DETAILED ACTION
This communication is a Final Office Action on the merits in response to communications received on 02/23/2026. Claims 1-2, 5-16, and 18-20 were previously presented and have been addressed below. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Rejections - 35 USC § 101
1. 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.
2. Claims 1-2, 5-16, and 18-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Under Step 1 of the two-part analysis from Alice Corp, claim 1 recites a process (i.e., an act or step, or a series of acts or steps), claim 19 recites a machine (i.e., a concrete thing, consisting of parts, or of certain devices and combination of devices), and claim 20 recites a manufacture (i.e., an article that is given a new form, quality, property, or combination through man-made or artificial means). Thus, each of the claims fall within one of the four statutory categories.
3. Under Step 2A – [Prong One] of the two-part analysis from Alice Corp, the claimed invention recites an abstract idea.
Claim 1 which represents claims 19 and 20 recites:
“receiving the fraud claim from the client, wherein the fraud claim is in respect of a potentially fraudulent transaction associated with the client; retrieving client data associated with the client, wherein the client data comprises (i) data relating to historical financial transactions associated with the client and (ii) data relating to one or more characteristics of the client;”, “assigning…the client to a cluster, based on the data relating to the one or more characteristics of the client, and wherein the cluster is selected from among multiple clusters generated based on data associated with characteristics of multiple other clients;”, “generating…an anomaly score for the potentially fraudulent transaction based on the data relating to the historical financial transactions associated with the client, based on one or more parameters of the potentially fraudulent transaction, and based on the assigned cluster;”, “generating…a fraud score associated with the fraud claim based on the anomaly score, based on the data relating to the historical financial transactions associated with the client, based on the assigned cluster, and based on the one or more parameters of the potentially fraudulent transaction;”, “determining, based on the fraud score, that the fraud claim is legitimate;” and “in response to determining that the fraud claim is legitimate, initiating an instruction so as to reverse the potentially fraudulent transaction”
The limitations as drafted are processes under their broadest reasonable recite an abstract idea of determining whether a fraud claim is legitimate and reversing a potentially fraudulent transaction which encompasses fundamental economic principles or practices (i.e., mitigating risk), commercial or legal interactions (including, legal obligations, marketing or sales activities or behaviors; business relations), mental processes, (i.e., observations, evaluations, judgments, and opinions) subject matter that falls within the certain methods of organizing human activity and mental processes groupings enumerated in MPEP 2106.05(a)(2)
Applicant’s Specification emphasizes in at least ¶ 0003 Real estate brokerages compete to provide faster, more accurate estimates for their clients to select properties for leasing/purchasing. Many factors affect the cost to a particular client for a particular property. A tool is needed to help brokers provide fast, accurate estimates and visualizations for clients to compare properties under consideration.
The limitations of “receiving”, “retrieving”, “assigning”, “generating”, “determining”, “initiating” in the context of the claim describe evaluations and activities that banks and card issuers typically perform to protect against false or fraudulent chargebacks from their customers. Thus, the series of steps cover concepts related to fundamental economic practices and commercial interactions - marketing or sales activities or behaviors, business relations. Also, the limitations of “assigning”, “generating”, and “determining” in the context of the claim cover mental processes for collecting and analyzing client/transaction data for fraud, which are steps that can be practically performed in the human mind with or without pen and paper. Accordingly, the claim recites an abstract idea.
4. Under Step 2A – Prong Two of the two-part analysis from Alice Corp, this judicial exception is not integrated into a practical application because the additional elements of: “being performed by one or more processors”, “using a clustering model”, “using a trained anomaly detection model”, “using a trained classification model”, – see claim 1, “a system”, “one or more processors”, ”one or more databases”– see claim 19, “a non-transitory computer-readable medium having stored thereon computer program code”, “one or more processors to cause the one or more processors”– see claim 20 are recited at a high-level of generality in light of the specification [Fig. 2, ¶ 0045, 0055, 0093-0094]. Thus, the specification describes the additional elements in general terms, without describing the particulars, the claim limitations may be broadly but reasonably construed as reciting generic computer components and functionalities in light of the disclosure. These claimed additional elements merely recite the words "apply it" (or an equivalent) with the judicial exception, or merely include instructions to implement an abstract idea on a computer, or merely using a computer as a tool to perform an abstract idea, as discussed in MPEP 2106.05(f)
The other additional element of: “a method of determining whether a fraud claim initiated by a client is legitimate, the method comprising:” is merely indicating a field of use or technological environment in which to apply a judicial exception. See MPEP 2106.05(h)
Thus, the additional claim elements are not indicative of integration into a practical application, because the claims do not involve improvements to the functioning of a computer, or to any other technology or technical field (MPEP 2106.05(a)), the claims do not apply or use the abstract idea to effect a particular treatment or prophylaxis for a disease or medical condition (Vanda Memo), the claims do not apply the abstract idea with, or by use of, a particular machine (MPEP 2106.05(b)), the claims do not effect a transformation or reduction of a particular article to a different state or thing (MPEP 2106.05(c)), and the claims do not apply or use the abstract idea in some other meaningful way beyond generally linking the use of the abstract idea to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception (MPEP 2106.05(e) and Vanda Memo). Therefore, the claims do not, for example, purport to improve the functioning of a computer. Nor do they effect an improvement in any other technology or technical field. Accordingly, the additional elements do not impose any meaningful limits on practicing the abstract idea and the claims are directed to an abstract idea.
5. The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception because, as discussed above with respect to integration of the abstract idea into a practical application, the additional element(s) of: “being performed by one or more processors”, “using a clustering model”, “using a trained anomaly detection model”, “using a trained classification model”, – see claim 1, “a system”, “one or more processors”, ”one or more databases”– see claim 19, “a non-transitory computer-readable medium having stored thereon computer program code”, “one or more processors to cause the one or more processors”– see claim 20 amount to no more than mere instructions in which to apply the judicial exception and do not provide an inventive concept at Step 2B.
6. Claims 2, 4-16, 18 are the dependent claims.
Claim 2 recites “wherein the one or more parameters comprise one or more of: data indicating a type of merchant associated with the potentially fraudulent transaction; an amount associated with the potentially fraudulent transaction; a time of day associated with the potentially fraudulent transaction; and a day of a week associated with the potentially fraudulent transaction.” which further describes the type of data/information that may be used in conjunction with the abstract idea, but does not make the claimed invention any less abstract. Claim 4 recites “wherein: the client data further comprises data relating to one or more characteristics of the client; and determining the fraud score is further based on the data relating to the one or more characteristics of the client.” which further describes the type of data/information that may be within the abstract idea, but does not make the claimed invention any less abstract. Claim 5 recites “wherein the one or more characteristics comprise one or more of: an age of the client; an earning potential or a salary of the client; a gender of the client; an address of the client; and a credit score of the client.” which further describes the type of data/information that may be within the abstract idea, but does not make the claimed invention any less abstract. Claim 6 recites “wherein determining the fraud score comprises: extracting, based on the data relating to the historical financial transactions associated with the client, one or more client transaction features; comparing the one or more client transaction features to stored client transaction features; and based on the comparison, determining the fraud score.” which further narrows how the abstract idea may be performed as it describes processes of parsing and comparing data at a high level of generality and merely uses generic computer components or machinery as a tool to perform the processes. Claim 7 recites “wherein the one or more client transaction features and the stored client transaction features are representative of one or more of: types of merchants; for each type of merchant from among multiple types of merchants, amounts associated with the type of merchant; one or more spending patterns; times of day; and days of a week.” which further describes the type of data/information that may be within the abstract idea, but does not make the claimed invention any less abstract. Claims 8 and 12 recite further comprising, prior to receiving the fraud claim from the client, obtaining the stored client transaction features by: retrieving other client data associated with multiple other clients, wherein the other client data comprises data relating to historical financial transactions associated with the other clients; extracting, based on the data relating to the historical financial transactions associated with the other clients, other client transaction features; and storing the other client transaction features.” which further narrows how the abstract idea may be performed as it describes processes of parsing and comparing data at a high level of generality and merely uses generic computer components or machinery as a tool to perform the processes. Claims 9 and 13 recite “wherein retrieving the other client data comprises: retrieving a dataset of client data; extracting features from the dataset of client data; based on one or more similarities between the extracted features, assigning each feature to one of multiple groups; and retrieving the other client data from one of the groups.” which further narrows how the abstract idea may be performed as it describes processes of parsing and comparing data at a high level of generality and merely uses generic computer components or machinery as a tool to perform the processes. Claims 10 and 14 recite wherein extracting the other client transaction features comprises: inputting the other client data to a trained machine learning model; and outputting the other client transaction features using the trained machine learning model to which the other client data was input” which further describes the how the other client transaction features are analyzed, but does not make the claimed invention any less abstract. The additional element recited in the claim “a trained machine learning model” is no more than generic computer components and known techniques and/or algorithms used as tools to perform the recited abstract idea. See MPEP 2016.05(f). Claims 11 and 15 recite “wherein determining the fraud score further comprises: extracting, based on the data relating to the one or more characteristics of the client, one or more client characteristic features; comparing the one or more client characteristic features to stored other client characteristic features; and based on the comparison, determining the fraud score.” which further narrows how the abstract idea may be performed as it describes processes of parsing and comparing data at a high level of generality and merely uses generic computer components or machinery as a tool to perform the processes. Claim 16 recites “wherein determining that the fraud claim is legitimate comprises: comparing the fraud score to a threshold; and based on the comparison, determining that the fraud claim is legitimate” which further narrow how the abstract idea may be performed, but does not make the claim any less abstract. Claim 18 recites “further comprising, prior to determining that the fraud claim is legitimate: determining a trust score associated with the client; and adjusting the fraud score based on the trust score, wherein determining that the fraud claim is legitimate is further based on the adjusted fraud score” which further narrow how the abstract idea may be performed, but does not make the claim any less abstract. Therefore, with respect to the dependent claims when viewed separately and in combination with the judicial exception, the recited limitations as whole fail to integrate the judicial exception into a practical application or provide an inventive concept.
Claim Rejections - 35 USC § 103
7. 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.
8. The factual inquiries 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 non-obviousness.
9. Claim(s) 1-2, 5-7, 11, 15-16, and 18-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Douglas (US 2021/0326884 A1) in view of Duan (US 2015/0095247 A1) in further view of Venturelli (US 11,288,673 B1).
With respect to claims 1, 19, and 20, Douglas discloses
a method (¶ 0073), system (abstract, ¶ 0006: discloses a system), and non-transitory computer-readable medium (¶ 0008: discloses a non-transitory computer readable medium) of determining whether a fraud claim initiated by a client is legitimate (¶ 0073: discloses a method for providing a dispute workflow.),
the method being performed by one or more processors (¶ 0033: discloses server 111 may include one or more processors 220) and comprising:
receiving the fraud claim from the client (¶ 0073: discloses server 111 may handle a dispute brought by user 120. For example, server 111 may receive an electronic message from user 120 who wishes to dispute a transaction made with the merchant 140.), wherein the fraud claim is in respect of a potentially fraudulent transaction associated with the client (¶ 0073: discloses user 120 may claim that the transaction is unauthorized, fraudulent, or that merchant 140 overcharged the user.);
retrieving client data associated with the client, wherein the client data comprises (i) data relating to historical financial transactions associated with the client and (ii) data relating to one or more characteristics of the client (¶ 0035-0036, 0049-0050: discloses the server may receive data relating to an activity of the user 120. The data relating to the activity of the user may include data indicating that the user has made a purchase. The data may also include data indicating suspicious transactions. Data stored may include historical fraud or disputes data, transaction data, credit rating of user 120.);
determining, based on the fraud score, that the fraud claim is legitimate (¶ 0050, 0076: discloses if the transaction score is higher than a predetermined threshold, the transaction may be legitimate.)
The Douglas reference does not explicitly the following limitations.
However, the Duan reference is related to fraud detection systems (¶ 0001) and teaches:
assigning, using a clustering model, the client to a cluster, based on the data relating to the one or more characteristics of the client (¶ 0018, 0024, 0046: discloses the event profile may include purchase information for the event. A purchase information may include a user’s name, payment information, credit card information, etc. The fraud detection system may access information associated with a set of event profiles. A clustering function may assign each sample from the set to a cluster.), and wherein the cluster is selected from among multiple clusters generated based on data associated with characteristics of multiple other clients (¶ 0018, 0024, 0046: discloses the first cluster may correspond to event profiles that have been previously classified as fraudulent. The second cluster may correspond to event profiles that have been previously classified as legitimate. Fraud detection system processes and arranges event profiles into a first or second cluster corresponding to their fraud classification.);
determining, based on the fraud score, that the fraud claim is legitimate (¶ 0038: discloses fraud-detection system 160 may approve or deny pay-out requests based on the fraud score of an event profile); and
in response to determining that the fraud claim is legitimate, initiating an instruction so as to reverse the potentially fraudulent transaction. (¶ 0038: discloses the fraud-detection system may receive a request to pay out funds. If the fraud score for the event profile is less than a threshold fraud score, the fraud detection system may approve the request to pay out the funds and facilitate the transfer of the requested funds.)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, to have modified the system and methods of Douglas, to include assigning, using a clustering model, the client to a cluster, based on the data relating to the one or more characteristics of the client, and wherein the cluster is selected from among multiple clusters generated based on data associated with characteristics of multiple other clients; determining, based on the fraud score, that the fraud claim is legitimate; in response to determining that the fraud claim is legitimate, initiating an instruction so as to reverse the potentially fraudulent transaction, as disclosed by Duan to achieve the claimed invention. As disclosed by Duan, the motivation for the combination would have been to prevent users from violating the terms of services of the system and making requests for fraudulent financial transactions. (¶ 0024)
The combination of Douglas and Duan do not explicitly disclose the following limitations. In the same field of endeavor, the Venturelli reference is related to utilizing multiple machine learning models to detect fraud (col. 2:43-60) and teaches:
generating, using a trained anomaly detection model, an anomaly score for the potentially fraudulent transaction based on the data relating to the historical financial transactions associated with the client, based on one or more parameters of the potentially fraudulent transaction, and based on the assigned cluster (col. 2:43-60, col. 5:30-35, cols. 5-6:65-5, col. 8:13-21: discloses an unsupervised ML anomaly detector exists for each cluster. The unsupervised ML anomaly detector generates an anomaly score. The anomaly score represents how similar or different is with historic requests having similar fraud scores.);
generating, using a trained classification model, a fraud score associated with the fraud claim based on the anomaly score (col. 8:27-29: discloses processing based on comparisons between the fraud score and anomaly score with various thresholds.), based on the data relating to the historical financial transactions associated with the client, based on the assigned cluster, and based on the one or more parameters of the potentially fraudulent transaction (col. 2:43-60, col. 8:27-29);
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, to have modified the combination of Douglas and Duan, to include generating, using a trained anomaly detection model, an anomaly score for the potentially fraudulent transaction based on the data relating to the historical financial transactions associated with the client, based on one or more parameters of the potentially fraudulent transaction, and based on the assigned cluster; generating, using a trained classification model, a fraud score associated with the fraud claim based on the anomaly score, based on the data relating to the historical financial transactions associated with the client, based on the assigned cluster, and based on the one or more parameters of the potentially fraudulent transaction, as disclosed by Venturelli to achieve the claimed invention. As disclosed by Venturelli, the motivation for the combination would have been to provide benefits for identifying possible false positives and/or false negatives before deciding that fraud is or is not likely as expressly suggested by Venturelli (col. 1:6-16, col. 2:55-60)
With respect to claim 2, the combination of Douglas, Duan, and Venturelli discloses the method of claim 1,
wherein the one or more parameters comprise one or more of:
data indicating a type of merchant associated with the potentially fraudulent transaction (¶ 0036, 0073: Douglas discloses analysis of the historical fraud or disputes data associated with merchant 140.); an amount associated with the potentially fraudulent transaction (¶ 0050 – see Douglas); a time of day associated with the potentially fraudulent transaction; and a day of a week associated with the potentially fraudulent transaction. (¶ 0050 – see Douglas)
With respect to claim 5, the combination of Douglas, Duan, and Venturelli discloses the method of claim 1,
wherein the one or more characteristics comprise one or more of:
an age of the client; an earning potential or a salary of the client; a gender of the client; an address of the client; and a credit score of the client (¶ 0036, 0075: Douglas discloses credit rating of user 120.).
With respect to claim 6, the combination of Douglas, Duan, and Venturelli discloses the method of claim 1, wherein determining the fraud score comprises:
extracting, based on the data relating to the historical financial transactions associated with the client, one or more client transaction features ¶ 0049-0050: Douglas discloses server may receive data relating to the activity of user 120. The data may include the type of transaction, the frequency of the transaction, the deviation of the amount of transaction, the location of the transaction.);
comparing the one or more client transaction features to stored client transaction features ¶ 0050: Douglas discloses server 111 may determine the type of transaction is out of the scope of a user’s normal purchases.); and based on the comparison, determining the fraud score. (¶ 0050: Douglas discloses server 111 may assign a transactional score based on any one or a combination of factors such as the type of transaction, the location of the transactions, etc.)
With respect to claim 7, the combination of Douglas, Duan, and Venturelli discloses the method of claim 6,
wherein the one or more client transaction features and the stored client transaction features are representative of one or more of:
types of merchants; for each type of merchant from among multiple types of merchants, amounts associated with the type of merchant; one or more spending patterns; times of day; and days of a week. (¶ 0050: Douglas discloses server 111 may detect that three transactions for TV’s were made within one week involving different types of merchants.)
With respect to claims 11 and 15, the combination of Douglas, Duan, Venturelli discloses the method of claim 1,
wherein determining the fraud score further comprises: extracting, based on the data relating to the one or more characteristics of the client, one or more client characteristic features (¶ 0049-0050: Douglas discloses server may receive data relating to the activity of user 120. The data may include the type of transaction, the frequency of the transaction, the deviation of the amount of transaction, the location of the transaction.); comparing the one or more client characteristic features to stored other client characteristic features (¶ 0050, 0057: Douglas discloses server may detect a charge has been made at a merchant located over 250 miles from where the user typically makes purchases.); and based on the comparison, determining the fraud score. (¶ 0050, 0077: Douglas discloses the transaction score may be based on data indicating suspicious transactions relating to activities of the user.)
With respect to claim 16, the combination of Douglas, Duan, Venturelli discloses the method of claim 1,
wherein determining that the fraud claim is legitimate (¶ 0038: Duan discloses fraud-detection system 160 may approve or deny pay-out requests based on the fraud score of an event profile) comprises:
comparing the fraud score to a threshold; and based on the comparison, determining that the fraud claim is legitimate. (¶ 0038: Duan discloses the fraud-detection system may receive a request to pay out funds. If the fraud score for the event profile is less than a threshold fraud score, the fraud detection system may approve the request to pay out the funds and facilitate the transfer of the requested funds.)
With respect to claim 18, the combination of Douglas, Duan, Venturelli discloses the method of claim 1, further comprising, prior to determining whether the fraud claim is legitimate: determining a trust score associated with the client (¶ 0075: Douglas discloses a user reputation score may identify whether the user has a history of disputing transactions, whether those disputes were resolved in favor of user 120 or merchant 140, whether the user is considered by FSP 110 to be in good standing, or whether the user is a participant in program(s) that include expedited dispute resolution as a benefit, etc.); and adjusting the fraud score based on the trust score (¶ 0075-0077: Douglas discloses the server may process the dispute with the expedited dispute resolution process when it determines that the user has a high reputation score.), wherein determining whether the fraud claim is legitimate is further based on the adjusted fraud score. (¶ 0075-0077: Douglas discloses the server may determine whether the transactional score is higher than a predetermined threshold score which may indicate the transactions is more likely valid.)
10. Claim(s) 8-10 and 12-14 is/are rejected under 35 U.S.C. 103 as being unpatentable over Douglas in view of Duan in view of Venturelli in further view of Kramme (US 2021/0374764 A1).
With respect to claims 8 and 12, the combination of Douglas, Duan, and Venturelli discloses the method of claim 6, further comprising,
prior to receiving the fraud claim from the client (¶ 0024: Douglas discloses the FSP may identify suspicious transactions before, during, or after the FSP customers conduct such transactions.),
The combination of Douglas, Duan, and Venturelli does not explicitly disclose the following limitations. However, Kramme discloses:
obtaining the stored client transaction features by: retrieving other client data associated with the multiple other clients (¶ 0041), wherein the other client data comprises data relating to historical financial transactions associated with the other clients (¶ 0043: discloses obtaining cardholder related or other customer related information.); extracting, based on the data relating to the historical financial transactions associated with the other clients, other client transaction features (¶ 0043: discloses analyzes information obtained from account records to identify spending patterns associated with different cardholders.); and storing the other client transaction features. ¶ 0043: discloses data indicative of the behavior patterns identified may be stored in an account holder behaviors database.)
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, to have modified the combination of Douglas, Duan, and Venturelli to include the steps for obtaining the stored client transaction features, as disclosed by Kramme to achieve the claimed invention. As disclosed by Kramme, the motivation for the combination would have been to learn from other financial accounts which types of data tend to be indicative of different classifications which then may be used to facilitate a more in-depth analysis or investigation. (¶ 0026-0027)
With respect to claims 9 and 13, the combination of Douglas, Duan, and Venturelli, Kramme discloses the method of claim 8,
wherein retrieving the other client data comprises:
retrieving a dataset of client data (¶ 0054: Kramme discloses multi-account data 82 may represent data associated with multiple financial accounts.); extracting features from the dataset of client data ¶ 0044-0045: Kramme discloses analyzes a transaction in view of past spending patterns of a particular cardholder.); based on one or more similarities between the extracted features, assigning each feature to one of multiple groups (¶ 0045: Kramme discloses utilizes the individual spending patterns when detecting and/or classifying fraud); and retrieving the other client data from one of the groups. (¶ 0043: Kramme discloses analyzes information obtained from account records to identify spending patterns associated with different cardholders).
With respect to claims 10 and 14, the combination of Douglas, Duan, Venturelli, Kramme discloses the method of claim 8, wherein extracting the other client transaction features comprises:
inputting the other client data to a trained machine learning model (¶ 0024, 0026: Kramme discloses a machine learning program may be trained using past dispute resolution interactions with customers and the associated outcomes.); and
outputting the other client transaction features using the trained machine learning model to which the other client data was input. ¶ 0026, 0058: Kramme discloses the machine learning program provides fraud classifications made in connection with multiple other financial accounts.)
Response to Arguments
11. Applicant's arguments filed 02/23/2026 have been fully considered but they are not persuasive.
With Respect to Rejections Under 35 USC 103
Applicant argues “Claim 1 recites, inter alia, "assigning, using a clustering model, the client to a cluster, based on the data relating to the one or more characteristics of the client, and wherein the cluster is selected from among multiple clusters generated based on data associated with characteristics of multiple other clients". The Office Action acknowledges that these features are absent from Douglas, and the Examiner therefore cites Duan as allegedly disclosing them. However, like Douglas, Duan also fails to teach or suggest these features. Duan is directed to fraud classification in an event-management context and describes clustering of event-profile samples, not clustering of clients based on client characteristics. In particular, Duan explains that a "clustering function may assign each sample from the set to a cluster" (Duan at [0046]), where the set is a set of event profiles. Duan further explains that the system may access "a first and second cluster of event profiles," where "the first cluster of event profiles may be identified as being associated with fraud and the second cluster of event profiles may be identified as legitimate" (Duan at [0062]). Accordingly, Duan's clustering is applied to event profiles, categorized as "fraud" versus "legitimate," rather than to clients based on characteristics of the client. The Examiner's reliance on Duan's discussion that an event profile may include "purchase information" (a user's name, payment information, credit card information, etc.) (see, e.g., Duan at [0018]) does not transform Duan's clustering of event-profile samples into the claimed step of assigning the client to a cluster based on data relating to one or more characteristics of the client. Therefore, Duan fails to teach or suggest at least "assigning, using a clustering model, the client to a cluster, based on the data relating to the one or more characteristics of the client" The Examiner respectfully disagrees.
The Applicant’s arguments are not persuasive. In the instant case, the response discusses the “assigning” step and attempts to distinguish the particular type of data or information being clustered from the prior art. As can be seen from the teachings of the Duan reference in at least [¶ 0018, 0024, 0038, 0046]: The fraud detection system includes an event profile with purchase information for the event. The purchase information may include a user’s name, payment information, credit card information, etc. The fraud detection system may access information associated with a set of event profiles. A clustering function may assign each sample, i.e., user information, from the set to a cluster. The first cluster may correspond to event profiles that have been previously classified as fraudulent. The second cluster may correspond to event profiles that have been previously classified as legitimate. Fraud detection system processes and arranges event profiles into a first or second cluster corresponding to their fraud classification. As best understood from the Duan reference, the reference provides a finding that using a clustering function for assigning user information to a cluster was a known technique in fraud detection systems. Duan’s clustering function assigns the user information from an event profile to a cluster which is performing the limitation of assigning a client to a cluster as claimed. For these reasons, the rejections under 103 are being maintained.
Applicant further argues “There is no disclosure in Venturelli of the anomaly detection being "based on the assigned cluster", as claimed, where the cluster is assigned to the client based on client characteristics. Rather, Venturelli's anomaly detection is only based on a fraud score. Therefore, Venturelli fails to teach or suggest at least "generating, using a trained anomaly detection model, an anomaly score for the potentially fraudulent transaction based on the data relating to the historical financial transactions associated with the client, based on one or more parameters of the potentially fraudulent transaction, and based on the assigned cluster" (emphasis added). The Examiner respectfully disagrees.
The Applicant’s arguments are not persuasive. In the instant case, the response discusses the “anomaly” score and attempts to distinguish the particular type of data or information being used to generate the score from the prior art. As can be seen from the teachings of the Venturelli reference in at least [col. 2:43-60, col. 5:30-35, cols. 5-6:65-5, col. 8:13-21]: discloses an unsupervised ML anomaly detector exists for each cluster. The unsupervised ML anomaly detector generates an anomaly score. The anomaly score represents how similar or different is with historic requests having similar fraud scores.] As best understood from the Venturelli reference, the reference provides a finding that generating anomaly score based on data/information from historical requests was a known technique in detecting fraud associated with requests. The prior art passages that discuss the technique for generating an anomaly score meet the limitation as claimed. For these reasons, the rejections under 103 are being maintained.
Applicant further argues “Claim 1 therefore claims a particular sequence: the output of the clustering model is provided to the anomaly detection model, and the output of the anomaly detection model is then provided to the trained classification model which then generates the fraud score. In contrast, Venturelli's described flow involves (i) a supervised classifier to generate a fraud score, and then (ii) an unsupervised anomaly detector which receives the fraud score. Clearly, therefore, Venturelli does not disclose or suggest the claimed sequence set out in claim 1. In particular Venturelli fails to teach or suggest fraud score generation before anomaly detection.” The Examiner respectfully disagrees.
The Applicant’s response to arguments are not persuasive. The response discusses the sequence of steps to distinguish the claims from the prior art. It is important for Applicant to note that "the prior art’s mere disclosure of more than one alternative does not constitute a teaching away from any of these alternatives because such disclosure does not criticize, discredit, or otherwise discourage the solution claimed…." In re Fulton, 391 F.3d 1195, 1201, 73 USPQ2d 1141, 1146 (Fed. Cir. 2004). For these reasons, the rejections under 103 are being maintained.
With Respect to Rejections Under 35 USC 101
Applicant argues “Applicant respectfully submits that the Examiner's characterization of the claims as "certain methods of organizing human activity" is inappropriate given the claim wording. While the subject matter arises in a financial context, it does not recite a fundamental economic principle (e.g., a pricing/contractual scheme), nor does it recite rules for managing a commercial or legal interaction in the abstract. Rather, the independent claims recite a particular computer-implemented, multi-model machine learning pipeline that processes client and transaction data in a defined sequence to generate model outputs and to initiate a transaction-reversal instruction. Accordingly, the subject matter is directed at a specific technical mechanism for operating a fraud- adjudication computing system, and the subject matter is not directed at organizing human or commercial activity as such.” The Examiner respectfully disagrees.
The Applicant’s arguments are not persuasive. The background of the Specification [¶ 0001-0003] makes clear the focus of the claimed invention is for methods of determining whether a fraud claim is legitimate, not improvements to machine learning technology. As for the features of “a particular computer-implemented multi-model machine learning pipeline” and “a fraud-adjudication computing system” the reply describes these features in a conclusory manner. At best, Applicant is relying upon these features as an attempt to limit the claimed invention to “a particular technological environment or field of use” which in insufficient to save the claim. See MPEP 2106.05(h) Under the BRI, the series of steps recited in the claim cover techniques for determining a fraud score for a fraud claim initiated by client. The claim uses the fraud score to indicate the fraud claim is legitimate and reverses a transaction for a client which is subject matter than may be reasonably characterized as fundamental economic practices (i.e., mitigating risk) and commercial interactions (i.e., sales/marketing activities, business relations) that falls within the certain methods of organizing human activity grouping. See Bozeman Fin. LLC v. Fed. Reserve Bank of Atlanta, 955 F.3d 971, 978 (Fed. Cir. 2020) - Claims to methods for detecting fraud in financial transactions is an example of a fundamental economic principle or practice. For these reasons, the rejections under 101 are being maintained.
.
Applicant argues “Regarding the Examiner's characterization of the claims as "mental processes", the Kim Memo explains that the mental process grouping applies when a claim contains limitation(s) that "can practically be performed in the human mind," and conversely "a claim does not recite a mental process when it contains limitation(s) that cannot practically be performed in the human mind." The Kim Memo further cautions that "[t]he mental process grouping is not without limits" and reminds examiners "not to expand this grouping" to encompass limitations that "cannot practically be performed in the human mind." (See Kim Memo, p. 2.) When considered as a whole, independent claim 1 recites a multi-model machine learning pipeline for fraud claim adjudication that is not practically performable in the human mind. Claim 1 recites, inter alia:
" assigning, using a clustering model, the client to a cluster, based on the data relating
to the one or more characteristics of the client, and wherein the cluster is selected from among multiple clusters generated based on data associated with
characteristics of multiple other clients; " generating, using a trained anomaly detection model, an anomaly score for the potentially fraudulent transaction based on the data relating to the historical financial transactions associated with the client, based on one or more parameters of the potentially fraudulent transaction, and based on the assigned cluster; " generating, using a trained classification model, a fraud score associated with the fraud claim based on the anomaly score, based on the data relating to the historical financial transactions associated with the client, based on the assigned cluster, and
based on the one or more parameters of the potentially fraudulent transaction
These limitations involve the coordinated operation of three distinct machine learning models processing client data and transaction parameters in a defined sequence. The specification (at [0063]-[0085]) describes training these models on historical transaction datasets, applying dimensionality reduction, feature extraction, and optimization techniques, and generating outputs that feed into subsequent model stages. These are not observations or evaluations that can practically be carried out by a person in their head, or with pencil and paper.” The Examiner respectfully disagrees.
The Applicant’s arguments are not persuasive. The reply merely cites to guidance from the Kim Memo which is insufficient to confer patent eligibility. The courts have previously held claims for collecting information and recognizing certain information to be abstract and mental processes that fall within the abstract idea groupings. As for the ordered combination or sequence of limitations, claims can recite a mental process even if they are being performed by a computer. See MPEP 2106.04(a)(2)(III)(c) Merely adding “one or more machine learning models” to a claim that recites a judicial exception does not preclude limitations from being in the mental processes grouping. As for the passages of the Specification being relied upon that discuss advantages/benefits of training the one or more machine learning models, however, there are no findings that support technical improvements to machine learning or computer technology. For these reasons, the rejections under 101 are being maintained.
Applicant further argues “At minimum, this is a "close call" under the Office's own guidance. The Kim Memo instructs that, if" it is a 'close call' as to whether a claim is eligible," examiners "should only make a rejection when it is more likely than not (i.e., more than SO%) that the claim is ineligible under 35 U.S. C. 101." (See Kim Memo, p. 5.) Applicant respectfully submits that the Examiner's mental-process characterization does not satisfy that "more likely than not" threshold when the claim is evaluated as a whole, including the coordinated multi-model machine learning pipeline operating on client and transaction data. Accordingly, Applicant respectfully submits that independent claims 1, 19, and 20, along with their dependent claims, are not directed to the mental process grouping of abstract ideas under Step 2A, Prong One.” The Examiner respectfully disagrees.
The Applicant’s arguments are not persuasive. The reply does not challenge any of the identified limitations from the previous rejection that were considered under Prong One of the analysis. The reply provides conclusory statements related to a close call and relies on guidance from the Kim Memo. Accordingly, the response does not make the claimed invention any less abstract or change the previous analysis. For these reasons, the rejections under 101 are being maintained.
Applicant further argues “Even assuming, arguendo, that the claims recite an abstract idea, the claims as a whole integrate any such concepts into a practical application. The Kim Memo reminds Examiners to avoid "oversimplifying claim limitations" and to consider: "Whether the claim recites only the idea of a solution or outcome... or the claim covers a particular solution to a problem or a particular way to achieve a desired outcome." (Kim Memo, p. 4.)”
“The present claims do not merely recite the "idea" of detecting fraud. Instead, they recite a specific "how": a multi-stage machine learning architecture in which (i) a clustering model assigns the client to a cluster based on the data relating to the one or more characteristics of the client, (ii) an anomaly detection model generates an anomaly score based on the data relating to the historical financial transactions associated with the client, based on one or more parameters of the potentially fraudulent transaction, and based on the assigned cluster, (iii) a classification model generates a fraud score based on the data relating to the historical financial transactions associated with the client, based on the assigned cluster, and based on the one or more parameters of the potentially fraudulent transaction, and (iv) an instruction is initiated to reverse the potentially fraudulent transaction. This coordinated sequence provides a particular way to improve fraud claim adjudication.” The Examiner respectfully disagrees.
The Applicant’s arguments are not persuasive. The specificity of the presently recited techniques in the claim does not automatically confer eligibility. The judicial exception cannot be relied upon alone for Applicant’s improvement. Other than restating the limitations as claimed, the response does explain how the ordered combination of limitations indicates a technological solution to a technological problem. As for the improvements to fraud claim adjudication, this is an alleged improvement to the abstract idea itself, not to any technological improvement. See BSG Tech LLC v. Buyseasons, Inc., 899 F. 3d1281, 1287-88 (Fed. Cir. 2018). For these reasons, the rejections under 101 are being maintained.
Applicant further argues “The Examiner treats the additional elements as "recited at a high-level of generality" and as "mere instructions to be performed by a computer as opposed to a technological solution to a technological problem." Respectfully, the claims do not merely recite generic "machine learning"; they recite a defined pipeline in which each model's output informs the next model's processing. This inter-stage dependency represents a specific technical architecture, and is not a mere invocation of generic computing.” The Examiner respectfully disagrees.
The Applicant’s arguments are not persuasive. Here, the response discusses features of the machine learning models at high-level of generality and does not provide any support from the Specification with respect to the technical details related to the machine learning models. As such, the recited machine learning technology being applied in the claim simply provides a generic environment in which the claimed method is performed. See also - BSG Tech LLC v. BuySeasons, Inc., 899 F.3d 1281, 1286 (Fed. Cir. 2018) ("[C]laims are not saved from abstraction merely because they recite components more specific than a generic computer.") For these reasons, the rejections under 101 are being maintained.
Applicant further argues “The Examiner further asserts that the technical improvements identified in Applicant's prior response (e.g., reducing false-positive fraud holds, reducing network bandwidth, etc.) should be disregarded because such improvements are allegedly "not discussed in the present Specification" and therefore "the improvements purported by Applicant cannot be relied upon to integrate [the] judicial exception into a practical application or provide an inventive concept." Applicant respectfully disagrees. Under MPEP § 2106.05(a), when an applicant asserts an improvement to computer functionality or another technology/technical field, the inquiry is whether the specification provides a technical explanation as to how to implement the invention such that one of ordinary skill in the art would recognize the claimed invention as providing an improvement. The MPEP further explains that the specification need not explicitly set forth the improvement, so long as it describes the invention such that the improvement would be apparent to a person of ordinary skill in the art.”
“Consistent with that guidance, and with the Kim Memo's reminder to avoid oversimplifying claim limitations and to focus on whether the claim recites a particular way of achieving an outcome rather than merely an outcome, the present specification's description of a coordinated, multi-model pipeline, as discussed above, provides the type of technical disclosure from which the skilled person would understand that the approach can reduce avoidable false- positive fraud holds and associated communications/workflow overhead compared with conventional one-size-fits-all scoring or purely manual review gating. Accordingly, the Examiner should consider these apparent, implementation-linked improvements when evaluating whether the claims integrate any alleged exception into a practical application at Step 2A, Prong Two and, as applicable, whether the claims provide an inventive concept at Step 2B.”
The Applicant’s arguments are not persuasive. Here the remarks are relying upon the guidance discussed from the MPEP 2106.05(a) and Kim Memo which is insufficient to attain eligibility. The remarks discuss in a conclusory manner how the claimed invention provides an inventive concept. As a matter of law, narrowing or reformulating an abstract idea does not add "significantly more" to it. See SAP Am., Inc. v. InvestPic, LLC, No. 2017-2081, slip op. at 14 (Fed. Cir. Aug. 2, 2018) ("What is needed is an inventive concept in the non-abstract application realm. . . . [L]imitation of the claims to a particular field of information. . . does not move the claims out of the realm of abstract ideas.") For these reasons, the rejections under 101 are being maintained.
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 EHRIN PRATT whose telephone number is (571)270-3184. The examiner can normally be reached 8-5 EST Monday-Friday.
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, Lynda Jasmin can be reached at 571-272-6782. 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.
/EHRIN L PRATT/ Examiner, Art Unit 3629
/LYNDA JASMIN/ Supervisory Patent Examiner, Art Unit 3629
Prosecution Insights