DETAILED ACTION
Acknowledgements
This action is in response to Applicant’s filing on Jun. 25, 2026, and is made Final. This action is being examined by James H. Miller, who is in the eastern time zone (EST), and who can be reached by email at James.Miller1@uspto.gov or by telephone at (469) 295-9082.
Interviews
Interviews are “indispensable to advance the prosecution of a patent application.” MPEP § 713. Accordingly, the following Examiner’s guidance and suggested workflow maximizes this benefit to Applicant by: (1) avoiding back and forth telephone calls for scheduling, (2) permitting Examiner out-of-office notifications to the Applicant when emailing the agenda, and (3) permitting real-time document collaboration and screen sharing.
Interviews are available by telephone or, preferably, by video conferencing using the USPTO’s web-based collaboration platform. Applicants are strongly encouraged to schedule via the USPTO Automated Interview Request (AIR) portal at http://www.uspto.gov/interviewpractice. If an interview is needed more quickly than permitted by the AIR scheduling tool, note this in the AIR remarks for consideration. The Examiner routinely considers such urgent requests when practicable.
An agenda submitted when filing the AIR is strongly encouraged, because Examiners use agendas when determining whether to grant an interview. The AIR has character limits, so send the agenda contemporaneously to James.Miller1@uspto.gov and reference the AIR.
After-Final Interviews Requests are granted only at the Examiner’s discretion and only if disposal or clarification for appeal may be accomplished with only nominal further consideration. MPEP § 713.09. An advance agenda explaining how the interview advances prosecution—e.g., through targeted arguments, identified Examiner error, or proposed claim amendments—is strongly suggested.
For GRANTED requests, expect an email within two (2) business days confirming a date/time slot and collaboration tool access instructions. For DENIED requests, the record will include an explanation for the denial.
The examiner is generally available for interviews, Monday through Friday, 10:00 a.m. to 4:00 p.m. ET.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Status
The status of claims is as follows:
Claims 1, 4–17, 19, 20, 24, and 25 are now pending and examined with Claims 1, 12, and 17 in independent form.
Claims 1, 12, 17 are presently amended.
No Claims are presently cancelled or added.
Response to Amendment
Applicant's Amendment has been reviewed against Applicant’s Specification filed Jun. 30, 2023, [“Applicant’s Specification”] and accepted for examination.
Response to Arguments
35 U.S.C. § 101 Argument
Applicant argues the claims “do not merely recite "methods of organizing human activity," but instead recite several detailed steps and features that point to specific improvements in computing device capabilities.” Applicant’s Reply at 12. Applicant argues the separate first and second user predictive models, real time merchant transaction data, storage of manual override data as “labeled training data,” and retraining “to adjust predictive weights” define a specific ML implementation for “modeling asymmetric account relationships.” Applicant’s Reply at 12–13.
Examiner respectfully disagrees. The claims as a whole remain directed to the commercial practice of determining whether a primary account holder’s spending limit for a secondary account holder should be overridden, rejected, or manually approved. The first/second predictive models, real time transaction data, labeled override data, and retraining are used to automate that financial permission determination. Non-Final Act. 5, 6, 10–13. Further, the claims define the models according to the financial account data used and the desired result, i.e., authorizing future transactions with similar data, but do not recite a particular model architecture, particular feature extraction procedure, training algorithm, particular technical implementation of the update/feedback mechanism, or particular technical similarity-analysis technique. The Specification teaches that the predictive models determine whether to authorize a spending limit override and can use primary/secondary user account data, merchant specific information, transaction amount, preferences, and geographic area in making that decision. Spec. ¶¶ 33, 41, 46.
Applicant argues “the claims are patent eligible because they integrate any alleged judicial exception into a practical application” because they “specify how the machine learning models are retrained (e.g., by storing labeled training data and using it to adjust predictive weights) … [and are] a specific technical implementation of machine learning, not a mental process. Applicant’s Reply at 12–13. The claims reflect a technical improvement to machine learning similar to that of Ex Parte Desjardins. Id. “Like Desjardins, which solved the technical problem of catastrophic forgetting by specifying how the model adjusts parameters to learn new tasks while protecting knowledge about previous tasks, the present claims solve the technical problem of modeling asymmetric account relationships. As described in the Specification, the system implements predictive models for user roles-"a primary user predictive model based on the primary user account data" and "a secondary user predictive model based on the secondary user account data," each based on different predictive variables tailored to their respective roles in the account relationship,” citing Spec. ¶¶ 9, 41. Specifically, amended Claims 1, 12, and 17 recite "store the transaction data and the manual override instruction as labeled training data" and "retrain at least one of the first predictive model and the second predictive model using the labeled training data to adjust predictive weights, wherein the retrained first predictive model and second predictive model automatically authorize subsequent transactions having transaction data similar to the stored transaction data without requiring manual override from the first user." Like Desjardins, the present claims specify how the models are retrained.” Applicant’s Reply at 14. Thus, claims reflect the technical improvement by specifying that the machine learning models are trained and retrained using "past transactions and associated indications of whether a first user did or did not manually execute a spending limitation override for each past transaction."” Id.
Examiner respectfully disagrees. Examiners acknowledge that the amended Independent Claims recite an adaptive, computer-implemented feedback loop and not merely a fixed spending threshold. However, the recited feedback loop is used to automate the financial determination of whether a secondary account holder’s over limit transaction should be permitted. The claims do not identify a particular model architecture, particular feature extraction procedure, training algorithm, particular technical implementation of the update/feedback mechanism, or particular technical similarity-analysis technique. The Specification describes “a machine learning model 290 that may be used to implement the one or more predictive model systems … [and] configured to facilitate the transaction management system 110 to determine whether to authorize a spending limitation override and/or determine a recurring purchase based on transaction history.” Spec. ¶ 33. Thus, the predictive models use primary and secondary account data to determine whether to authorize spending limit override. Spec. ¶ 41.
Desjardins is distinguishable. There, the specification identified an ML specific technical problem (i.e., “catastrophic forgetting in continual learning systems” p. 7) and described an improvement in the training and operation of the machine learning model itself (i.e., "effectively learn new tasks in succession whilst protecting knowledge about previous tasks" p. 9). The claims in Desjardins reflected that improvement by requiring adjustment of parameters to optimize performance on the second task “while protecting performance of the machine learning model on the first machine learning task.” p. 9. The disclosed benefits there included “learn[ing] new tasks in succession whilst protecting knowledge about previous tasks,” and “allow[ing] artificial intelligence systems to ‘us[e] less of their storage capacity’ and enable[ ] ‘reduced system complexity.’” p. 9.
Here, the Specification identifies manual intervention and lagging indicators in a spending limit authorization workflow and it states that a manual input override may update predictive models so that a future transaction may be automatically authorized. Spec. ¶¶ 17, 66. The claims do not identify a corresponding technical deficiency in predictive model operation or failure mode, such as degradation in learned task performance, and the claims do not require a particular model update that addresses such a deficiency. Instead, the claims use a manual override as labeled training data so that the model can make a future financial authorization decision. Non-Final Act. at 6–9, 16–21.
Applicant argues the recitation of two models each addressing role-specific predictive variables is a “specific machine learning model design that addresses the technical problem of modeling asymmetric account relationships.” Applicant’s Reply at 12. Applicant further argues the “two-model architecture and labeled-data retraining mechanism recited by the amended claims enable the disclosed system to ‘improve the speed with which computers can determine recurring transactions and spending limitations and allows spending limitation overrides to be conducted in near real-time, unlike current methods which only use lagging indicators and/or require manual input from a primary account holder," citing Spec. ¶ 17. Applicant’s Reply at 15.
Examiner respectfully disagrees. The first and second models are defined according to financial account data, and their use is in deciding whether to authorize a spending limit override. The Specification describes “a machine learning model 290 that may be used to implement the one or more predictive model systems … [and] configured to facilitate the transaction management system 110 to determine whether to authorize a spending limitation override and/or determine a recurring purchase based on transaction history.” Spec. ¶ 33. Thus, the predictive models use primary and secondary account data to determine whether to authorize spending limit override. Spec. ¶ 41. The claims do not recite how maintaining first and second models technically improves model architecture or operation, rather than using different financial data to improve a financial determination.
The pending claims are more like Recentive Analytics v. Fox than Desjardins. In Recentive, the claims used machine learning in a commercial context to improve television scheduling and network map generation, but they did not claim an improvement in ML techniques themselves. The Federal Circuit noted that “the claimed methods are not rendered patent eligible by the fact that (using existing machine learning technology) they perform a task previously undertaken by humans with greater speed and efficiency than could previously be achieved.” The pending claims here similarly use predictive models, labels, retraining, and weight adjustment to automate the commercial decision whether a secondary user’s over limit transaction should be approved, rather than reciting a specific solution to an identified technical deficiency in predictive model training or operation, as in Desjardins.
Applicant argues the amended claims are “patent eligible because they recite an inventive concept that is significantly more than the alleged abstract idea. … These elements are significantly more than an abstract idea because they specify a two-model architecture with role- specific predictive variables that addresses the technical problem of modeling asymmetric account relationships and real-time technical integration with merchant systems.” Applicant’s Reply at 17–18. Applicant further argues these elements are not well understood routine and conventional. Applicant’s Reply at 17–18.
Examiner respectfully disagrees. The claims automate a commercial decision about spending limit overrides between a primary user and a secondary user. The recited feedback loop is used to automate the financial determination of whether a secondary account holder’s over limit transaction should be permitted. The claims do not identify a particular model architecture, particular feature extraction procedure, training algorithm, particular technical implementation of the update/feedback mechanism, or particular technical similarity-analysis technique. The claims merely using a predictive model to “improve the speed with which computers can determine recurring transactions and spending limitations and allows spending limitation overrides to be conducted in near real-time.” Spec. ¶ 17. The Specification describes generic processors, memory, networks, databases, merchant systems, user interfaces, and conventional computing and predictive model functionality only at a high level. The Specification does not disclose a technical arrangement that produces an unconventional result beyond automated financial authorization.
Claim Objections
Claim 25 is objected to because of the following informalities. Appropriate correction is required.
Claims 25: The status identified of Claim 25 indicates “(New)” but Claim 25 was previously presented. Please update the status identifier next round accordingly.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1, 4–17, 19, 20, 24, and 25 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claims 1, 4–17, 19, 20, 24, and 25: Applicant amended Independent Claims this round to recite: “wherein the retrained first predictive model and second predictive model automatically authorize subsequent transactions having transaction data similar to the stored transaction data without requiring manual override from the first user.” The term “similar to” is a relative term which renders the claim indefinite. The term “similar to” is not defined by the claim, the specification does not provide a standard for ascertaining the requisite degree, and one of ordinary skill in the art would not be reasonably apprised of the scope of the invention. The term “similar to” controls the scope of the automatic authorization function. See, Spec. ¶¶ 41, 46, 50.
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, 4–17, 19, 20, 24, and 25 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.
Analysis
Step 1: Claims 1, 4–17, 19, 20, 24, and 25 are directed to a statutory category. Claims 1, 4–11, 24, and 25 recite “a system” and are therefore, directed to the statutory category of “a machine.” Claims 12–16 also recite “a system” and are therefore, directed to the statutory category of “a machine.” Claims 17, 19, and 20 also recite “a system” and are therefore, directed to the statutory category of “a machine.”
Representative Claim
Claim 1 is representative [“Rep. Claim 1”] of the subject matter under examination and recites, in part, emphasis added by Examiner to identify limitations with normal font indicating the abstract idea exception, bold limitations indicating additional elements. Each limitation is identified by a letter for later use as a shorthand notation in referencing/describing each limitation. Portions of the claim use italics to identify intended use limitations1 and underline, as needed, in further describing the abstract idea exception:
[A] 1. A system comprising: one or more processors; and memory in communication with the one or more processors and storing instructions that, when executed by the one or more processors, are configured to cause the system to:
[B] train at least one of a first predictive model and a second predictive model using historical transaction data associated with past transactions and associated indications of whether a first user did or did not manually execute a spending limitation override for each past transaction;
[C] implement [use] the first predictive model based on first account data associated with a first user, the first predictive model based on one or more predictive variables selected from primary merchant locations, account spending, a repayment schedule, or combinations thereof;
[D] implement [use] the second predictive model based on second account data associated with a second user, the second predictive model based on one or more predictive variables selected from secondary merchant locations, account spending, spending limitations, or combinations thereof;
[E] receive a first user input from the first user, the first user input corresponding to a spending limitation associated with the second user, and
[F] receive in real-time during the transaction and prior to completion of the transaction, transaction data from a merchant system, the transaction data comprising a transaction amount and a merchant identifier;
[G] identify that a transaction associated with the second user exceeds the spending limitation based on the transaction amount and merchant identifier;
[H] automatically override the spending limitation when the transaction exceeds the spending limitation by less than a threshold; and
[I] when the transaction exceeds the spending limitation by the threshold or greater than the threshold: automatically reject the spending limitation override;
[J] in response to receiving a manual override instruction from the first user, store the transaction data and the manual override instruction as labeled training data; and
[K] retrain at least one of the first predictive model and the second predictive model using the labeled training data to adjust predictive weights,
[L] wherein the retrained first predictive model and second predictive model automatically authorize subsequent transactions having transaction data similar to the stored transaction data without requiring manual override from the first user.
Claims are directed to an abstract idea exception.
Step 2A, Prong One: Rep. Claim 1 recites “when the transaction exceeds the spending limitation by less than a threshold” “automatically override the spending limitation”2 (Limitation H); “when the transaction exceeds the spending limitation by the threshold or greater than the threshold: automatically reject the spending limitation override” (Limitation I), and “automatically authorize subsequent transactions having transaction data similar to the stored transaction data without requiring manual override from the first user” (Limitation K), which recites commercial or legal interactions under the organizing human activity exception because overriding or rejecting a spending limitation is a pre-sale or sale activity, MPEP § 2106.04(a)(2)(II)(B), or a fundamental economic principle/practice also under organizing human activity because said Limitations “describe concepts relating to the economy and commerce,” such as “mitigating risks” of overspending on, for example, a credit card. MPEP § 2106.04(a)(2)(II)(A). Rep. Claim 1 describes the commercial practice of determining whether a primary account holder’s spending limitations for a secondary account holder should be overridden, rejected, or manually approved. This activity recites a commercial practice and a fundamental economic practice because it describes permissions for a secondary user to make financial transactions using an account subject to primary user’s spending limitations. Limitations E, F, and G are the required to “automatically override” or “automatically reject” the spending limitation and thus, recite the same exception.
The predictive model features do not alter the focus of the claims. Rep. Claim 1 uses first and second user account data, historical transactions, indication of a manual override, transaction amount, merchant identifier, merchant locations, repayment schedule, account spending, and spending limitations to determine whether an over limit transaction should be permitted. The storage of the “transaction data and the manual override instruction as labeled training data” and retraining a model “to adjust predictive weights” is used to learn a primary user’s over limit transaction authorization preferences and automatically authorize later transactions having similar data. Thus, the claim uses predictive model tools to automate a financial determination and does not claim an improvement is predictive model technology, training, or operation itself.
Step 2A, Prong Two: Rep. Claim 1 does not contain additional elements that integrate the abstract idea exception into a practical application because the additional elements are mere instructions to apply the abstract idea exception. MPEP § 2106.05(f). The additional elements are: a system comprising one or more processors, and memory storing instructions; Limitation B (“train one or more machine learning models …”); Limitation C (“implement [use] the first predictive mode …”); Limitation D (“implement [use] the second predictive model …”); a merchant system; store the transaction data and the manual override instruction as labeled training data; Limitation K (“retrain …”).
Regarding the system comprising one or more processors, and memory storing instructions; Limitation B (“train one or more machine learning models …”); Limitation C (“implement [use] the first predictive mode …”); Limitation D (“implement [use] the second predictive model …”); a merchant system; store the transaction data and the manual override instruction as labeled training data; Limitation K (“retrain …”), Applicant’s Specification does not otherwise describe them or describes them using exemplary language as a general-purpose computer, as a part of a general-purpose computer, or as any known and exemplary (generic) computer component known in the prior art. Thus, Applicant takes the position that such hardware/software is so well known to those of ordinary skill in the art that no explanation is needed under 35 U.S.C. § 112(a). Lindemann Maschinenfabrik GMBH v. Am. Hoist & Derrick Co., 730 F.2d 1452, 1463 (Fed. Cir. 1984) (citing In re Meyers, 410 F.2d 420, 424 (CCPA 1969) (“[T]he specification need not disclose what is well known in the art”). E.g., Spec. ¶ 25 (“Network 150 may be of any suitable type”); ¶ 27 (“A peripheral interface, for example, may include the hardware, firmware and/or software that enable(s) communication with various peripheral devices”); ¶ 32 (“the transaction management system 110 may include the memory 230 that includes instructions to enable the processor 210 to execute one or more applications, such as server applications, network communication processes, and any other type of application or software known to be available on computer systems.”); ¶ 34 (“The memory 230 may also include any combination of one or more databases controlled by memory controller devices (e.g., server(s), etc.) or software”); ¶ 37 (“the transaction management system 110 may include any number of hardware and/or software applications that are executed to facilitate any of the operations”); ¶ 44 (“the secondary user is free to transact with any merchant with transactions that are not otherwise limited by the spending limitation”); ¶ 44 (“the spending limitation may be a combination of any of the above-described spending limitations”); ¶ 55 (These computer-executable program instructions may be loaded onto a general-purpose computer”); ¶ 85 (any known “mobile” or “portable computing” device); ¶¶ 26, 30, 31, Fig. 2 (known and generic (exemplary) processor, memory, and instructions); ¶ 29 (any “mobile network interface … known in the art.”); ¶ 31 (“The processor 210 may be one or more known processing devices”); ¶ 59 (“well-known methods, structures and techniques have not been shown in detail in order not to obscure an understanding of this description”) ¶ 17 (“Machine learning models are a unique computer technology that involves training the models to complete tasks, such as labeling, categorizing, and identifying recurring transactions.”). The Specification describes the generic processor, here, performing calculations (functions) that are programmed by software. Spec. ¶ 33. This describes a computer doing what it is designed to do—performing directions it is given to follow. Thus, the Specification describes generic computer components and conventional communication and storage arrangement. Spec. ¶¶ ¶¶ 25, 26, 27, 30, 31, 32, 33, 34, 37, 40, 41, 44, 55, 59, 85, Fig. 2.
The Specification further describes the claimed first/second prediction model (Limitations B, C, D, and K) may built with “software known to be available on computer systems.” Spec., ¶ 32. The computer system may be a “general purpose computer.” Id. ¶ 55. Applicant’s Specification discloses the model “may include a machine learning model 290 … to facilitate the transaction management system 110 to determine whether to authorize a spending limitation override and/or determine a recurring purchase.” Spec. ¶ 33. The Specification further describes producing primary and secondary user predictive models “by feeding the primary user account data and the secondary user account data into a machine learning model … [to] determine whether to allow a spending limit override.” Spec. ¶ 41. Rep. Claim 1 does not describe a particular predictive model architecture; a technical interaction between the first and second models; a specific feature extraction technique; a label format or labeling technique, a training objective; a training or weight update/feedback mechanism or technique; a technical similarity analysis technique, a performance constraint; a model accuracy improvement technique; a processing resource reduction; or defined latency reduction.
Limitations J and K recite a feedback implementation, and Examiner considers the claimed feedback loop in combination with the real time merchant system receipt (Limitation F) and later automatic authorization function (Limitation L) to the extent they are additional elements, arguendo. The Specification does not describe training of the models and “training” itself is recited in one location. Spec. ¶ 17. The Specification teaches that “[t]raditional payment systems required manual intervention from primary account holders of an account to authorize transactions which exceed spending limits set for secondary account holders of the account. … [and] use lagging indicators and/or require manual input from a primary account holder.” Spec. ¶ 17. The Specification further teaches that “the manual override provided by the primary account user will be used by the system to update the one or more predictive model systems so that in the future, such a transaction may be automatically authorized.” Spec. ¶ 66. Thus, the Specification identifies an adaptive financial authorization workflow whose stated benefit is reduced repeated manual intervention and more timely authorization. The Specification does not identify a technical deficiency in predictive model training, retraining, or operation, payment network communication, merchant system operation, database functionality, or computer architecture.
Limitation F, to the extent it is an additional element, arguendo, does not specify an improvement in merchant system data transmission, timing, routing, security, or processing, and merely invokes computers or other machinery in its ordinary capacity to receive, store, or transmit data. MPEP § 2106.05(f)(2). The Specification describes the merchant systems as a “general purpose computer.” Spec. ¶ 55.
The Specification asserts “examples of the present disclosure improve the speed with which computers can determine recurring transactions and spending limitations and allows spending limitation overrides to be conducted in near real-time, unlike current methods which only use lagging indicators and/or require manual input from a primary account holder [automating authorizations].” Spec. ¶ 17. Rep. Claim 1 does not recite determining recurring transactions and does not recite a particular technical mechanism that produces the asserted real-time result. Thus, even considered in combination, the additional elements apply predictive models, data storage, and transaction processing functions to the financial permission abstract idea exception and do not improve the computer, predictive models, merchant systems, the payment network, or database functionality. Rather, the asserted improvement in speed concerns “relying on a computer to perform routine tasks more quickly or more accurately is insufficient to render a claim patent eligible.” OIP Technologies, Inc. v. Amazon.com, Inc., 788 F.3d 1359, 1363 (Fed. Cir. 2015) (citing Alice Corp. Pty. Ltd. v. CLS Bank Intern., 134 S.Ct. 2347, 2359 (2014)); MPEP § 2106.05(f)(2).
The additional elements do not provide a practical application under Desjardins and are more analogous to the use of generally described machine learning in Recentive for the same reasons stated above in the Response to Arguments point heading, which are incorporated here as if set out in length. Rep. Claim 1 does not identify a comparable technical problem with predictive model operations/training or require a particular technical mechanism that corrects it and is distinguished from Desjardins.
Therefore, the claim as a whole, looking at the additional elements individually and in combination, are no more than mere instructions to apply the exception using generic computer components and is not a practical application. MPEP § 2106.05(f). The additional elements do not integrate the abstract idea exception into a practical application because they do not impose any meaningful limits on the abstract idea exception. Accordingly, Rep. Claim 1 is directed to an abstract idea.
Independent Claims 12 and 17 are not substantially different than Rep. Claim 1, recite the same abstract idea as Rep. Claim 11, and contain no additional elements not otherwise analyzed for Rep. Claim 1. Therefore, Independent Claims 12 and 17 are also directed to the same abstract idea.
The claims do not provide an inventive concept.
Rep. Claim 1 fails Step 2B because the claim as a whole, even when considering the additional elements individually and in combination, does not amount to significantly more than the abstract idea. MPEP § 2106.05. The additional elements (i.e., a system comprising one or more processors, and memory storing instructions; Limitation B (“train one or more machine learning models …”); Limitation C (“implement [use] the first predictive mode …”); Limitation D (“implement [use] the second predictive model …”); a merchant system; store the transaction data and the manual override instruction as labeled training data; Limitation K (“retrain …”), are each well-understood, routine, and conventional (“WRC”) computer components and functions in the relevant field, as evidenced by Applicant’s own disclosure[3]. Further, Applicant’s Specification discloses that these components operate in no particular order and are implemented using generic, off-the-shelf computing technology. Spec. ¶ 53 (steps/functions may be performed in any order or concurrently); Spec., ¶¶ 25, 26, 27, 30, 31, 32, 33, 34, 37, 40, 41, 44, 55, 59, 85, Fig. 2 (describing each component using exemplary language as generic or known computing equipment and networks).
A system comprising one or more processors, and memory storing instructions is WRC in the financial technology field. Spec. ¶¶ 25–38, 53–57.
A merchant system transmitting transaction and transaction data is WRC transaction processing communications. Spec. ¶¶ 21, 22, 23, 45.
Store the transaction data and the manual override instruction as labeled training data is a WRC data storage operation. The claim does not recite a particular storage structure, labeling structure or format, storage retrieval technique or improvement in a database. Spec. ¶¶ 30, 32, 34, 35.
Limitation C (“implement [use] the first predictive mode …”); Limitation D (“implement [use] the second predictive model …”) are recited at a high level of generality. Spec. ¶¶ 33, 41 (cites supra).
Limitation B (“train one or more machine learning models …”) and Limitation K (“retrain …”) are recited at a high level of generality and do not recite a particular model architecture, particular feature extraction procedure, training algorithm, particular technical implementation of the update/feedback mechanism, or particular technical similarity-analysis technique.
Applying such predictive models to financial transaction data to pre-authorize or score card transactions is conventional in the art, as evidenced by U.S. Pat. No. 6,430,539 (Aug. 6, 2002) (“Predictive Modeling of Consumer Financial Behavior”), which describes building predictive models from historical account/transaction data to predict future spending behavior, using features and labels derived from past transactions, and U.S. Pat. Pub. No. 2010/0280927 (Nov. 4, 2010) (“Pre-authorization of a Transaction Using Predictive Modeling”), which describes using predictive scoring models on transaction data to authorize payment card transactions in real time or near real time.
Using AI/ML/predictive model in payment transaction processing, including receiving transaction data from merchant systems and generating authorization decisions is conventional in the art, as evidenced by U.S. Pat. Pub. No. 2018/0183737 (“Processing Payment Transactions Using Artificial Intelligence Messaging Services”).
The Examiner also finds the functions of training, implementing, receiving, identifying, overriding (authorizing), rejecting (denying), storing, and retraining, and processing (e.g., performing mathematical operations on) data, described in Limitations A–L are all normal functions of a generic computer.
Additionally, NPL Leskovec further supports that the broad predictive model operations recited in Rep. Claim 1 were known. NPL Leskovec describes supervised machine learning using labeled feature vector/output pairs to learn a predictive function. NPL Leskovec at pp. 463–466, § 12.1.1 and Example 12.2. NPL Leskovec further teaches training classifiers by modifying features weights based on labeled training examples, including perceptron and Winnow weight-update techniques. NPL Leskovec, pp. 471–477, §§ 12.2.1, 12.2.3, Examples 12.4, 12.5. NPL Leskovec further describes online learning in which newly arriving training examples are used to modify an existing model, and nearest neighbor techniques that classify a future input based on stored examples and a similarity distance measure. NPL Leskovec at pp. 468–470, 483–485, 497–500, §§ 3.5, 12.1.4, 12.2.8, 12.4.1, 12.4.2. Thus, NPL Leskovec supports the use of labeled historical information, model weight adjustment, updating a model using new labeled information, and similarity/distance based classification were known techniques. Rep. Claim 1 does not recite a particular predictive model, weight update algorithm, similarity technique, or data structure that distinguishes the claimed training and retraining steps from these known techniques.
There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. Thus, taken alone, the additional elements do not amount to significantly more than the above-identified judicial exception (the abstract idea). Looking at the additional elements in combination does not provide an inventive concept because the claim does not recite a nonconventional predictive model architecture, model training or weight update technique, similarity technician, data storage technique, merchant system communication protocol, payment network arrangement, or technical performance improvement produced by the claimed combination. The claims feedback loop is considered as an ordered combination but it merely automates the financial authorization decision using broadly recited computer and predictive model functions at a high level of generality. Thus, Rep. Claim 1 does not provide an inventive concept.
Independent Claims 12 and 17 are system claims having elements that perform the same abstract processing and generic computer operations recited in Rep. Claim 1. Independent Claims 12 and 17 add no additional elements beyond those of Rep. Claim 1 that would amount to significantly more than the abstract idea. Therefore, Independent Claims 12 and 17 also do not recite an inventive concept under Step 2B.
Dependent Claims Not Significantly More
The dependent claims have been given the full two-part analysis including analyzing the additional limitations both individually and in combination. The dependent claim(s) when analyzed both individually and in combination are also held to be patent ineligible under 35 U.S.C. § 101. Dependent claims are dependent on Independent Claims and include all the limitations of the Independent Claims. Therefore, all dependent claims recite the same Abstract Idea. Dependent claims do not contain additional elements that integrate the abstract idea exception into a practical application or recite an inventive concept because the additional elements: (1) are mere instructions to apply the abstract idea exception; and/or (2) further limit the abstract idea exception of the Independent Claims. The abstract idea itself cannot provide the inventive concept or practical application. MPEP §§ 2106.05(I), 2106.04(d)(III).
Dependent Claims 4–11, 13–16, 19, 20, 24, and 25 all recite “wherein” clauses or limitations that further limit the abstract idea of the Independent Claims and contain no additional elements not otherwise analyzed with Rep. Claim 1. Claim 25 is not an additional element because it does not change how the predictive model operates or is trained. The decision logic is not altered. Thus, this feature is mere extra solution activity or post solution activity. Once an abstract decision “approve/deny” s made, sending a signal to a merchant system is a conventional way to communicate a result in the payment system arts, not a non-conventional computer specific improvement. The timing of the transmission is in furtherance of the abstract idea and forms part of the same abstract idea. An inventive concept or practical application cannot be furnished by an abstract idea exception itself. MPEP §§ 2106.05(I), 2106.04(d)(III).
Conclusion
Claims 1, 4–17, 19, 20, 24, and 25 are therefore drawn to ineligible subject matter as they are directed to an abstract idea without significantly more. The analysis above applies to all statutory categories of invention. As such, the presentment of Rep. Claim 1 otherwise styled as another statutory category is subject to the same analysis.
Examiner Statement of Prior Art—No Prior Art Rejections
Based on the prior art search results, the prior art of record fails to anticipate or render obvious the claimed subject matter of the instant application. While some individual features of Claims 1, 4–17, 19, 20, 24, and 25 may be shown in the prior art of record—no known reference, alone or in combination, would provide the invention of Claims 1, 4–17, 19, 20, 24, and 25. The prior art most closely resembling the applicant’s claimed invention are:
Marx et al. (U.S. Pat. Pub. No. 2014/0379576) is pertinent because it discloses a primary user of a payment account may designate one or more secondary users who may access and utilize the primary user's payment account. When a secondary user is making a transaction using the payment account, the primary user may be notified in real time.
Zhang et al. (U.S. Pat. Pub. No 2020/0314101) is pertinent because it discloses machine learning models trained on historical authorization requests and predictive model systems that automatically authorize/decline transactions based on scores versus thresholds.
Unnerstall et al. (U.S. Pat. Pub. No. 2018/0121913) is pertinent because it discloses location sensing and proximity checks using geofencing and whether a transaction satisfies or violates a geofence rule.
Daetz (U.S. Pat. Pub. No. 2019/0385166) is pertinent because it discloses system and methods to enable a user to set a spend limits associated with a payment account, a current balance on the account, and a time period. In the event of a transaction amount above the spending limit, the system automatically may approve or deny is the amount is “below a certain amount (e.g., $10, $100, etc.). ¶ 70.
Sahni et al. (U.S. Pat. No. 10,445,739) is pertinent because it discloses systems and methods for sharing financial accounts via a mobile wallet system where the mobile wallet system allows for a master wallet associated with a primary account holder to provide limited access to an account of the primary account holder to secondary users. The primary account holder can limit a secondary user's level of access to the funds in the account by establishing spending rules and limits for each secondary user, which limits restrict the secondary users' abilities to spend funds in the account.
FOR: (CA. Pat. App. CA 2790529 A1) is pertinent because it discloses a system and method for managing transaction restrictions for an account in a family of accounts through a network.
NPL: System and Method for Credit Constraints (2004) is pertinent because it discloses a method and system to provide the ability to constrain the usage of an individual's credit card or similar payment device, such as in the parent-child relationship.
Conclusion
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 JAMES H MILLER whose telephone number is (469)295-9082. The examiner can normally be reached M-F: 10- 4 PM (EST).
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, Bennett M Sigmond can be reached at (303) 297-4411. 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.
/JAMES H MILLER/ Primary Examiner, Art Unit 3694
1 Statements of intended use fail to limit the scope of the claim under BRI. MPEP § 2103(I)(C).
2 Examiner modified the as-claimed limitation by placing the condition first.
Prosecution Insights