract idea is also processing and analyzing credit risks of users to assist lenders and financial entities in making credit decisions, which is a commercial interaction. Therefore, a commercial interaction is also a Certain Methods of Organizing Human Activity. The system limitations, e.g., a non-transitory computer-readable storage medium, one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, and one or more user interfaces, do not necessarily restrict the claim from reciting an abstract idea. Thus, claim 19 recites an abstract idea (Step 2A-Prong 1: YES).
This judicial exception is not integrated into a practical application because the additional elements of a non-transitory computer-readable storage medium, one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, and one or more user interfaces, are no more than simply applying the abstract idea using generic computer elements. The additional elements listed above are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Specifically, the additional elements, a non-transitory computer-readable storage medium, one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, and one or more user interfaces, are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Merely invoking a non-transitory computer-readable storage medium, one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, and one or more user interfaces, is similar to invoking software and software components. The presence of a generic computer arrangement is nothing more than to implement the claimed invention (MPEP 2106.05(f)). Therefore, the recitations of additional elements do not meaningfully apply the abstract idea and hence do not integrate the abstract idea into a practical application. Thus, claim 19 does not integrate the abstract idea into a practical application (Step 2A-Prong 2: NO).
Claim 19 does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements of a non-transitory computer-readable storage medium, one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, and one or more user interfaces, are recited at a high level of generality in that it results in no more than simply applying the abstract idea using generic computer elements. The additional elements when considered separately and as an ordered combination do not amount to add significantly more as these limitations provide nothing more than to simply apply the exception in a generic computer environment (Step 2B: NO). Thus, claim 19 is not patent eligible.
Dependent claim 20 is directed to a non-transitory computer-readable storage medium that performs a series steps that describe the abstract idea of processing credit decisions and recommended credit values and limits. These series of steps describe the abstract idea of processing credit decisions and recommended credit values and limits, which is mitigating risk of potential losses arising from credit risks by determining and managing the credit risks of entities associated with users based on preprocessed data; therefore, corresponding to a fundamental economic principle or practice (including mitigating risk). Hence, a fundamental economic principle or practice (mitigating risk) is a Certain Methods of Organizing Human Activity. The abstract idea is also processing and analyzing credit risks of users to assist lenders and financial entities in making credit decisions, which is a commercial interaction. Therefore, a commercial interaction is also a Certain Methods of Organizing Human Activity. The system limitations, e.g., a non-transitory computer-readable storage medium, one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, and one or more user interfaces, do not necessarily restrict the claim from reciting an abstract idea. Thus, claim 20 recites an abstract idea. The additional elements listed above are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Specifically, the additional elements, a non-transitory computer-readable storage medium, one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, and one or more user interfaces, are all recited at a high level of generality and under their broadest reasonable interpretation comprises a generic computing arrangement. Merely invoking a non-transitory computer-readable storage medium, one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, and one or more user interfaces, is similar to invoking software and software components. The presence of a generic computer arrangement is nothing more than to implement the claimed invention (MPEP 2106.05(f)). Therefore, the recitations of additional elements do not meaningfully apply the abstract idea, and hence do not integrate the abstract idea into a practical application. Furthermore, the additional elements: a non-transitory computer-readable storage medium, one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, trained one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, and one or more user interfaces, do not amount to add significantly more as these limitations provide nothing more than to simply apply the exception in a generic computer environment.
Dependent claims 2-9, 11-18, and 20 have further defined the abstract idea that is present in their respective independent claims: Claim 1, 10, and 19, and thus correspond to Certain Methods of Organizing Human Activity, and hence are abstract in nature for the reason presented above. The dependent claims 2-9, 11-18, and 20 do not include any additional elements that integrate the abstract idea into a practical application or are sufficient to amount to significantly more than the judicial exception when considered both individually and as an ordered combination. Therefore, dependent claims 2-9, 11-18, and 20 are directed to an abstract idea without significantly more.
Thus, claims 1-20 are not patent-eligible.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-2, 10-11, 18, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Relova (U.S. Patent Application Publication No. US 2025/0029178 A1 hereinafter “Relova”), in view of Speirs (U.S. Patent Application Publication No. US 2023/0274349 A1; hereinafter “Speirs”), and further in view of Zimmerman (U.S. Patent Application Publication No. US 2025/0200654 A1; hereinafter “Zimmerman”).
Regarding Claim 1:
Relova teaches:
A machine-learning based (ML-based) computing method for managing one or more credit risks of one or more first users, the ML-based computing method comprising: (Relova, a computer-based credit evaluation system that uses a machine learning-based credit risk model with an associated adverse action methodology to assess applicant credit profiles and identify adverse action factors for credit request denials (See, Para. 4, 6; Abstract));
receiving, by one or more hardware processors, one or more inputs from one or more electronic devices associated with one or more second users, wherein the one or more inputs comprise information related to at least one of: one or more entities associated with the one or more first users; (Relova, Credit evaluation system 108 receives a credit request of an applicant, e.g., a user of one of user devices 116 from FIG. 1 (170). In some examples, credit evaluation system 108 may retrieve the credit profile of the applicant from credit profiles 104 in database 102 from FIG. 1 . In other examples, credit evaluation system 108 may receive the credit profile of the applicant either directly from the one of user devices 116 of the applicant and store the applicant's credit profile in memory 126. (See, Para. 7, 17, 73; Abstract; Fig. 1));
retrieving, by the one or more hardware processors, one or more data associated with the one or more first users from one or more databases, based on the one or more inputs received from the one or more electronic devices associated with the one or more second users, wherein the one or more data comprise at least one of: one or more credit agency data, one or more accounts receivables data, one or more financial metrics, and one or more entity data, associated with the one or more first users; (Relova, database 102 may be managed by a credit reporting agency or credit bureau that captures, updates, and stores credit histories, e.g., credit profiles 104, on a majority of consumers. In other examples, database 102 may be managed by a specific lending institution and store credit profiles 104 for its customers and potential customers. (See, Para. 7, 17; Abstract; Fig. 1); Credit evaluation system 108 receives a credit request of an applicant, e.g., a user of one of user devices 116 from FIG. 1 (170). In some examples, credit evaluation system 108 may retrieve the credit profile of the applicant from credit profiles 104 in database 102 from FIG. 1 . In other examples, credit evaluation system 108 may receive the credit profile of the applicant either directly from the one of user devices 116 of the applicant and store the applicant's credit profile in memory 126. Credit risk scoring unit 134 uses credit risk model 132 to calculate a credit risk score for the applicant based on a credit profile of the applicant that includes applicant values for a plurality of characteristics assessed by credit risk model 132 (172). Credit risk model 132 may be a machine learning-based model that is trained based on values for a plurality of characteristics from a population of consumers, as described with respect to FIG. 4 . Credit risk scoring unit 134 may store the credit risk score for the applicant as one of applicant scores 128 in memory 126. (See, Para. 73));
determining, by the one or more hardware processors, the one or more credit risks of the one or more entities associated with the one or more first users based on preprocessed one or more data, by the trained one or more machine learning models; (Relova, Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision; Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. Reporting unit 112 may generate a report indicating whether the applicant's credit request is approved or denied and, if denied, further indicating one or more adverse action factors for the denial. Credit evaluation system 108 may then transmit the report to user device 116A of the applicant via network 114 and/or to a regulatory agency. (See, Para. 4, 7, 16-19, 73; Fig. 1));
generating, by the one or more hardware processors, one or more credit decisions for the one or more entities associated with the one or more first users based on the determined one or more credit risks of the one or more entities associated with the one or more first users, by the trained one or more machine learning models, wherein the one or more credit decisions comprise at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; (Relova, Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. Reporting unit 112 may generate a report indicating whether the applicant's credit request is approved or denied and, if denied, further indicating one or more adverse action factors for the denial. Credit evaluation system 108 may then transmit the report to user device 116A of the applicant via network 114 and/or to a regulatory agency. (See, Para. 4, 7, 16-19, 73: Fig. 1));
generating, by the one or more hardware processors, one or more confidence scores for each credit decision of the one or more credit decisions to classify the one or more credit decisions, based on a correlation between the one or more data and the one or more credit decisions and based on the trained one or more machine learning models, wherein the classification of the one or more credit decisions comprises at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; (Relova, Upon receipt of a credit request from an applicant via one of user devices 116, risk model unit 110 of credit evaluation system 108 retrieves the one of credit profiles 104 from database 102 for the applicant. For example, credit evaluation system 108 may send a query to database 102 for the particular one of credit profiles 104 associated with the applicant. In other examples, risk model unit 110 may receive the applicant's credit profile directly from the one of user devices 116 of the applicant. Risk model unit 110 uses a credit risk model to assess the applicant's credit profile according to multiple characteristics, which were used to build the credit risk model being applied. As some specific examples, the characteristics may include delinquency history, collections history, number of credit inquiries, bankcard balance, maximum credit amount, and the like. Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision. (See, Para.18-23; Fig.1));
determining, by the one or more hardware processors, at least one of: one or more recommended credit values, one or more recommended first credit limits, one or more recommended second credit limits, and one or more recommended third credit limits, based on the classification of at least one of: the one or more first credit decisions, the one or more second credit decisions, and the one or more third credit decisions; (Relova, Risk model unit 110 uses a credit risk model to assess the applicant's credit profile according to multiple characteristics, which were used to build the credit risk model being applied. As some specific examples, the characteristics may include delinquency history, collections history, number of credit inquiries, bankcard balance, maximum credit amount, and the like. Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision. (See, Para. 18-25; Fig.1; Abstract));
providing, by the one or more hardware processors, an output of at least one of: the one or more credit decisions, the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, to the one or more second users on one or more user interfaces associated with the one or more electronic devices. (Relova, Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. Reporting unit 112 may generate a report indicating whether the applicant's credit request is approved or denied and, if denied, further indicating one or more adverse action factors for the denial. Credit evaluation system 108 may then transmit the report to user device 116A of the applicant via network 114 and/or to a regulatory agency. (See, Para.16; Abstract; Fig. 1); credit evaluation system 108 utilizes interfaces 124 to wirelessly communicate with external systems, e.g., database 102 and/or user devices 116 from FIG. 1. (See, Para.34)).
Relova does not specifically teach training, by the one or more hardware processors, one or more machine learning models, by obtaining, by the one or more hardware processors, one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; selecting, by the one or more hardware processors, one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; segmenting, by the one or more hardware processors, the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; training, by the one or more hardware processors, the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decisions, and wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model; providing, by the one or more hardware processors, one or more automated approvals for the one or more credit decisions, based on one or more second pre-configured rules and parameters.
However, Speirs further teaches the following limitations:
training, by the one or more hardware processors, one or more machine learning models, by obtaining, by the one or more hardware processors, one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; (Speirs, a predictive machine learning model is used to predict an individual's probability of delinquency on a utility bill payment. The model is a predictive model that was trained, tested, and validated using a data set associated with account-level credit score and monthly payment performance between December 2009 and November 2016, obtained from a credit reporting agency (CRA), along with other financial and demographic data. Records with at least 24 months of consecutive utility payment performance data in the period (December 2014 to November 2016) were used, as one goal was to predict payment performance in the last 12 months of the data. (See, Para. 39-42, ; Abstract; Fig. 1-5));
selecting, by the one or more hardware processors, one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; (Speirs, Since it uses decision trees, the random forest algorithm is particularly appropriate for this application due to the fact that the dataset includes many variables (also known as features) of varying importance, on different scales. Decision trees are useful for finding the appropriate feature to split on, and for finding the value of that feature in order to minimize the cost function (See, Para. 65-71; Abstract; Fig. 1-5));
segmenting, by the one or more hardware processors, the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; and (Speirs, The random forest algorithm, another supervised machine learning technique, was also examined. The random forest technique involves separating the training and validation data set into multiple smaller datasets, or bags, forming decision trees with the smaller data sets, and using the many decision trees to classify the input parameters (See, Para. 65-71; Abstract; Fig. 1-5));
training, by the one or more hardware processors, the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decisions, and wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model. (Speirs, Since it uses decision trees, the random forest algorithm is particularly appropriate for this application due to the fact that the dataset includes many variables (also known as features) of varying importance, on different scales. Decision trees are useful for finding the appropriate feature to split on, and for finding the value of that feature in order to minimize the cost function (See, Para. 26, 29-34, 41, 65-71; Abstract; Fig. 1-5)).
It would have been obvious to one of ordinary skill in the art before the effective filing of the claimed invention to have modified Relova with the features of Speirs’ provide that a “predictive model may be a model that was trained, tested, and validated according to a machine learning technique. In certain embodiments, the machine learning technique comprises random forest classification. To generate the predictive model, in certain embodiments such as the embodiments described above, the number of individual utility service account holders is at least 800,000. In further embodiments the predictive model includes at least 5,000 features, each of the features being weighted according to the feature's contribution in the predictive model for predicting probability of delinquency on utility bill payment, wherein none of the twenty (20) highest-weighted features is a demographic variable. In some embodiments, none of the 50 highest-weighted features is a demographic variable, and in further embodiments, none of the 100 highest-weighted features is a demographic variable.” (Speirs, Para. 10).
Relova and Speirs do not specifically teach providing, by the one or more hardware processors, one or more automated approvals for the one or more credit decisions, based on one or more second pre-configured rules and parameters.
However, Zimmerman further teaches the following limitation:
providing, by the one or more hardware processors, one or more automated approvals for the one or more credit decisions, based on one or more second pre-configured rules and parameters; (Zimmerman, The credit issuer service 302 may also utilize decision engines that apply predefined rules and algorithms to evaluate credit applications, as described in more detail below. These engines process the credit scores and other relevant data to make automated decisions regarding credit approval, credit limits, and interest rates. The credit issuer service 302 may employ data analytics tools and techniques to gain insights from credit-related data. These technologies enable the identification of patterns, trends, and risk factors, facilitating more informed decision-making. (See, Para.63; Abstract; Fig. 3)).
It would have been obvious to one of ordinary skill in the art before the effective filing of the claimed invention to have modified Relova and Speirs with the features of Zimmerman’s system because “credit application processes typically require obtaining information from various external sources, which often come in different formats, making it challenging to analyze and process. The sources for this information can change frequently, leading to additional complexities. Traditional methods for authorizing user access have been manual, time-consuming, and prone to errors. Moreover, once credit issuers make a decision, there's generally no recourse for merchants or applicants. This disclosure introduces a novel system that addresses these challenges. It centralizes the acquisition of data from multiple external sources, normalizing this data, and converting it into a single risk metric. This allows merchants to assess an applicant's creditworthiness without seeing the comprehensive data, providing a more streamlined, adaptable, and efficient solution. Furthermore, this system permits merchants a say in the approval process, offering more flexibility in decision-making based on the risk metric, even if the applicant lacks ideal creditworthiness. This is a significant departure from the current technology, emphasizing automation and recourse.” (Zimmerman, Para. 3).
Regarding Claim 2:
Relova teaches:
further comprising retrieving, by the one or more hardware processors, one or more current credit limits and maintaining the one or more current credit limits, for the one or more entities associated with the one or more first users when the one or more third credit decisions are generated. (Relova, Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. ….. Upon receipt of a credit request from an applicant via one of user devices 116, risk model unit 110 of credit evaluation system 108 retrieves the one of credit profiles 104 from database 102 for the applicant……risk model unit 110 may receive the applicant's credit profile directly from the one of user devices 116 of the applicant. Risk model unit 110 uses a credit risk model to assess the applicant's credit profile according to multiple characteristics, which were used to build the credit risk model being applied. As some specific examples, the characteristics may include delinquency history, collections history, number of credit inquiries, bankcard balance, maximum credit amount, and the like. Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision. (See, Para. 16-23; Abstract; Fig. 1)).
Regarding Claim 10:
Relova teaches:
A machine learning based (ML-based) computing system for managing one or more credit risks of one or more first users, the ML-based computing system comprising: one or more hardware processors; a memory coupled to the one or more hardware processors, wherein the memory comprises a plurality of subsystems in form of programmable instructions executable by the one or more hardware processors, and wherein the plurality of subsystems comprises: (Relova, a computer-based credit evaluation system that uses a machine learning-based credit risk model with an associated adverse action methodology to assess applicant credit profiles and identify adverse action factors for credit request denials; a computing system comprising a memory, a credit risk model, and one or more processors in communication with the memory (See, Para. 4, 7; Abstract));
an input receiving subsystem configured to receive one or more inputs from one or more electronic devices associated with one or more second users, wherein the one or more inputs comprise information related to at least one of: one or more entities associated with the one or more first users; (Relova, Credit evaluation system 108 receives a credit request of an applicant, e.g., a user of one of user devices 116 from FIG. 1 (170). In some examples, credit evaluation system 108 may retrieve the credit profile of the applicant from credit profiles 104 in database 102 from FIG. 1 . In other examples, credit evaluation system 108 may receive the credit profile of the applicant either directly from the one of user devices 116 of the applicant and store the applicant's credit profile in memory 126. (See, Para. 7, 17, 73; Abstract; Fig. 1));
a data retrieval subsystem configured to retrieve one or more data associated with the one or more first users from one or more databases, based on the one or more inputs received from the one or more electronic devices associated with the one or more second users, wherein the one or more data comprise at least one of: one or more credit agency data, one or more accounts receivables data, one or more financial metrics, and one or more entity data, associated with the one or more first users; (Relova, database 102 may be managed by a credit reporting agency or credit bureau that captures, updates, and stores credit histories, e.g., credit profiles 104, on a majority of consumers. In other examples, database 102 may be managed by a specific lending institution and store credit profiles 104 for its customers and potential customers. (See, Para. 7, 17; Abstract; Fig. 1); Credit evaluation system 108 receives a credit request of an applicant, e.g., a user of one of user devices 116 from FIG. 1 (170). In some examples, credit evaluation system 108 may retrieve the credit profile of the applicant from credit profiles 104 in database 102 from FIG. 1 . In other examples, credit evaluation system 108 may receive the credit profile of the applicant either directly from the one of user devices 116 of the applicant and store the applicant's credit profile in memory 126. Credit risk scoring unit 134 uses credit risk model 132 to calculate a credit risk score for the applicant based on a credit profile of the applicant that includes applicant values for a plurality of characteristics assessed by credit risk model 132 (172). Credit risk model 132 may be a machine learning-based model that is trained based on values for a plurality of characteristics from a population of consumers, as described with respect to FIG. 4 . Credit risk scoring unit 134 may store the credit risk score for the applicant as one of applicant scores 128 in memory 126. (See, Para. 73));
a credit risk determining subsystem configured to determine the one or more credit risks of the one or more entities associated with the one or more first users based on preprocessed one or more data, by the trained one or more machine learning models; (Relova, Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision; Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. Reporting unit 112 may generate a report indicating whether the applicant's credit request is approved or denied and, if denied, further indicating one or more adverse action factors for the denial. Credit evaluation system 108 may then transmit the report to user device 116A of the applicant via network 114 and/or to a regulatory agency. (See, Para. 4, 7, 16-19, 73; Fig. 1));
a credit decision generation subsystem configured to generate one or more credit decisions for the one or more entities associated with the one or more first users based on the determined one or more credit risks of the one or more entities associated with the one or more first users, by the trained one or more machine learning models, wherein the one or more credit decisions comprise at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; (Relova, Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. Reporting unit 112 may generate a report indicating whether the applicant's credit request is approved or denied and, if denied, further indicating one or more adverse action factors for the denial. Credit evaluation system 108 may then transmit the report to user device 116A of the applicant via network 114 and/or to a regulatory agency. (See, Para. 4, 7, 16-19, 73: Fig. 1));
a confidence score generation subsystem configured to generate one or more confidence scores for each credit decision of the one or more credit decisions to classify the one or more credit decisions, based on a correlation between the one or more data and the one or more credit decisions and based on the trained one or more machine learning models, wherein the classification of the one or more credit decisions comprises at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; (Relova, Upon receipt of a credit request from an applicant via one of user devices 116, risk model unit 110 of credit evaluation system 108 retrieves the one of credit profiles 104 from database 102 for the applicant. For example, credit evaluation system 108 may send a query to database 102 for the particular one of credit profiles 104 associated with the applicant. In other examples, risk model unit 110 may receive the applicant's credit profile directly from the one of user devices 116 of the applicant. Risk model unit 110 uses a credit risk model to assess the applicant's credit profile according to multiple characteristics, which were used to build the credit risk model being applied. As some specific examples, the characteristics may include delinquency history, collections history, number of credit inquiries, bankcard balance, maximum credit amount, and the like. Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision. (See, Para.18-23; Fig.1));
a credit limit determining subsystem configured to determine at least one of: one or more recommended credit values, one or more recommended first credit limits, one or more recommended second credit limits, and one or more recommended third credit limits, based on the classification of at least one of: the one or more first credit decisions, the one or more second credit decisions, and the one or more third credit decisions; (Relova, Risk model unit 110 uses a credit risk model to assess the applicant's credit profile according to multiple characteristics, which were used to build the credit risk model being applied. As some specific examples, the characteristics may include delinquency history, collections history, number of credit inquiries, bankcard balance, maximum credit amount, and the like. Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision. (See, Para. 18-25; Fig.1; Abstract));
an output subsystem configured to provide an output of at least one of: the one or more credit decisions, the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, to one or more second users on one or more user interfaces associated with the one or more electronic devices. (Relova, Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. Reporting unit 112 may generate a report indicating whether the applicant's credit request is approved or denied and, if denied, further indicating one or more adverse action factors for the denial. Credit evaluation system 108 may then transmit the report to user device 116A of the applicant via network 114 and/or to a regulatory agency. (See, Para.16; Abstract; Fig. 1); credit evaluation system 108 utilizes interfaces 124 to wirelessly communicate with external systems, e.g., database 102 and/or user devices 116 from FIG. 1. (See, Para.34)).
Relova does not specifically teach a training subsystem configured to train one or more machine learning models, wherein in training the one or more machine learning models, the training subsystem is configured to: obtain one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; select one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; segment the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; train the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decision, wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model; an auto-approval subsystem configured to provide one or more automated approvals for the one or more credit decisions, based on one or more second pre-configured rules and parameters.
However, Speirs further teaches the following limitations:
a training subsystem configured to train one or more machine learning models, wherein in training the one or more machine learning models, the training subsystem is configured to: obtain one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; (Speirs, a predictive machine learning model is used to predict an individual's probability of delinquency on a utility bill payment. The model is a predictive model that was trained, tested, and validated using a data set associated with account-level credit score and monthly payment performance between December 2009 and November 2016, obtained from a credit reporting agency (CRA), along with other financial and demographic data. Records with at least 24 months of consecutive utility payment performance data in the period (December 2014 to November 2016) were used, as one goal was to predict payment performance in the last 12 months of the data. (See, Para. 39-42, ; Abstract; Fig. 1-5));
select one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; (Speirs, Since it uses decision trees, the random forest algorithm is particularly appropriate for this application due to the fact that the dataset includes many variables (also known as features) of varying importance, on different scales. Decision trees are useful for finding the appropriate feature to split on, and for finding the value of that feature in order to minimize the cost function (See, Para. 65-71; Abstract; Fig. 1-5));
segment the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; and (Speirs, The random forest algorithm, another supervised machine learning technique, was also examined. The random forest technique involves separating the training and validation data set into multiple smaller datasets, or bags, forming decision trees with the smaller data sets, and using the many decision trees to classify the input parameters (See, Para. 65-71; Abstract; Fig. 1-5));
train the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decision, wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model; (Speirs, Since it uses decision trees, the random forest algorithm is particularly appropriate for this application due to the fact that the dataset includes many variables (also known as features) of varying importance, on different scales. Decision trees are useful for finding the appropriate feature to split on, and for finding the value of that feature in order to minimize the cost function (See, Para. 26, 29-34, 41, 65-71; Abstract; Fig.1-5)).
It would have been obvious to one of ordinary skill in the art before the effective filing of the claimed invention to have modified Relova with the features of Speirs’ provide that a “predictive model may be a model that was trained, tested, and validated according to a machine learning technique. In certain embodiments, the machine learning technique comprises random forest classification. To generate the predictive model, in certain embodiments such as the embodiments described above, the number of individual utility service account holders is at least 800,000. In further embodiments the predictive model includes at least 5,000 features, each of the features being weighted according to the feature's contribution in the predictive model for predicting probability of delinquency on utility bill payment, wherein none of the twenty (20) highest-weighted features is a demographic variable. In some embodiments, none of the 50 highest-weighted features is a demographic variable, and in further embodiments, none of the 100 highest-weighted features is a demographic variable.” (Speirs, Para. 10).
Relova and Speirs do not specifically teach providing, by the one or more hardware processors, one or more automated approvals for the one or more credit decisions, based on one or more second pre-configured rules and parameters.
However, Zimmerman further teaches the following limitation:
an auto-approval subsystem configured to provide one or more automated approvals for the one or more credit decisions, based on one or more second pre-configured rules and parameters; (Zimmerman, The credit issuer service 302 may also utilize decision engines that apply predefined rules and algorithms to evaluate credit applications, as described in more detail below. These engines process the credit scores and other relevant data to make automated decisions regarding credit approval, credit limits, and interest rates. The credit issuer service 302 may employ data analytics tools and techniques to gain insights from credit-related data. These technologies enable the identification of patterns, trends, and risk factors, facilitating more informed decision-making. (See, Para.63; Abstract; Fig. 3)).
It would have been obvious to one of ordinary skill in the art before the effective filing of the claimed invention to have modified Relova and Speirs with the features of Zimmerman’s system because “credit application processes typically require obtaining information from various external sources, which often come in different formats, making it challenging to analyze and process. The sources for this information can change frequently, leading to additional complexities. Traditional methods for authorizing user access have been manual, time-consuming, and prone to errors. Moreover, once credit issuers make a decision, there's generally no recourse for merchants or applicants. This disclosure introduces a novel system that addresses these challenges. It centralizes the acquisition of data from multiple external sources, normalizing this data, and converting it into a single risk metric. This allows merchants to assess an applicant's creditworthiness without seeing the comprehensive data, providing a more streamlined, adaptable, and efficient solution. Furthermore, this system permits merchants a say in the approval process, offering more flexibility in decision-making based on the risk metric, even if the applicant lacks ideal creditworthiness. This is a significant departure from the current technology, emphasizing automation and recourse.” (Zimmerman, Para. 3).
Regarding Claim 11:
Relova teaches:
wherein the credit limit determining subsystem is further configured to retrieve one or more current credit limits and to maintain the one or more current credit limits, for the one or more entities associated with the one or more first users when the one or more third credit decisions are generated. (Relova, Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. ….. Upon receipt of a credit request from an applicant via one of user devices 116, risk model unit 110 of credit evaluation system 108 retrieves the one of credit profiles 104 from database 102 for the applicant……risk model unit 110 may receive the applicant's credit profile directly from the one of user devices 116 of the applicant. Risk model unit 110 uses a credit risk model to assess the applicant's credit profile according to multiple characteristics, which were used to build the credit risk model being applied. As some specific examples, the characteristics may include delinquency history, collections history, number of credit inquiries, bankcard balance, maximum credit amount, and the like. Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision. (See, Para. 16-23; Abstract; Fig. 1)).
Regarding Claim 18:
Relova teaches:
wherein the training subsystem is further configured to re-train the one or more machine learning models over a plurality of time intervals based on one or more training data, wherein in re-training the one or more machine learning models over the plurality of time intervals, the training subsystem is configured to: (Relova, Training unit 130 may periodically (e.g., monthly, bi-monthly, yearly, or the like) re-train credit risk model 132 based on an updated set of training data. The updated set of training data may include values for additional characteristics to be included in a new version of the credit risk model. Additionally, or alternatively, the updated set of training data may include characteristic values for the applicants that were scored during the time since credit risk model 132 was last trained by training unit 130. For example, the credit risk scores for new applicants calculated by credit risk model 132 itself may be used as feedback to update training data 103 for a future re-training of credit risk model 132. (See, Para. 36-38, 42-43; Abstract; Fig. 1, 2));
receive the one or more training data corresponding to the one or more data associated with the one or more second users; (Relova, After credit risk model 132 is trained and anchor values 127 are determined, credit scoring unit 134 may apply credit risk model 132 to assess new data in the form of applicant values of the characteristics from an applicant's credit profile, e.g., one of credit profiles 104 in database 102 from FIG. 1 . More specifically, risk model unit 110 receives a credit request from a user device of an applicant, e.g., one of user devices 116 from FIG. 1 , via interfaces 124. In response to the credit request, credit risk scoring unit 134 may retrieve the one of credit profiles 104 for the applicant from database 102. In other examples, credit risk scoring unit 134 may receive the applicant's credit profile directly from the one of user devices 116 and store the applicant's credit profile in memory 126. (See, Para. 36-43; Abstract; Fig. 1, 2));
add the one or more training data with the one or more labelled datasets to generate one or more updated training datasets; (Relova, Training unit 130 may periodically (e.g., monthly, bi-monthly, yearly, or the like) re-train credit risk model 132 based on an updated set of training data. The updated set of training data may include values for additional characteristics to be included in a new version of the credit risk model. Additionally, or alternatively, the updated set of training data may include characteristic values for the applicants that were scored during the time since credit risk model 132 was last trained by training unit 130. For example, the credit risk scores for new applicants calculated by credit risk model 132 itself may be used as feedback to update training data 103 for a future re-training of credit risk model 132 (See, Para. 36-43; Abstract; Fig. 1, 2));
re-train the one or more machine learning models to correlate the one or more features associated with the one or more data, with the one or more historical credit decisions; and (Relova, Training unit 130 may periodically (e.g., monthly, bi-monthly, yearly, or the like) re-train credit risk model 132 based on an updated set of training data. …..the updated set of training data may include characteristic values for the applicants that were scored during the time since credit risk model 132 was last trained by training unit 130. For example, the credit risk scores for new applicants calculated by credit risk model 132 itself may be used as feedback to update training data 103 for a future re-training of credit risk model 132 ….. Anchor values 127 are static values associated with credit risk model 132 such that anchor values 127 may only be updated when credit risk model 132 is re-trained.….. FIG. 3A shows chart 150 including a set of credit profile characteristics and applicant values, anchor values (i.e., “best values”), definitions, and an average type (e.g., mean or mode) for the set of credit profile characteristics. The credit profile characteristics include values related to delinquency history, collections history, number of credit inquiries, bankcard balance, maximum credit amount, and the like. (See, Para. 36-43, 50; Fig. 1, 2, 3A-B));
execute the re-trained one or more machine learning models in the credit decision generation subsystem to generate the one or more credit decisions for the one or more entities associated with the one or more first users. (Relova, Credit scoring unit 134 applies credit risk model 132 to calculate a credit risk score for the applicant based on the applicant values of the characteristics from the applicant's credit profile. Credit scoring unit 134 may store the credit risk score for the applicant as one of applicant scores 128 in memory 126. Based on the applicant's credit risk score, risk model unit 110 may determine whether to approve or deny the applicant's credit request. For example, risk model unit 110 may approve the credit request if the applicant's credit risk score is greater than or equal to a credit approval threshold defined by the lending institution from which the credit is being requested. Conversely, risk model unit 110 may deny the credit request if the applicant's credit risk score is less than the credit approval threshold. (See, Para. 36-44, 50; Fig. 1, 2, 3A-B)).
Regarding Claim 19:
Relova teaches:
A non-transitory computer-readable storage medium having instructions stored therein that when executed by one or more hardware processors, cause the one or more hardware processors to execute operations of: (Relova, a computer readable medium comprising instructions that when executed cause one or more processors to (See, Para. 8; Abstract)).
receiving one or more inputs from one or more electronic devices associated with one or more second users, wherein the one or more inputs comprise information related to at least one of: one or more entities associated with the one or more first users; (Relova, Credit evaluation system 108 receives a credit request of an applicant, e.g., a user of one of user devices 116 from FIG. 1 (170). In some examples, credit evaluation system 108 may retrieve the credit profile of the applicant from credit profiles 104 in database 102 from FIG. 1 . In other examples, credit evaluation system 108 may receive the credit profile of the applicant either directly from the one of user devices 116 of the applicant and store the applicant's credit profile in memory 126. (See, Para. 7, 8, 17, 73; Abstract; Fig. 1));
retrieving one or more data associated with the one or more first users from one or more databases, based on the one or more inputs received from the one or more electronic devices associated with the one or more second users, wherein the one or more data comprise at least one of: one or more credit agency data, one or more accounts receivables data, one or more financial metrics, and one or more entity data, associated with the one or more first users; (Relova, database 102 may be managed by a credit reporting agency or credit bureau that captures, updates, and stores credit histories, e.g., credit profiles 104, on a majority of consumers. In other examples, database 102 may be managed by a specific lending institution and store credit profiles 104 for its customers and potential customers. (See, Para. 7, 17; Abstract; Fig. 1); Credit evaluation system 108 receives a credit request of an applicant, e.g., a user of one of user devices 116 from FIG. 1 (170). In some examples, credit evaluation system 108 may retrieve the credit profile of the applicant from credit profiles 104 in database 102 from FIG. 1 . In other examples, credit evaluation system 108 may receive the credit profile of the applicant either directly from the one of user devices 116 of the applicant and store the applicant's credit profile in memory 126. Credit risk scoring unit 134 uses credit risk model 132 to calculate a credit risk score for the applicant based on a credit profile of the applicant that includes applicant values for a plurality of characteristics assessed by credit risk model 132 (172). Credit risk model 132 may be a machine learning-based model that is trained based on values for a plurality of characteristics from a population of consumers, as described with respect to FIG. 4 . Credit risk scoring unit 134 may store the credit risk score for the applicant as one of applicant scores 128 in memory 126. (See, Para. 73));
determining the one or more credit risks of the one or more entities associated with the one or more first users based on preprocessed one or more data, by the trained one or more machine learning models; (Relova, Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision; Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. Reporting unit 112 may generate a report indicating whether the applicant's credit request is approved or denied and, if denied, further indicating one or more adverse action factors for the denial. Credit evaluation system 108 may then transmit the report to user device 116A of the applicant via network 114 and/or to a regulatory agency. (See, Para. 4, 7, 16-19, 73; Fig. 1));
generating one or more credit decisions for the one or more entities associated with the one or more first users based on the determined one or more credit risks of the one or more entities associated with the one or more first users, by the trained one or more machine learning models, wherein the one or more credit decisions comprise at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; (Relova, Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. Reporting unit 112 may generate a report indicating whether the applicant's credit request is approved or denied and, if denied, further indicating one or more adverse action factors for the denial. Credit evaluation system 108 may then transmit the report to user device 116A of the applicant via network 114 and/or to a regulatory agency. (See, Para. 4, 7, 16-19, 73: Fig. 1));
generating one or more confidence scores for each credit decision of the one or more credit decisions to classify the one or more credit decisions, based on a correlation between the one or more data and the one or more credit decisions and the trained one or more machine learning models, wherein the classification of the one or more credit decisions comprises at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; (Relova, Upon receipt of a credit request from an applicant via one of user devices 116, risk model unit 110 of credit evaluation system 108 retrieves the one of credit profiles 104 from database 102 for the applicant. For example, credit evaluation system 108 may send a query to database 102 for the particular one of credit profiles 104 associated with the applicant. In other examples, risk model unit 110 may receive the applicant's credit profile directly from the one of user devices 116 of the applicant. Risk model unit 110 uses a credit risk model to assess the applicant's credit profile according to multiple characteristics, which were used to build the credit risk model being applied. As some specific examples, the characteristics may include delinquency history, collections history, number of credit inquiries, bankcard balance, maximum credit amount, and the like. Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision. (See, Para.18-23; Fig.1));
determining at least one of: one or more recommended credit values, one or more recommended first credit limits, one or more recommended second credit limits, and one or more recommended third credit limits, based on the classification of at least one of: the one or more first credit decisions, the one or more second credit decisions, and the one or more third credit decisions; (Relova, Risk model unit 110 uses a credit risk model to assess the applicant's credit profile according to multiple characteristics, which were used to build the credit risk model being applied. As some specific examples, the characteristics may include delinquency history, collections history, number of credit inquiries, bankcard balance, maximum credit amount, and the like. Based on the credit risk model, risk model unit 110 determines an overall credit risk score for the applicant that is used to make the lending decision. (See, Para. 18-25; Fig.1; Abstract));
providing an output of at least one of: the one or more credit decisions, the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, to one or more second users on one or more user interfaces associated with the one or more electronic devices. (Relova, Credit evaluation system 108 may retrieve the applicant's credit profile from credit profiles 104 stored in database 102 and analyze the applicant's credit profile with respect to multiple characteristics. Reporting unit 112 may generate a report indicating whether the applicant's credit request is approved or denied and, if denied, further indicating one or more adverse action factors for the denial. Credit evaluation system 108 may then transmit the report to user device 116A of the applicant via network 114 and/or to a regulatory agency. (See, Para.16; Abstract; Fig. 1); credit evaluation system 108 utilizes interfaces 124 to wirelessly communicate with external systems, e.g., database 102 and/or user devices 116 from FIG. 1 . (See, Para.34)).
Relova does not specifically teach training one or more machine learning models, by: obtaining one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; selecting one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; segmenting the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; training the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decisions, wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model; and providing one or more automated approvals for the one or more credit decisions with at least one of: the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, based on one or more second preconfigured rules and parameters.
However, Speirs further teaches the following limitations:
training one or more machine learning models, by: obtaining one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; (Speirs, a predictive machine learning model is used to predict an individual's probability of delinquency on a utility bill payment. The model is a predictive model that was trained, tested, and validated using a data set associated with account-level credit score and monthly payment performance between December 2009 and November 2016, obtained from a credit reporting agency (CRA), along with other financial and demographic data. Records with at least 24 months of consecutive utility payment performance data in the period (December 2014 to November 2016) were used, as one goal was to predict payment performance in the last 12 months of the data. (See, Para. 39-42, ; Abstract; Fig. 1-5));
selecting one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; (Speirs, Since it uses decision trees, the random forest algorithm is particularly appropriate for this application due to the fact that the dataset includes many variables (also known as features) of varying importance, on different scales. Decision trees are useful for finding the appropriate feature to split on, and for finding the value of that feature in order to minimize the cost function (See, Para. 65-71; Abstract; Fig. 1-5));
segmenting the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; and (Speirs, The random forest algorithm, another supervised machine learning technique, was also examined. The random forest technique involves separating the training and validation data set into multiple smaller datasets, or bags, forming decision trees with the smaller data sets, and using the many decision trees to classify the input parameters (See, Para. 65-71; Abstract; Fig. 1-5));
training the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decisions, wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model; (Speirs, Since it uses decision trees, the random forest algorithm is particularly appropriate for this application due to the fact that the dataset includes many variables (also known as features) of varying importance, on different scales. Decision trees are useful for finding the appropriate feature to split on, and for finding the value of that feature in order to minimize the cost function (See, Para. 26, 29-34, 41, 65-71; Abstract; Fig.1-5)).
It would have been obvious to one of ordinary skill in the art before the effective filing of the claimed invention to have modified Relova with the features of Speirs’ provide that a “predictive model may be a model that was trained, tested, and validated according to a machine learning technique. In certain embodiments, the machine learning technique comprises random forest classification. To generate the predictive model, in certain embodiments such as the embodiments described above, the number of individual utility service account holders is at least 800,000. In further embodiments the predictive model includes at least 5,000 features, each of the features being weighted according to the feature's contribution in the predictive model for predicting probability of delinquency on utility bill payment, wherein none of the twenty (20) highest-weighted features is a demographic variable. In some embodiments, none of the 50 highest-weighted features is a demographic variable, and in further embodiments, none of the 100 highest-weighted features is a demographic variable.” (Speirs, Para. 10).
Relova and Speirs do not specifically teach providing one or more automated approvals for the one or more credit decisions with at least one of: the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, based on one or more second preconfigured rules and parameters.
However, Zimmerman further teaches the following limitation:
providing one or more automated approvals for the one or more credit decisions with at least one of: the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, based on one or more second preconfigured rules and parameters; (Zimmerman, The credit issuer service 302 may also utilize decision engines that apply predefined rules and algorithms to evaluate credit applications, as described in more detail below. These engines process the credit scores and other relevant data to make automated decisions regarding credit approval, credit limits, and interest rates. The credit issuer service 302 may employ data analytics tools and techniques to gain insights from credit-related data. These technologies enable the identification of patterns, trends, and risk factors, facilitating more informed decision-making. (See, Para.63; Abstract; Fig. 3)).
It would have been obvious to one of ordinary skill in the art before the effective filing of the claimed invention to have modified Relova and Speirs with the features of Zimmerman’s system because “credit application processes typically require obtaining information from various external sources, which often come in different formats, making it challenging to analyze and process. The sources for this information can change frequently, leading to additional complexities. Traditional methods for authorizing user access have been manual, time-consuming, and prone to errors. Moreover, once credit issuers make a decision, there's generally no recourse for merchants or applicants. This disclosure introduces a novel system that addresses these challenges. It centralizes the acquisition of data from multiple external sources, normalizing this data, and converting it into a single risk metric. This allows merchants to assess an applicant's creditworthiness without seeing the comprehensive data, providing a more streamlined, adaptable, and efficient solution. Furthermore, this system permits merchants a say in the approval process, offering more flexibility in decision-making based on the risk metric, even if the applicant lacks ideal creditworthiness. This is a significant departure from the current technology, emphasizing automation and recourse.” (Zimmerman, Para. 3).
Allowable Subject Matter
Dependent claims 3, 6, 12, 15, and 20, as well as dependent claims 4-5, 7-9, 13-14, and 16-17 due to their dependencies on claims 3, 6, 12, and 15, respectively, would be allowable if rewritten or amended to overcome the rejection(s) under 35 U.S.C. 101, set forth in this Office action. The following is a statement of reasons for the indication of allowable subject matter: Independently the claims are obvious however the claims as a whole are not obvious because the examiner would have to improperly use the claims as a road map to combine the individual obvious claims together. However, the claims still do not overcome the 35 U.S.C. 101 rejection.
Response to Arguments
Applicant's arguments filed on 04/06/2026 have been fully considered, but are not persuasive due to the following reasons:
With respect to the rejection of claims 1-20 under 35 U.S.C. 101, Applicant arguments are moot in view of the grounds of rejections presented above in this office action. The arguments are addressed to the extent they apply to the amended claims.
Applicant argues that “while Applicant acknowledges that the end result of the system facilitates a commercial interaction, the claims as amended are fundamentally directed to a specific technological process: the specialized configuration and training of a machine learning computing system using precise feature engineering techniques and specific model architectures (e.g., XGBoost, LightGBM). Therefore, Applicant requests that the 35 U.S.C. Q 101 rejection of claims as allegedly being directed toward an abstract idea be reconsidered and removed, as Step 2A Prong One: NO. ”
Examiner respectfully disagrees.
Under Step 2A: Prong 1, Examiner respectfully notes that claims 1, 10, and 19, as amended, are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea of processing credit decisions and recommended credit values and limits; without significantly more. The series of steps recited in claims1, 10, and 19, as amended, describe the abstract idea of processing credit decisions and recommended credit values and limits, which is mitigating risk of potential losses arising from credit risks by determining and managing the credit risks of entities associated with users based on preprocessed data; therefore, corresponding to a fundamental economic principle or practice (including mitigating risk). Hence, a fundamental economic principle or practice (mitigating risk) is a Certain Methods of Organizing Human Activity. The abstract idea is also processing and analyzing credit risks of users to assist lenders and financial entities in making credit decisions, which is a commercial interaction. Therefore, a commercial interaction is also a Certain Methods of Organizing Human Activity. Furthermore, the system limitations (claim 1), e.g., one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces do not necessarily restrict the claim from reciting an abstract idea.
Moreover, Examiner respectfully notes that the claims are first analyzed in the absence of technology to determine if it recites an abstract idea. The additional limitations of technology are then considered to determine if it restricts the claim from reciting an abstract idea. In this case, it is determined that the additional limitations of technology do not necessarily restrict claims 1, 10, and 19, as amended, from reciting an abstract idea. Furthermore, Examiner respectfully notes that the recited features in the limitations: “receiving, by one or more hardware processors, one or more inputs from one or more electronic devices associated with one or more second users, wherein the one or more inputs comprise information related to at least one of: one or more entities associated with the one or more first users; retrieving, by the one or more hardware processors, one or more data associated with the one or more first users from one or more databases, based on the one or more inputs received from the one or more electronic devices associated with the one or more second users, wherein the one or more data comprise at least one of: one or more credit agency data, one or more accounts receivables data, one or more financial metrics, and one or more entity data, associated with the one or more first users; training, by the one or more hardware processors, one or more machine learning models, by obtaining, by the one or more hardware processors, one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; selecting, by the one or more hardware processors, one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; segmenting, by the one or more hardware processors, the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; and training, by the one or more hardware processors, the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decisions, and wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model; determining, by the one or more hardware processors, the one or more credit risks of the one or more entities associated with the one or more first users based on preprocessed one or more data, by the trained one or more machine learning models; generating, by the one or more hardware processors, one or more credit decisions for the one or more entities associated with the one or more first users based on the determined one or more credit risks of the one or more entities associated with the one or more first users, by the trained one or more machine learning models, wherein the one or more credit decisions comprise at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; generating, by the one or more hardware processors, one or more confidence scores for each credit decision of the one or more credit decisions to classify the one or more credit decisions, based on a correlation between the one or more data and the one or more credit decisions and based on the trained one or more machine learning models, wherein the classification of the one or more credit decisions comprises at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; determining, by the one or more hardware processors, at least one of: one or more recommended credit values, one or more recommended first credit limits, one or more recommended second credit limits, and one or more recommended third credit limits, based on the classification of at least one of: the one or more first credit decisions, the one or more second credit decisions, and the one or more third credit decisions; providing, by the one or more hardware processors, one or more automated approvals for the one or more credit decisions, based on one or more second pre-configured rules and parameters; and providing, by the one or more hardware processors, an output of at least one of: the one or more credit decisions, the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, to the one or more second users on one or more user interfaces associated with the one or more electronic devices” are simply making use of a computer and the computer limitations do not necessarily restrict the claim from reciting an abstract idea as discussed above under Step 2A-Prong 1 of the 35 U.S.C. 101 rejection.
Hence, Examiner has also considered each and every arguments under Step 2A-Prong 1 and concludes that these arguments are not persuasive. For example, under Step 2A-Prong 1, Examiner considers each and every limitation to determine if the claim recites an abstract idea. In this case, it is determined that the claim recites an abstract idea and the additional limitations of a computer device does not necessarily restrict the claim from reciting an abstract idea. The recited steps, as amended, are abstract in nature as there are no technical/technology improvements as a result of these steps. Thus, the claim recites an abstract idea. Whether the claim integrates the abstract idea into a practical application by providing technical/technology improvements are considered under Step 2A-Prong 2.
Applicant argues that “even if the Examiner maintains that the claims recite a judicial exception under Step 2A, Prong 1, Applicant respectfully submits that the claims are fully eligible under Step 2A, Prong 2….. even if the claims are found to recite a judicial exception under Prong 1, they integrate that exception into a practical application….. Therefore, the claims do not simply use a computer as a generic tool, they recite a specific, technical improvement in how an ML-based computing system is trained and deployed, effectively integrating the alleged abstract idea into a practical application. Thus, the claims are integrated into a practical application and are eligible as Step 2A Prong Two: YES.”
Examiner respectfully disagrees.
Under Step 2A: Prong II, Examiner respectfully notes that there is no improved technology in simply receiving, inputting, retrieving, training (processing), obtaining, selecting, segmenting, correlating, determining, preprocessing, generating, classifying, providing, approving, outputting, and presenting (displaying) data (i.e., user input data, entity information, user data, credit agency data, account receivables data, financial metrics, entity data, labelled datasets, validation datasets, historical credit decisions data, credit risks data, credit limits data, credit decision approval data, recommended credit values, and etc.). The disclosed invention simply cannot be equated to improvement to technological practices or computers. There is no technical improvement at all. Instead, Applicant recites in claim 1: “receiving, by one or more hardware processors, one or more inputs from one or more electronic devices associated with one or more second users, wherein the one or more inputs comprise information related to at least one of: one or more entities associated with the one or more first users; retrieving, by the one or more hardware processors, one or more data associated with the one or more first users from one or more databases, based on the one or more inputs received from the one or more electronic devices associated with the one or more second users, wherein the one or more data comprise at least one of: one or more credit agency data, one or more accounts receivables data, one or more financial metrics, and one or more entity data, associated with the one or more first users; training, by the one or more hardware processors, one or more machine learning models, by obtaining, by the one or more hardware processors, one or more labelled datasets from the one or more databases, wherein the one or more labelled datasets comprise the one or more data; selecting, by the one or more hardware processors, one or more features associated with the one or more data for training the one or more machine learning models based on one or more feature engineering processes; segmenting, by the one or more hardware processors, the one or more labelled datasets into at least one of: one or more training datasets and one or more validation datasets; and training, by the one or more hardware processors, the one or more machine learning models to correlate the one or more features associated with the one or more data and one or more historical credit decisions, and wherein the one or more machine learning models comprise at least one of: a random forest model, an extreme gradient boosting (XGBoost) classifier model, a K-means clustering model, a light gradient-boosting machine (LightGBM) classifier model; determining, by the one or more hardware processors, the one or more credit risks of the one or more entities associated with the one or more first users based on preprocessed one or more data, by the trained one or more machine learning models; generating, by the one or more hardware processors, one or more credit decisions for the one or more entities associated with the one or more first users based on the determined one or more credit risks of the one or more entities associated with the one or more first users, by the trained one or more machine learning models, wherein the one or more credit decisions comprise at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; generating, by the one or more hardware processors, one or more confidence scores for each credit decision of the one or more credit decisions to classify the one or more credit decisions, based on a correlation between the one or more data and the one or more credit decisions and based on the trained one or more machine learning models, wherein the classification of the one or more credit decisions comprises at least one of: one or more first credit decisions, one or more second credit decisions, one or more third credit decisions; determining, by the one or more hardware processors, at least one of: one or more recommended credit values, one or more recommended first credit limits, one or more recommended second credit limits, and one or more recommended third credit limits, based on the classification of at least one of: the one or more first credit decisions, the one or more second credit decisions, and the one or more third credit decisions; providing, by the one or more hardware processors, one or more automated approvals for the one or more credit decisions, based on one or more second pre-configured rules and parameters; and providing, by the one or more hardware processors, an output of at least one of: the one or more credit decisions, the one or more recommended credit values, the one or more recommended first credit limits, and the one or more recommended second credit limits, to the one or more second users on one or more user interfaces associated with the one or more electronic devices.” The recited features in the limitations do not result in computer functionality or technical improvement. Examiner respectfully notes that Applicant is simply using a computer to input, process, and output data. The recited features in the limitations does not disclose a technical solution to technical problem, but simply a business solution. Specifically, the recited steps, as amended, are merely managing/processing data (MPEP 2106.05(d)(II)) and does not result in computer functionality or technical improvement. Thus, Applicant has simply provided a business method practice of processing data (user input data, entity information, user data, credit agency data, account receivables data, financial metrics, entity data, labelled datasets, validation datasets, historical credit decisions data, credit risks data, credit limits data, credit decision approval data, recommended credit values, and etc.), and no technical solution or improvement has been disclosed.
Moreover, there is no technology/technical improvement as a result of implementing the abstract idea. The recited features in the claims, as amended, simply amount to the abstract idea of processing credit decisions and recommended credit values and limits; and, there is no computer functionality improvement or technology improvement. The claim does not provide a technical solution to a technical problem. If there is an improvement, it is to the abstract idea and not to technology. Additionally, Examiner notes that it is important to keep in mind that an improvement in the judicial exception itself (e.g., recited fundamental economic principle or practice and/or commercial interaction) is not an improvement in technology (See, MPEP 2106.05(a)(II)).
Additionally, Claim 1, as amended, recites steps at a high level of generality. In addition, all uses of the recited judicial exceptions require such data gathering and outputting, and, as such, these limitations do not impose any meaningful limits on the claim. These limitations amount to necessary data gathering and output. See MPEP 2106.05. The claim simply makes use of a computer as a tool to apply the abstract idea without transforming the abstract idea into a patent eligible subject matter. Furthermore, these steps, as amended, are recited as being performed by one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces. The additional elements: one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces are recited at a high level of generality, and are used as a tool to perform the generic computer function of receiving, processing, and outputting data. See MPEP 2106.05(f). Amended claim 1 recites one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces, which are simply used to perform an abstract idea, as discussed above in Step 2A, Prong I of the 35 U.S.C. 101 rejection, such that it amounts to no more than mere instructions to apply the exception using a generic computer. See MPEP 2106.05(f). Specifically, the recitation of “one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces” in the limitations merely indicates a field of use or technological environment in which the judicial exception is performed. The claims, as amended, merely confines the use of the abstract idea to a particular technological environment; and thus fails to add an inventive concept to the claims. See MPEP 2106.05(h). Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application, and the claim is directed to the judicial exception. Hence, claim 1, 10, and 19, as amended, do not integrate the abstract idea into a practical application. Thus, these arguments are not persuasive.
Applicant argues that “the ordered combination of elements-specifically, segmenting the datasets, extracting features via recursive or boosted extraction processes, and training specific ensemble models (XGBoost/LightGBM) to dynamically correlate historical decisions and generate specific confidence scores-amounts to significantly more than simply applying an abstract idea on a generic computer. This specific combination of ML techniques is not a well-understood, routine, or conventional activity in the industry, and it fundamentally transforms the claim into a specific, patent- eligible technological invention. Accordingly, the amended claims possess an inventive concept that transforms any alleged abstract idea into a patent-eligible application as Step 2B: YES. For these reasons, the withdrawal of the 35 U.S.C. § 101 rejection is respectfully requested.”
Examiner respectfully disagrees.
Under Step 2B, Examiner respectfully notes that all of Applicant's arguments have been reviewed, and the inventive concept cannot be furnished by a judicial exception. The improvements argued are to the abstract idea and not to technology. The technical limitations are simply utilized as a tool to implement the abstract idea without adding significantly more. Thus, the claim is directed to an abstract idea, and hence these arguments are not persuasive. The presence of a computer does not make the claimed solution necessarily rooted in computer technology. Furthermore, Examiner notes that the courts have determined that processing data is well-understood, routine, and conventional functions of a computer when they are claimed in a merely generic manner (see MPEP 2106.05(d)(II)). Thus, the recited combination of steps in claims 1, 10, and 19 operate in a well-understood, routine, conventional and generic way. As noted above, the claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements of “one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces are recited at a high level of generality in that it results in no more than simply applying the abstract idea using generic computer elements. The additional elements when considered separately and as an ordered combination do not amount to add significantly more as these limitations provide nothing more than to simply apply the exception in a generic computer environment.
Applying the 2024 Guidance Update on Patent Subject Matter Eligibility here, and as explained with respect to Step 2A, Prong 2, the additional elements: one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces, are at best mere instructions to “apply” the abstract idea, which cannot provide an inventive concept. See MPEP 2106.05(f). The additional elements: one or more hardware processors, one or more electronic devices, one or more databases, one or more machine learning models, random forest model, extreme gradient boosting (XGBoost) classifier model, K-means clustering model, light gradient-boosting machine (LightGBM) classifier model, one or more trained machine learning models, and one or more user interfaces, were found to be insignificant extra-solution activity in Step 2A, Prong II, because they were determined to be insignificant limitations as necessary for data gathering, processing, and outputting. The evaluation of the insignificant extra-solution activity consideration takes into account whether or not the extra-solution activity is well understood, routine, and conventional in the field. See MPEP 2106.05(g). As discussed in Step 2A, Prong II above, the claims’ limitations are recited at a high level of generality. These elements simply amount to receiving and outputting data and are well-understood, routine, conventional activity. See MPEP 2106.05(d)(II). As discussed in Step 2A, Prong II above, the recitation of a computer/processor to perform recited limitations, as amended, amounts to no more than mere instructions to apply the exception using a generic computer component. Even when considered in combination, these additional elements represent mere instructions to implement an abstract idea or other exception on a computer and insignificant extra-solution activity, which do not provide an inventive concept. (Step 2B: NO).
Hence, Examiner respectfully declines Applicant’s request to withdraw the 35 U.S.C. 101 rejection of claims 1-20.
With respect to the rejection of claims 1-2, 10-11, 18, and 19 under 35 U.S.C. 103, Applicant arguments are moot in view of new grounds of rejections presented above in this office action. Hence, Examiner respectfully declines Applicant’s request to withdraw the 35 U.S.C. 103 rejection of Claims 1-2, 10-11, 18, and 19 .
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure is the following:
Vinay (U.S. Patent Pub. No. US-2018/0182029-A1) “Systems and methods for custom ranking objectives for machine learning models applicable to fraud and credit risk assessments”
Dalinina (U.S. Patent Pub. No. US-2020/0357060-A1) “Rules/model-based data processing system for intelligent default risk prediction”
Hubard (U.S. Patent Pub. No. US- 2022/0122171-A1) “Client server system for financial scoring with cash transactions”
Anasta (U.S. Patent Pub. No. US-2022/0383406-A1) “Account risk detection and account limitation generation using machine learning”
Bradford (U.S. Patent Pub. No. US-2023/0084370-A1) “Dynamically updating account access based on employment data”
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MOHAMMED H MUSTAFA whose telephone number is (571)270-7978. The examiner can normally be reached M-F 8:00 - 5:00.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Michael W Anderson can be reached on 571-270-0508. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/MOHAMMED H MUSTAFA/Examiner, Art Unit 3693
/BRUCE I EBERSMAN/Primary Examiner, Art Unit 3693
Prosecution Insights