DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
The office action is being examined in response to the application filed by the applicant on 09/15/2025.
Claims 1-20 are pending and have been examined.
This action is made FINAL.
Claims 1, 13, and 20, have been Amended, adding limitations to clause 9 with no new matter.
Response to Arguments
35 U.S.C. § 101 Arguments – Abstract Idea
Applicant’s Remarks, see pages 11-12, filed 9/15/2025, with respect to the rejection of claims 1-20 under 35 U.S.C. § 101, have been fully considered and are not persuasive.
On page 11, the Applicant argues with respect to 35 U.S.C. § 101, and respectfully asserts that the claims integrate an abstract idea into a practical application. The Applicant also asserts that the claims are not an abstract idea category at all, but instead, the claimed operations improve the function of an online system using a machine-learning segmentation model trained to classify users into segments, develop an appeasement model, and produce an outlier score to find unusual request rates. The Applicant asserts that the functions performed by claims improves the technical field of automated user support and exception handing in online systems by leveraging machine learning to tailor actions based on statistics of user’s behaviors, enabling accurate, efficient, and consistent processing via performance by the computer system. On pages 11-12, the Applicant argues that the computer is not merely a tool for implementing a human activity, but instead requires specific capabilities of executing a trained machine-learning model to compute scores and generate the outlier. The Applicant further asserts that the segmentation model is recited with specificity, resulting in a specific improvement to computer-based systems, reducing error, and improving speed and accuracy through automating statistical modeling.
The Examiner respectfully disagrees. Improving the speed, efficiency, accuracy, and consistency of the function of a general purpose computing system or systems over a network, either online via the internet or locally or by use of a automation via a machine learning model, is explicitly disclosed in MPEP 2106.05(f)(2), "claiming the improved speed or efficiency inherent with applying the abstract idea on a computer" (or with a machine-learning model) does not integrate a judicial exception into a practical application or provide an inventive concept. Intellectual Ventures I LLC v. Capital One Bank (USA), 792 F.3d 1363, 1367, 115 USPQ2d 1636, 1639 (Fed. Cir. 2015). Further, as disclosed in the 35 U.S.C. § 101 rejection below, the additional elements, i.e. a user device, an online concierge system, processors, systems, non-transitory computer readable medium, executable instructions, or a computer implemented method, instructions, and graphical user interface of the user device, a segmentation model, and an appeasement model, are merely tools used to implement the abstract ideas, i.e. the claims are merely adding the words “apply it” with the abstract ideas. The claims are recited as intended results, depicting all means or methods of resolving the problem encompassed by the claim without explicit or implicit limitation beyond mere instructions to implement the abstract ides, applying the additional elements as tools to perform the instructions, i.e. abstract ideas, and are therefore, not an improvement to the functioning of a computer, or to any other technology or technical field (MPEP 2106.05(a)). Finally, the limitations are generally linking the use of the judicial exception to the particular technological environment or field of use (MPEP 2016.05(h)).
On page 12, the Applicant argues that claim 7 recites a probability mass function used as an appeasement model that reduces processing power, and is therefore an improvement to a technical field, such that it is patent eligible.
The Examiner respectfully disagrees. Reducing processing power to a general-purpose computing structure via automating process via statistical methods is merely applying the mathematical probability mass function as a tool to improve performance of the computing structures, and is therefore disclosed under MPEP 2106.05(f), such that the probability mass function is merely a tool used to implement the abstract ideas, i.e. adding the words “apply it.”
On page 12, the Applicants asserts that the claims are patent eligible and respectfully requests that the rejection should be withdrawn. The Examiner respectfully disagrees and the 5 U.S.C. § 101 rejection is maintained. Please find an updated rejection for 35 U.S.C. § 101 below, reflecting the amendments.
35 U.S.C. § 103 Arguments
Applicant’s arguments, see pages 10-11, filed 9/15/2025, with respect to 35 U.S.C. § 103 in claims 1-20, have been fully considered and are not persuasive.
On page 10, the Applicant asserts that the prior art disclosure of Stashluk does not teach Claim 1, which recites “generating an appeasement model…representing appeasement rates of users in the identified user segments,” and instead, only teaches a returns frequency on a per-user basis.. The Applicant also asserts that Ford does not teach “generating an outlier score for the user…[that] represents a likelihood that a random user within the identified user segment would have the appeasement request rate associated with the user,” because Ford does not describe this sample as being a user’s appeasement request rate, nor describe the PDF distribution as appeasement request rates with other users.
The Examiner respectfully disagrees. Stashluk does, in fact, disclose generating an appeasement model that returns rates of users in various user segments as shown below. Additionally, the combination of Stashluk in view of Ford does, in fact, teach a probability that a user and a random user would have the same outlier score when comparing the probability of a random user’s rate to the user’s rate. While Ford does not explicitly disclose a user’s appeasement request rate or appeasement request rates of other user’s, the combination of Stashluk, comprising of the appeasement model and appeasement request rates of a user among other user’s grouped in segments, where data is stored and accessed for analysis, and Ford, comprising the probability mass function, probability distribution, and outlier probability value (score) between a random value (random user score) and a focus value (user score) in a group, are presented below as an obviousness rejection according to 35 U.S.C. § 103.
The Examiner respectfully disagrees and the 5 U.S.C. § 103 rejection is maintained. Please find an updated rejection for 35 U.S.C. § 103 below, reflecting the Applicant’s amendments.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Independent Claims:
Regarding Claims 1, 13, and 20: The claims recite the following functions: compute a rate, generate a segmentation score, all based on a user’s request for retail delivery order conciliation, which are abstract ideas in the category of “methods of organizing human activity,” more specifically, “Commercial or Legal Interactions” which “include agreements in the form of contracts, legal obligations, advertising, marketing or sales activities or behaviors, and business relations,” (MPEP 2106.04(a)(2)(II)).
The claims recite: generate an outlier score, calculate a chance of matching outlier scores, compare outlier score to threshold, which are abstract ideas in the category of “methods of organizing human activity,” more specifically, “fundamental economic principles or practices” due to “mitigating risk” because the functions culminate in determining if a retail delivery order conciliation request is valid or evidences user misuse or fraud when compared to similarly grouped data (MPEP 2105.04(a)(2)(II)).
The claims also recite: group users into clusters according to attributes, an abstract idea in the category of “Managing Personal Behavior or Relationships or Interactions Between People” because the claims utilize the function above to compare and group customers according to similarities in characterizations of user and order data to best issue an appropriate conciliation to a retail delivery order request. (MPEP 2105.04(a)(2)(II)).
The claims recite: receive an order conciliation request, identify an order, access user data, identify a segment containing a user, and apply an action, which are abstract ideas in the category of “mental processes” or “things that can be performed in the human mind” since these functions “include observations, evaluations, judgments, and opinion” (MPEP 2105.04(a)(2)(III)).
Step 2A Prong 2: The claims also recite data or groups of data that are merely non-functional descriptive information limitations that do not carry patentable weight in the claim because they do not positively recite any additional functions that limit the claims or the structures of the claims. Instead, the specification is focused on the nature of the data being manipulated – i.e., the descriptive nature of the data (MPEP 2106.05(e)).
Insofar as the claims recite the additional elements of hardware in the preambles and in the claims themselves, i.e. a user device, an online concierge system, processors, systems, non-transitory computer readable medium, executable instructions, or a computer implemented method, instructions, and graphical user interface of the user device, each claim and the applicant’s specification at paragraphs [0016 and 0020] (graphical user interface), and [0078-0079] are simply reciting or disclosing generic computing structures at a high level of generality without providing advances or improvements to the technology or structures themselves. These recitations amount to “apply it,” mere instructions to apply the Judicial Exceptions on generic computing structures (MPEP 2106.05(f)).
Lastly, the claims recite training, generating, and applying machine learning models to perform general machine learning model tasks, which are not abstract ideas. These machine learning models, which are additional elements, are recited at a high level of generality without providing advances or improvements to the technical field of Machine Learning. This amounts to “apply it,” mere instructions to generally link the Judicial Exception to a technological environment that merely confines the use of the abstract ideas to general machine learning models without the steps required to achieve the results, and thus fail to add an inventive concept to the claims. Therefore, the claims do not provide for practical applications of the judicial exceptions (MPEP 2106.05(e and h)). The claims as a whole, while looking at additional elements individually and in combination, do not integrate the judicial exceptions into a practical application (MPEP 2106.04).
Step 2B: The analysis above for Step 2A is commensurate with the analysis for this Step 2B, such that the same additional elements described above, taken individually and in combination do not result in the claims, as a whole, amounting to significantly more than the judicial exceptions (MPEP 2106.05).
Dependent Claims:
Claims 2-12, and 14-19 recite further elements related to the computing of request rates, segmenting user data, utilizing machine learning models, finding outlier scores, comparing outliers to thresholds, and applying remediation steps according to steps of the parent claims. These activities fail to differentiate the claims from the related activities in the parent claims and fail to provide any material to render the claimed invention to be significantly more than the identified abstract ideas.
Claims 2-4 and 14-16 recite further limitations for claims 2 and 14: computing request rates by computing a ratio, number of requests received, number of orders, claims 3 and 15: accessing user data based on characteristics of order data, and claims 4 and 16: identifying order data according to characteristics of location data. These limitations narrow how the abstract ideas may be performed but do not make the claims any less abstract.
Claims 5-7 and 17-19 merely impart, state use, or state inclusions of characterizations of machine learning models that narrow how the abstract ideas may be performed but do not make the claims any less abstract. Claims 5 and 17 recite further limitations to the user segmentation model, specifying that it is an unsupervised mode. Claims 6 and 18 recite further limitations to the user segmentation model, specifying that the model uses k-means clustering. Claims 7 and 19 recite further limitations to the appeasement model, wherein it includes a probability mass function that gives a probability distribution comparing values and random values. Furthermore, the machine learning models implemented to perform the steps are recited at a high-level of generality and are only nominally and generically disclosed by the specification as tools for performing these steps.
Claim 8 further limits generating the outlier score comprising comparing user request rates to average segment request rates. These limitations narrow how the abstract idea may be performed but do not make the claims any less abstract.
Claim 9 further limits applying an appeasement action to comprise selecting from a list based on the request made, wherein the possibilities include redelivery and refund issuance to the user. These limitations narrow how the abstract idea may be performed but do not make the claims any less abstract.
Claims 10-12 merely respond to an outlier score by allocating a select a security action based on characterizations of data. Claim 10 recites selecting a security action from available options if the outlier score does not exceed the threshold. Claim 11 recites selecting a security action based on comparing the outlier score to a set of threshold values comprising the threshold value. Claim 12 recites selecting security actions based on unspecified user order associated values. These limitations further narrow how the abstract ideas may be performed but do not make the claim any less abstract.
The claims do not provide any new additional limitations or meaningful limits beyond abstract ideas that are not addressed above in the independent claims. Therefore, they do not integrate the abstract idea into a practical application nor do they provide significantly more to the abstract idea. Thus, after considering all claim elements, both individually and as a whole, it has been determined that the claims do not integrate the judicial exception into a practical application or provide an inventive concept. Therefore, claims 2-12 and 14-20 are ineligible.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
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 nonobviousness.
Claims 1-9 and 13-20 are rejected under 35 U.S.C. 103 as being unpatentable over Stashluk, US20060149577A1 in view of Kansal, et al., hereinafter “Kansal,” “Customer Segmentation Using K-Means Clustering.” 2018 International Conference on Computational Techniques, Electronics and Mechanical Systems (CTEMS), Dec. 2018, pp. 135–139, https://doi.org/10.1109/ctems.2018.8769171., and in further view of Ford, US20190036971A1.
Regarding Claims 1, 13, and 20, Stashluk discloses:
receiving an appeasement request from a user device corresponding to a user of an online concierge system, wherein the appeasement request identifies an order associated with the user for which the user requests that an appeasement be applied; [0057] (user device used to send and receive transaction communications), [0007] “provision of…customer-specific, or item-specific return services…to provide rules-based returns processing that are initiated by the occurrence and detection of a triggering-event” (receive a request for order conciliation), [0020] (the request identifies user’s order);
computing an appeasement request rate for the user based on a set of prior appeasement requests from the user; [0058] (real-time and batch accumulation of individual and aggregated transaction and customer information, where an appeasement request is merely a characterization of a transaction), and [0046] “The extraction, transfer, and loading of information may result in return analytics such as the percentage of customers returning items, the frequency of returns on a per-customer basis, the frequency of returns on a per-item basis, or other analytics useful to a remote retailer 304 in determining general or specific customer satisfaction and retention;”
accessing user data for a set of candidate users, wherein the set of candidate users comprises the user; [0049] “The extraction, transfer, and loading of information may result in return analytics such as the percentage of customers returning items, the frequency of returns on a per-customer basis, the frequency of returns on a per-item basis, or other analytics useful to a remote retailer 304 in determining general or specific customer satisfaction and retention,” and [0051] (databases contain data for individuals and sets of individuals together);
generating a set of user segments based on the set of candidate users, wherein a user segment comprises a distinct subset of users within the set of candidate users, and wherein generating the set of user segments comprises: [0010] and [0020] “In particular embodiments, return analytics may be gathered on a customer or a group of customers. The return analytics may enable a remote retailer providing goods to customers on the Internet, by catalog, or by phone to implement segment-based marketing programs, offer dynamic pricing and/or personalized promotions, and improve customer retention and satisfaction.”
applying user segmentation to the user data for the set of candidate users to generate a segmentation score for each candidate user in the set of candidate users, to generate a segmentation score for a user based on user data describing that user, and wherein a segmentation score for a user represents the user segment to which the user belongs; and [0010] and [0020] “In particular embodiments, return analytics may be gathered on a customer or a group of customers. The return analytics may enable a remote retailer providing goods to customers on the Internet, by catalog, or by phone to implement segment-based marketing programs, offer dynamic pricing and/or personalized promotions, and improve customer retention and satisfaction,” [0077] (customers are categorized and segmented based on transaction analytics which include individual and group return rates that incorporate at least user data, order data, or data characterizations, and may influence future return decisions) (Examiner note: the claim does not limit what metrics are evaluated to compute the appeasement request rate as long as the rate is based on prior appeasement requests therefore the prior art’s rate is being utilized to represent score);
grouping the set of candidate users into the set of user segments based on the segmentation score for each candidate user; [0010] and [0020] “In particular embodiments, return analytics may be gathered on a customer or a group of customers. The return analytics may enable a remote retailer providing goods to customers on the Internet, by catalog, or by phone to implement segment-based marketing programs, offer dynamic pricing and/or personalized promotions, and improve customer retention and satisfaction,” [0077] (customers are categorized and segmented based on transaction analytics which include individual and group return rates that incorporate at least user data, order data, or data characterizations, and may influence future return decisions) (Examiner note: the claim does not limit what metrics are evaluated to compute the appeasement request rate as long as the rate is based on prior appeasement requests therefore the prior art’s rate is being utilized to represent score),
identifying a user segment of the set of user segments of which the user is a member; [0047] (The system can identify user and the segment that the user is in to perform individual actions)
wherein applying the appeasement action comprises transmitting to cause a notification through the user device, wherein the notification comprises text describing the appeasement action. [0081-0082 and Table 2] (deliver communications to user, including the communications listed in [Table 2]);
generating an appeasement model based on users in the identified user segment, wherein the appeasement model is a model representing appeasement request rates of users in the identified user segment; [0047] (system that applies retailer-specific, customer-specific, and returns server-specific rules, i.e. a model, to create segmentation and categorizations of user data, order data, return data, and characterizations of data, models that perform instructions based on different rules for tailoring customer experiences according to metrics and analytics), [0077] (user data regarding purchases and returns are used to categorize customers based on analytics rate calculations including return request rate, i.e. appeasement request rate), [0088] “reported return analytics may be organized by logical customer groupings,” [0088] “Return analytics may also be performed to provide reports describing customer behavior ... Returns server 306 may profile the customer transaction data … to determine a systemic linkage between return behavior and purchase behavior;”
based on an appeasement model, a user and other users within user segments, appeasement request rate associated with the user; [0047] (system that applies retailer-specific, customer-specific, and returns server-specific rules, i.e. a model), [0077] (user data regarding purchases and returns are used to categorize customers based on analytics rate calculations including return rate, i.e. appeasement request rate, associated with a user), [0088] “reported return analytics may be organized by logical customer groupings,” (i.e. customer segments), [0088] “Return analytics may also be performed to provide reports describing customer behavior ... Returns server 306 may profile the customer transaction data … to determine a systemic linkage between return behavior and purchase behavior,” (i.e. determining appeasement request rates for a user);
comparing the score to a threshold value; and [0154] “generate a rating score for each candidate object. The objects can then be ranked by their scores, and the highest scoring set of X objects, … deliver to a recommendation recipient …. In some embodiments, scoring thresholds may be set and used in addition to just relative ranking of the candidate objects,” [0236] “the Ranking Value Range is the indexed attribute values,” [0241] “thresholds are illustrated in this example, … thresholds may be applied as required for best results.”
While Stashluk does not disclose, Kansal Teaches:
wherein the user segmentation model is a machine learning model trained to generate a segmentation score. [Page (pg.) 135, Abstract] “The process of segmenting the customers with similar behaviours into the same segment and with different patterns into different segments is called customer segmentation... This is where machine learning comes into play, various algorithms are applied for unravelling the hidden patterns in the data for better decision making … is been trained by applying… onto a dataset;” and [pg. 135 Introduction] “ Clustering comes under unsupervised learning … There are a number of clustering algorithm …like k-means,” and [pg. 139, Silhouette Score] “It is a way of measuring how well the data point has been clustered into the correct cluster;”
It would be obvious to a person having ordinary skill in the art to combine the disclosures of the prior art before the filing date of the instant application because Stashluk and Kansal are related by the fields of computer science, more specifically data science systems for processing and grouping customer data using machine learning models, more specifically, user segmentation models. Stashluk and Kansal both utilize clustering via the segmentation model, i.e. algorithmic and statistical methods performed by machine learning models, large language models, and neural networks, which include training of models and implementation of trained models to cluster customer data in order to manage customer interaction within an e-commerce environment. Kansal specifically utilizes unsupervised -k means clustering. Both Stashluk and Kansal disclose methods of analyzing customer data to compute rates and scores, compare scores, and group users according to data and characteristics of said data. The analytics and machine learning models and methods of the prior art combine to yield the predictable results of the applicant’s disclosure and patent application claims, which lead to a markedly advantageous system of Managing Appeasement Requests Using User Segmentation.
Where Stashluk does not disclose and Kansal does not teach, Ford Teaches:
generating an outlier score, wherein the outlier score represents a likelihood that a random [data within a data group] would have [the same value as the initial data]; (where Stashluk discloses the [0085]-[0088] (computing a risk score, i.e. outlier score, based on observed events, i.e. appeasement requests as disclosed in Stashluk, a probability distribution of a sample is compared to a sample space, i.e. the user’s request rate as compared to other user’s request rates in the the same user segment, [0089-0090] (a probability mass function utilizes probability distributions to determine the likelihood or uniklihood of the probability that a random variable, i.e. the other random users’ request rate {from Stashluk}, is equal to a value of the user’s request rate {from Stashluk}, where the outlier score is the probability);
comparing the outlier score to a threshold value; [0088] “the resulting risk score may be assigned,” [0090] (outlier score is a design choice and could be the outlier score), [0092] “the sample may be the occurrence of a feature associated with a corresponding event. As used herein, a feature, as it relates to an event, broadly refers to a property, characteristic or attribute of a particular event,” [0101] “configured events … may be processed by the feature matching,” [0150], where [0162], [0163] “security policy elements may include a rule and an associated action, where the rule defines the criteria by which the security policy is enforced, and the action describes the corresponding response should the bounds of the rule be met or exceeded. … the rule may include an event, an allowable behavior, or a combination thereof,” [0166], [0182];
based on the comparing, applying an appeasement action to the user, wherein applying the appeasement action comprises transmitting instructions to the user device associated with the user, the user device associated with the user to display through a graphical user interface of the user device. [0163] “As used herein, a security policy broadly refers to a set of security policy elements providing the criteria by which the security policy is enforced, and one or more associated responses should it be violated,” (threshold and action according to the threshold), [0164], [0166], [0182], [0025] (a display device, i.e. a graphical user interface of a user device), [0057-0058] (transmit instructions to an end user device to update contextual information), [0215] (instructions are implemented by the processor to cause the processor to perform the functions or acts specified);
It would be obvious to a person having ordinary skill in the art to combine the disclosures of the prior art before the filing date of the instant application because Stashluk, Kansal, and Ford are related by the fields of computer science, more specifically data science systems for processing and grouping customer data. Stashluk and Kansal are also related by clustering via algorithmic and statistical methods performed by machine learning models, large language models, and neural networks including training of models to cluster customer data in order to manage customer interaction with e-commerce, Kansal specifically utilizes unsupervised -k means clustering. Ford is related to Stashluk and Kansal as it discloses clustering, but employs managing customer data, interactions and requests via machine learning models including probability mass functions that comprise algorithmic and statistical methods. All three disclose methods of analyzing customer data to compute rates and scores, compare scores to group users according to data and characteristics of said data. Stashluk and Ford further utilize alternate models for making response decision based on user data analytics. Ford expands on this by implementing ranges of scores based off of analytics to identify the appropriate response actions to customer requests. Ford further implements specific security scoring features obvious to those recited in the instant application. The analytics and machine learning models and methods of the prior art combine to yield the predictable results of the applicant’s disclosure and patent application claims, which lead to a markedly advantageous system of Managing Appeasement Requests Using User Segmentation.
Regarding Claims 2 and 14: Stashluk discloses, and Kansal and Ford teach: The method of claim 1,
Stashluk discloses: wherein computing the appeasement request rate comprises computing a ratio of a number of appeasement requests received from the user to a number of orders placed by the user. [0088] “Return analytics may also be performed to provide reports describing customer behavior ... Returns server 306 may profile the customer transaction data … to determine a systemic linkage between return behavior and purchase behavior.”
Regarding Claims 3 and 15: Stashluk discloses, Kansal teaches, and Ford further teaches: The method of claim 1,
Stashluk discloses: wherein accessing user data for a set of candidate users comprises identifying the set of candidate users based on order data describing characteristics of orders placed by the set of candidate users. [0049] “The extraction, transfer, and loading of information may result in return analytics such as the percentage of customers returning items, the frequency of returns on a per-customer basis, the frequency of returns on a per-item basis, or other analytics useful to a remote retailer 304 in determining general or specific customer satisfaction and retention,” and [0051] (databases contain data for individuals and sets of individuals together along with order data and characteristics of order data), and [0053] (order data and characteristics of order data).
Regarding Claim 4 and 16: Stashluk discloses, and Kansal and Ford teach: The method of claim 3, wherein orders placed by the set of candidate users were placed from a same geographic region as the order identified by the appeasement request.
Stashluk discloses: [0096] “the categorization of customer 302 relative to other customers… may take into account customer location;” (Examiner note: while this paragraph is pertinent to pricing and shipping, within context, the instant application and the prior art similarly utilize geographic comparisons of customers to drive post order conciliation actions).
Regarding Claims 5 and 17: Stashluk discloses, and Kansal and Ford teach: The method of claim 1,
Kansal teaches: wherein the user segmentation model is an unsupervised machine learning model. [Page (pg.) 135, Abstract] “The process of segmenting the customers with similar behaviours into the same segment and with different patterns into different segments is called customer segmentation... This is where machine learning comes into play, various algorithms are applied for unravelling the hidden patterns in the data for better decision making … is been trained by applying… onto a dataset;” and [pg. 135 Introduction] “ Clustering comes under unsupervised learning … There are a number of clustering algorithm …like k-means.”
Regarding Claims 6 and 18: Stashluk discloses, and Kansal and Ford teach: The method of claim 1,
Kansal teaches: wherein the user segmentation model uses k-means clustering to generate the user segments. [Page (pg.) 135, Abstract] “The process of segmenting the customers with similar behaviours into the same segment and with different patterns into different segments is called customer segmentation... This is where machine learning comes into play, various algorithms are applied for unravelling the hidden patterns in the data for better decision making … is been trained by applying… onto a dataset;” and [pg. 135 Introduction] “ Clustering comes under unsupervised learning … There are a number of clustering algorithm …like k-means,” and [pg. 139, Silhouette Score] “It is a way of measuring how well the data point has been clustered into the correct cluster,”
Regarding Claims 7 and 19: Stashluk discloses, and Kansal and Ford teach: The method of claim 1,
Ford teaches: wherein the appeasement model includes a probability mass function, wherein the probability mass function gives a probability that a random variable is equal to a value based on a probability distribution. [0089-0090] (a probability mass function utilizes probability distributions to determine the likelihood or uniklihood of the probability that a random variable is equal to a value).
Regarding Claim 8: Stashluk discloses, Kansal teaches, and Ford further teaches: The method of claim 1,
Stashluk discloses: wherein comprises comparing the appeasement request rate of the user to an average appeasement request rate of users in the identified segment. [0088] “reported return analytics may be organized by logical customer groupings, [0088] “Return analytics may also be performed to provide reports describing customer behavior ... Returns server 306 may profile the customer transaction data … to determine a systemic linkage between return behavior and purchase behavior;
Ford teaches: wherein generating the outlier score comprises comparing the appeasement request rate of the user to an average appeasement request rate of users in the identified segment generating the outlier score [0089-0090] (a probability mass function utilizes probability distributions to determine the likelihood or uniklihood of the probability that a random variable is equal to a value).;
Regarding Claim 9: Stashluk discloses, and Kansal and Ford teach: The method of claim 1,
Stashluk discloses: wherein applying the appeasement action comprises selecting the appeasement action from a set of possible appeasement actions based on the appeasement request, wherein the set of possible appeasement actions includes redelivering the order and issuing a refund to the user. [0048] “identify the needs of core business logic 336, and provision the logic necessary to perform the return services requested by retailer 304 or a customer 302 of retailer 304,” [0054] “When communications are received from returns server 306 that indicate that a return is pending and/or authorized, accounting system 368 may be used in a similar manner to refund all or a portion of the customer's payment,” [0005] “there exists a myriad of issues with which an Internet customer must contend to return this remotely purchased merchandise. For example, these issues include whether the item is returnable, how will the refund be paid, what shipping agent should be used, how efficient is the long-distance returns process, and what happens if the package is lost or damaged in transit,” “information may be useful to returns server 306 for the provision of a number of returns services.”
(Examiner note: Claim 9 is an intended result that depends upon the possibility of an action occurring to return the results, therefore it carries no patentable weight.)
Claims 10 - 12 are rejected under 35 U.S.C. 103 unpatentable over Stashluk, US20060149577A1 in view of Kansal, et al., hereinafter “Kansal,” “Customer Segmentation Using K-Means Clustering.” 2018 International Conference on Computational Techniques, Electronics and Mechanical Systems (CTEMS), Dec. 2018, pp. 135–139, https://doi.org/10.1109/ctems.2018.8769171., in further view of Ford, US20190036971A1, and in further view of Jia, US20200134628A1.
Regarding Claim 10: Stashluk discloses, and Kansal and Ford teach: The method of claim 1 further comprising:
Where Stashluk does not disclose, and Kansal and Ford do not teach, Jia teaches: responsive to the outlier score not exceeding the threshold value, selecting a security action from a set of possible security actions; and applying the security action to the user. [0068] “if the estimated re-presentment score is higher than the high cutoff threshold, the merchant control action prediction service 141 assigns a “won” label, and if the estimated re-presentment score is lower than the low cutoff threshold, the merchant control action prediction service 141 assigns a “not won” label;”
It would be obvious to a person having ordinary skill in the art to combine the disclosures of the prior art before the filing date of the instant application because Stashluk, Kansal, Ford, and Jia are related by the fields of computer science, more specifically data science systems for processing and grouping customer data. Stashluk and Kansal are also related by clustering via algorithmic and statistical methods performed by machine learning models, large language models, and neural networks including training of models to cluster customer data in order to manage customer interaction with e-commerce, Kansal specifically utilizes unsupervised -k means clustering. Jia utilizes machine learning models and scoring to manage commerce transaction options for security via probability, similar to Ford. They are is related to Stashluk and Kansal as Ford briefly discloses clustering, while Jia and Ford employ managing customer data, interactions and requests via machine learning models including probability or probability mass functions that comprise algorithmic and statistical methods. where all four disclose scoring users and user actions, comparing scores, and Stashluk, Jia, and Ford further utilize ranges of values to identify response actions to customer requests. Ford and Jia further implement specific security scoring features obvious to those recited in the instant application. The machine learning models and methods of the prior art combine to yield the predictable results of the applicant’s disclosure and patent application claims, which lead to a markedly advantageous system of Managing Appeasement Requests Using User Segmentation.
Regarding Claim 11: Stashluk discloses, and Kansal and Ford teach: The method of claim 10,
Where Stashluk does not disclose, and Kansal and Ford do not teach, Jia teaches: wherein selecting the security action comprises comparing the outlier score to a set of threshold values, wherein the set of threshold values comprises the threshold value. [0028] “the routing decision may be output by a routing model as a routing score, which can be compared to one or more routing thresholds used to decide which processor (or gateway) to select to send this transaction to be processed to increase a settle rate;”
It would be obvious to a person having ordinary skill in the art to combine the disclosures of the prior art before the filing date of the instant application because Stashluk, Kansal, Ford, and Jia are related by the fields of computer science, more specifically data science systems for processing and grouping customer data. Stashluk and Kansal are also related by clustering via algorithmic and statistical methods performed by machine learning models, large language models, and neural networks including training of models to cluster customer data in order to manage customer interaction with e-commerce, Kansal specifically utilizes unsupervised -k means clustering. Jia utilizes machine learning models and scoring to manage commerce transaction options for security via probability, similar to Ford. They are is related to Stashluk and Kansal as Ford briefly discloses clustering, while Jia and Ford employ managing customer data, interactions and requests via machine learning models including probability or probability mass functions that comprise algorithmic and statistical methods. where all four disclose scoring users and user actions, comparing scores, and Stashluk, Jia, and Ford further utilize ranges of values to identify response actions to customer requests. Ford and Jia further implement specific security scoring features obvious to those recited in the instant application. The machine learning models and methods of the prior art combine to yield the predictable results of the applicant’s disclosure and patent application claims, which lead to a markedly advantageous system of Managing Appeasement Requests Using User Segmentation.
Regarding Claim 12: Stashluk discloses, Kansal and Ford teach: The method of claim 10,
Where Stashluk does not disclose, and Kansal and Ford do not teach, Jia teaches: wherein selecting the security action comprises selecting the security action based on a value associated with orders placed by the user. [0038] and [0030] “for evaluating and processing an online purchase transaction 125 associated with orders from a customer.”
It would be obvious to a person having ordinary skill in the art to combine the disclosures of the prior art before the filing date of the instant application because Stashluk, Kansal, Ford, and Jia are related by the fields of computer science, more specifically data science systems for processing and grouping customer data. Stashluk and Kansal are also related by clustering via algorithmic and statistical methods performed by machine learning models, large language models, and neural networks including training of models to cluster customer data in order to manage customer interaction with e-commerce, Kansal specifically utilizes unsupervised -k means clustering. Jia utilizes machine learning models and scoring to manage commerce transaction options for security via probability, similar to Ford. They are is related to Stashluk and Kansal as Ford briefly discloses clustering, while Jia and Ford employ managing customer data, interactions and requests via machine learning models including probability or probability mass functions that comprise algorithmic and statistical methods. where all four disclose scoring users and user actions, comparing scores, and Stashluk, Jia, and Ford further utilize ranges of values to identify response actions to customer requests. Ford and Jia further implement specific security scoring features obvious to those recited in the instant application. The machine learning models and methods of the prior art combine to yield the predictable results of the applicant’s disclosure and patent application claims, which lead to a markedly advantageous system of Managing Appeasement Requests Using User Segmentation.
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 ANGELA HATCH whose telephone number is (571)270-1393. The examiner can normally be reached 10:00-6:00.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Nathan Uber can be reached at (571)270-3923. 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.
ANGELA HATCH
Examiner
Art Unit 3626
/ANGELA HATCH/ Examiner, Art Unit 3626
/KIERSTEN V SUMMERS/ Primary Examiner, Art Unit 3626