DETAILED ACTION
Notice of Pre-AIA or AIA Status
1. 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
2. A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 06/17/2026 has been entered.
3. Claims 1, 3-5, and 7-9 are currently pending and are rejected for the reasons set forth below.
Claim Rejections - 35 USC § 101
4. 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.
5. Claims 1, 3-5, and 7-9 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., an abstract idea) without significantly more.
6. Analysis:
Step 1: Statutory Category?: (is the claim(s) directed to a process, machine, manufacture or composition of matter?) - YES: In the instant case, claims 1, 3-5, and 7-9 are directed to computerized-method (i.e., process).
Regarding independent claim 1:
Step 2A - Prong 1: Judicial Exception Recited?: (is the claim(s) recited a judicial exception (an abstract idea enumerated in the 2019 PEG, a law of nature, or a natural phenomenon) – YES: Independent claim 1 recites the at least following limitations of “… in … transactions which are fraud-labeled transactions and legit-labeled transactions, … : (i) retrieving … a dataset of fraud-labeled transactions … on the dataset of fraud-labeled transactions, to mark transactions as 'similar' or 'novel'; and (ii) … a dataset of legit-labeled transactions … on the dataset of legit-labeled transactions, to mark transactions as 'similar' or 'novel', wherein … to mark a transaction as ‘similar’ when a pattern of the transaction is similar to transactions provided in the dataset of fraud-labeled transactions during training, wherein … to mark a transaction as ‘novel’ when a pattern of the transaction is not similar to transactions provided in the dataset of legit-labeled transactions during training, wherein … classifying transactions as either similar or different to the provided dataset, and wherein … all transactions, by separating transactions inside and outside …; … identify fraud transactions in transactions which have been classified as legit transactions …, to mark transactions as 'legit' or 'fraud', send transactions classified as ‘legit’ transactions … to be processed and marked as ‘similar’ or as ‘novel’; in response to transactions classified as ‘legit’ transactions … being marked as ‘similar’ …, determining that no further investigation is required for the transactions marked as ‘similar’ …; send transactions marked as ‘novel’ … to be processed and marked as ‘similar’ or as ‘novel’; and identify transactions marked as ‘novel’ … as potentially unknown fraud transactions and transactions marked as ‘similar’ … as potentially missed fraud, for alert distribution.” These recited limitations of the claim, as drafted, under its broadest reasonable interpretation, fall within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas as they cover performance of the limitations in fundamental economic principles or practices (including mitigating risk for classifying financial transactions into fraud or legit transactions). Accordingly, the claim recites an abstract idea.
Step 2A - Prong 2: Integrated into a Practical Application?: (is the claim(s) recited additional elements that integrate the exception into a practical application of the exception) - NO: This judicial exception is not integrated into a practical application. In particular, independent claim 1 further to the abstract idea includes additional elements of “a classification Machine Learning (ML) model”, “a financial system”, “a system”, “one or more processors”, “a data store”, “a ML fraud model”, “a ML legit model”, “a trained legit model”, “a trained ML fraud model”, “an unsupervised algorithm”, “a one-class Support Vector Machine (SVM)”, and “a hypersphere”. However, the additional elements recite generic computer components such as a computer, computing devices, a server, and/or software programing that are recited a high-level of generality that merely perform, conduct, carry out, implement, and/or narrow the abstract idea itself. Accordingly, the additional elements evaluated individually and in combination do not integrate the abstract idea into a practical application because they comprise or include limitations that are not indicative of integration into a practical application such as adding the words "apply it" (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea -- See MPEP 2106.05(f). The claim is directed to an abstract idea.
Step 2B: Claim provides an Inventive Concept?: (is the claim(s) recited additional elements that amount to an inventive concept (aka “significantly more”) than the recited judicial exception) - NO: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of “a classification Machine Learning (ML) model”, “a financial system”, “a system”, “one or more processors”, “a data store”, “a ML fraud model”, “a ML legit model”, “a trained legit model”, “a trained ML fraud model”, “an unsupervised algorithm”, “a one-class Support Vector Machine (SVM)”, and “a hypersphere” evaluated individually and in combination do not amount to more than a recitation of the words "apply it" (or an equivalent) or are not more than mere instructions to implement an abstract idea or other exception on a computer, or are not more than merely using a computer as a tool to perform an abstract idea. Use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general-purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application or provide significantly more - See MPEP 2106.05(f)(2). None of the additional elements taken individually or when taken as an ordered combination amount to significantly more than the abstract idea. Accordingly, the claim is patent-ineligible.
Dependent claims 3-5 and 7-9 have been given the full two-part analysis, analyzing the additional limitations both individually and in combination. The dependent claims, when analyzed individually and in combination, are also held to be patent-ineligible under 35 U.S.C. 101.
Dependent claims 3 and 4: simply refine the abstract idea because they recite limitations (e.g., the computerized-method further comprising calculating a novelty-score for transactions that have been marked as 'novel' by the ML fraud model, and wherein a preconfigured number of transactions having highest novelty-score are transmitted to a user for investigation; the computerized-method further comprising calculating a similarity-score for transactions that have been marked as 'similar' by the ML fraud model, and wherein a preconfigured number of transactions having highest similarity- score are transmitted to a user for investigation), that fall under the category of organizing human activity as described above in independent claim 1. Additionally, merely stating that these process steps are performed by the ML fraud model amounts to no more than merely applying generic computer components (i.e., the ML fraud model) to implement the abstract idea on a computer. Thus, the dependent claims do not add any additional element or subject matter that provides a technological improvement (i.e., an integration into a practical application) that results in the claims being directed to patent eligible subject matter or include an element or feature that is significantly more than the recited abstract idea (i.e., a technological inventive concept under Step 2B).
Dependent claim 5: simply provides further definition to “the computerized environment” recited in independent claim 1. Simply stating that wherein the computerized environment is at least one of: test environment, production environment or staging environment does not add any additional element or subject matter that provides a technological improvement (i.e., an integration into a practical application, a play interface) that results in the claim being directed to patent eligible subject matter or include an element or feature that is significantly more than the recited abstract idea (i.e., a technological inventive concept under Step 2B).
Dependent claims 7 and 8: simply provide further definition to “the retrieved fraud-labeled transactions and the retrieved legit-labeled transactions” recited in dependent claim 1. Simply stating that wherein the retrieved fraud-labeled transactions are transactions from a preconfigured time; wherein the retrieved legit-labeled transactions are a sample retrieved randomly from the legit-labeled transactions in the data store do not add any additional element or subject matter that provides a technological improvement (i.e., an integration into a practical application, a play interface) that results in the claims being directed to patent eligible subject matter or include an element or feature that is significantly more than the recited abstract idea (i.e., a technological inventive concept under Step 2B).
Dependent claim 9: simply provides further definition to “marking transactions as 'similar’” recited in independent claim 1. Simply stating that wherein marking transactions as 'similar' indicates that a pattern of the transactions is similar to transactions provided during training and wherein transactions marked as 'novel' indicates that the pattern of the transactions is not similar to transactions provided during the training does not add any additional element or subject matter that provides a technological improvement (i.e., an integration into a practical application, a play interface) that results in the claim being directed to patent eligible subject matter or include an element or feature that is significantly more than the recited abstract idea (i.e., a technological inventive concept under Step 2B).
Response to Applicant’s Arguments
7. 35 U.S.C. §101 Rejections: Applicant’s arguments with respect to amended claims 1, 3-5, and 7-9 that are rejected under 35 U.S.C. 101 have been considered but they are not persuasive because the claimed invention is directed to a judicial exception (i.e., an abstract idea) without significantly more.
1. Applicant’s Argument: Step 2A, Prong One: From Applicant Arguments/Remarks, Applicant respectfully disagrees that Claim 1, as amended, merely recites a method of organizing human activity. Claim 1 is not directed to the general concept of mitigating financial risk or classifying transactions as fraud or legitimate. Rather, Claim 1 recites a specific computerized machine-learning architecture in which transactions already classified as “legit” by a classification ML model are processed by two separately trained unsupervised ML models in a particular order. The claim requires training an ML fraud model using fraud-labeled transactions and training an ML legit model using legit-labeled transactions. The claim further requires each model to mark transactions as “similar” or “novel” based on its own respective training dataset. The trained ML legit model first processes transactions that the classification ML model classified as “legit.” Transactions marked as “similar” by the trained ML legit model are determined as requiring no further investigation. Only transactions marked as “novel” by the trained ML legit model are routed to the trained ML fraud model, which then marks those transactions as “similar” or “novel” to distinguish potentially missed fraud from potentially unknown fraud. Thus, Claim 1 does not merely recite the result of identifying fraud. Claim 1 recites a particular ML-model arrangement and a particular routing sequence among differently trained models (See Applicant Arguments/Remarks Pages 1-2).
In response to Applicant’s arguments, Examiner respectfully disagrees and submits that amended independent claim 1 at issue recite limitations as drafted, under its broadest reasonable interpretation, fall within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas as they cover performance of the limitations in fundamental economic principles or practices (including mitigating risk for classifying financial transactions into fraud or legit transactions). In addition, Examiner respectfully submits that geometric topology, particularly in the context of a sphere, involves the study of manifolds and maps between them, focusing on the geometric structures of low-dimensional manifolds like spheres, therefore the amended limitations of independent claim 1 of “by separating transaction inside the hypersphere when marked as “similar” and outside of the hypersphere, when marked as “novel”” still fall within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas as they cover performance of the limitations in fundamental economic principles or practices (including mitigating risk for classifying financial transactions into fraud or legit transactions). See details of Claim Rejections - 35 USC § 101 of claims 1, 3-5, and 7-9 in the section above.
2. Applicant’s Argument: Step 2A, Prong Two: From Applicant Arguments/Remarks, Applicant asserts that even if claim 1 were considered to recite an abstract idea, the claim integrates any alleged abstract idea into a practical application. Under MPEP §2106.04(d), a claim integrates a judicial exception into a practical application when the claim applies, relies on, or uses the judicial exception in a manner that imposes a meaningful limit on the judicial exception. Here, the claim does not merely instruct a generic computer to classify financial transactions. Instead, the claim recites a specific architecture that uses: 1. a classification ML model trained to mark transactions as “legit” or “fraud”; 2. a separately trained ML legit model trained on legit-labeled transactions to mark transactions as “similar” or “novel” relative to legit training data; 3. a separately trained ML fraud model trained on fraud-labeled transactions to mark transactions as “similar” or “novel” relative to fraud training data; 4. a one-class SVM hypersphere-based novelty detection mechanism for the legit and fraud models; and 5. a specific routing path in which transactions classified as “legit” by the classification ML model are first processed by the trained ML legit model, with only transactions marked as “novel” by the trained ML legit model being sent to the trained ML fraud model … Accordingly, the claim applies the trained ML models in a particular technological process for processing model outputs and routing transaction data between distinct trained models. The claim therefore imposes meaningful limits on any alleged abstract idea and integrates the alleged abstract idea into a practical application. The Claim Is Not Merely a Generic Computer Implementation. The Office Action characterizes the claim elements as generic computer components. Applicant respectfully disagrees. The eligibility analysis should consider the claim as a whole, including the ordered combination of limitations. Claim 1 does not merely recite a processor, data store, and generic ML model. Rather, claim 1 recites an arrangement in which: the ML fraud model and ML legit model are separately trained on different labeled datasets; …This is not a generic instruction to “apply” fraud classification using a computer. The claimed process changes how transaction data classified as “legit” by a classification ML model is further processed by a downstream ML architecture. The claimed architecture also produces specific downstream categories—potentially unknown fraud and potentially missed fraud—based on the output of the trained ML fraud model after filtering by the trained ML legit model. The Office Action does not address this ordered ML-model routing architecture as a whole. Instead, the Office Action treats the individual components in isolation and concludes that they are generic. Applicant respectfully submits that this approach overlooks the specific claimed interaction among the classification ML model, the trained ML legit model, and the trained ML fraud model (See Applicant Arguments/Remarks Pages 2-4).
In response to Applicant’s arguments, Examiner respectfully disagrees and submits that the current application does not improve the technical field of fraud detection by machine learning models and the amended independent claim 1 does not integrate the judicial exception into a practical application because the amended independent claim 1 further to the abstract idea includes additional elements of “a classification Machine Learning (ML) model”, “a financial system”, “a system”, “one or more processors”, “a data store”, “a ML fraud model”, “a ML legit model”, “a trained legit model”, “a trained ML fraud model”, “an unsupervised algorithm”, “a one-class Support Vector Machine (SVM)”, and “a hypersphere”. However, the additional elements recite generic computer components such as a computer, computing devices, a server, and/or software programing that are recited a high-level of generality that merely perform, conduct, carry out, implement, and/or narrow the abstract idea itself. Accordingly, the additional elements evaluated individually and in combination do not integrate the abstract idea into a practical application because they comprise or include limitations that are not indicative of integration into a practical application such as adding the words "apply it" (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, or merely uses a computer as a tool to perform an abstract idea -- See MPEP 2106.05(f). See details of Claim Rejections - 35 USC § 101 in the section above.
3. Applicant’s Argument: Step 2B: From Applicant Arguments/Remarks, Applicant asserts that For at least the reasons discussed above, Claim 1 recites significantly more than any alleged judicial exception. The ordered combination of claim limitations is not merely a generic computer implementation of fraud classification. Claim 1 recites a specific ML architecture that first applies a trained ML legit model to transactions classified as “legit” by a classification ML model, determines that transactions marked as “similar” by the trained ML legit model require no further investigation, routes transactions marked as “novel” by the trained ML legit model to a trained ML fraud model, and then identifies the transactions as potentially unknown fraud or potentially missed fraud based on whether the trained ML fraud model marks the transactions as “novel” or “similar.” The claim therefore includes an inventive concept in the ordered combination of the separately trained ML models, the hypersphere-based novelty/similarity mechanism, and the specific routing of transactions between the trained ML legit model and the trained ML fraud model. For at least the foregoing reasons, Applicant respectfully submits that claims 1, 3–5, and 7–9 are patent eligible under 35 U.S.C. §101. Applicant respectfully requests withdrawal of the §101 rejection (See Applicant Arguments/Remarks Page 4).
In response to Applicant’s arguments, Examiner respectfully disagrees and submits that amended independent claim 1 does not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of “a classification Machine Learning (ML) model”, “a financial system”, “a system”, “one or more processors”, “a data store”, “a ML fraud model”, “a ML legit model”, “a trained legit model”, “a trained ML fraud model”, “an unsupervised algorithm”, “a one-class Support Vector Machine (SVM)”, and “a hypersphere” evaluated individually and in combination do not amount to more than a recitation of the words "apply it" (or an equivalent) or are not more than mere instructions to implement an abstract idea or other exception on a computer, or are not more than merely using a computer as a tool to perform an abstract idea. Use of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general-purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application or provide significantly more - See MPEP 2106.05(f)(2). None of the additional elements taken individually or when taken as an ordered combination amount to significantly more than the abstract idea. See details of Claim Rejections - 35 USC § 101 in the section above. Examiner also notes that dependent claims 3-5, and 7-9 are rejected under 35 USC § 101 since they depend from the rejected independent claim 1.
Relevant Prior Art
8. The prior art made of record and not relied upon are considered pertinent to Applicant’s disclosure. The following references are pertinent for disclosing various features relevant to the claimed invention, but they do not disclose all the claimed features, as explained below.
9. The best prior art of record, Kala et al. (U.S. Pub. No. 2021/0081948), hereinafter, “Kala”, in view of Kramme et al. (U.S. Patent No. 10,825,028), hereinafter, “Kramme”, alone or in combination, neither discloses nor fairly suggests the instant application amended claim limitations of “operate training by: (i) retrieving from the data store a dataset of fraud-labeled transactions to train a ML fraud model on the dataset of fraud-labeled transactions, to mark transactions as ‘similar’ or ‘novel’; and(ii) retrieving from the data store a dataset of legit-labeled transactions to train a ML legit model on the dataset of legit-labeled transactions, to mark transactions as ‘similar’ or ‘novel’, wherein the ML fraud model is trained to mark a transaction as ‘similar’ when a pattern of the transaction is similar to transactions provided in the dataset of fraud-labeled transactions during training, wherein the ML legit model is trained to mark a transaction as ‘novel’ when a pattern of the transaction is not similar to transactions provided in the dataset of legit-labeled transactions during training, wherein the legit model and the fraud model are trained by an unsupervised algorithm that learns a decision function for classifying transactions as either similar or different to the provided dataset, and wherein the unsupervised algorithm is a one-class Support Vector Machine (SVM) that uses a hypersphere to encompass all transactions, by separating transactions inside the hypersphere when marked as ‘similar’ and outside of the hypersphere, when marked as ‘novel’, deploy a classification ML model, a trained ML fraud model and a trained ML legit model in a computerized environment to identify fraud transactions in transactions which have been classified as legit transactions by the classification ML model, wherein the trained classification ML model has been trained on a dataset of preconfigured transactions from the data store, to mark transactions as ‘legit’ or ‘fraud’.”
Conclusion
10. Any inquiry concerning this communication or earlier communications from the examiner should be directed to Liz Nguyen whose telephone number is (571) 272-5414. The examiner can normally be reached on Monday to Friday 8:00 A.M to 5:00 P.M.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Matthew Gart, can be reached on (571) 272-3955. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Center system (visit: https://patentcenter.uspto.gov). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call (800) 786-9199 (USA or CANADA) or (571) 272-1000.
/LIZ P NGUYEN/
Examiner, Art Unit 3696
/SCOTT S TROTTER/Primary Examiner, Art Unit 3696
Prosecution Insights