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 .
This communication is in response to Applicant’s Arguments/Remarks made in an amendment filed 3/2/2026 regarding application 18/086,394 filed 12/21/2022. Claims 1, 3-11, and 13-20 are amended and hereby entered.
Response to Arguments
Applicant's arguments filed 3/2/2026 are fully considered but they are not persuasive.
Regarding 35 USC 101:
The applicant submits the claims do not recite a mental process because they cannot be practically performed in the human mind. Specifically, the applicant cites the MPEP examples of claims that do not recite a mental process stating, “a claim to a specific data encryption method for computer communication involving a several-step manipulation of data, Synopsys., 839 F.3d at 1148, 120 USPQ2d at 1481 (distinguishing the claims in TQP Development, LLC v. Intuit Inc., 2014 WL 651935 (E.D. Tex. 2014))”. However, the claims are not directed towards a specific data encryption method. The claims are directed to fraud detection, in which the method recites receiving data, processing the data at a high level (with general purpose computing components and machine learning model), and presenting the result of analysis. The claims perform a mental process in a computer environment, see MPEP 2106.04(a)(2)(III)(C) stating, “Performing a mental process in a computer environment… The patentee in FairWarning claimed a system and method of detecting fraud and/or misuse in a computer environment, in which information regarding accesses of a patient’s personal health information was analyzed according to one of several rules (i.e., related to accesses in excess of a specific volume, accesses during a pre-determined time interval, or accesses by a specific user) to determine if the activity indicates improper access. 839 F.3d. at 1092, 120 USPQ2d at 1294. The court determined that these claims were directed to a mental process of detecting misuse, and that the claimed rules here were "the same questions (though perhaps phrased with different words) that humans in analogous situations detecting fraud have asked for decades, if not centuries." 839 F.3d. at 1094-95, 120 USPQ2d at 1296”.
Further, the applicant submits the claims are integrated into a practical application because the claims recite a technical improvement. Specifically, the applicant recites the data processing steps solve technical problems associated with fraud detection and prevention. However, the claims are using general purpose computing components and machine learning models recited at a high level to perform a commonplace business method of fraud detection with standard data processing steps, see MPEP 2106.05(f)(2). The claims recite the use of a computer to perform standard data processing steps such as encoding, processing, converting, and inputting, which is then used by a machine learning model to detect fraud. The claims as a whole do not describe a technical improvement to the underlying computer, technology, or technical field.
Additionally, the applicant improperly lists claim limitations as additional elements. Much of the applicant listed additional elements are considered a part of the abstract idea, not additional elements that can integrate the judicial exception into practical application or amount to significantly more. The office nonfinal action dated 12/01/2025 clearly states the additional elements as “a non-transitory computer readable storage medium, machine learning model, server computer system, and one or more processors”. Therefore, the applicant listed elements are part of the abstract idea and cannot amount to significantly more.
Further, the applicant submits the claims amount to significantly more because the combination operates in a non-conventional and non-generic way. Specifically, the applicant submits the examiner fails to form a prima facie case of subject matter eligibility because no evidence was provided that elements are “well-understood, routine, and conventional”. However, Berkheimer analysis is necessary in the presence of insignificant extra solution activity. The rejection does not rely on “well-understood, routine, and conventional activity” or “insignificant extra-solution activity”. The rejection relies on “mere instructions to apply an exception”, see MPEP 2106.05(f). Therefore, the applicants’ arguments are not persuasive, and the rejection is maintained.
Regarding 35 USC 103:
The applicant submits prior art Agrawal only teaches deactivation, rather than deactivation in response to detecting that a fraud score being a first value and that the score is a second value less than the first value. The applicant discusses paragraph 70 failing to teach this limitation because it only teaches deactivation. However, both paragraphs 73 and 72 were also cited to teach these limitations. Paragraph 73 of prior art Agrawal states, “The generated scores may be used for, and individually trigger, a number of computer-driven services 225 for resolving merchant breach”, and paragraph 72 of prior art Agrawal states, “Each model 218, 220 may output a score that, when compared to a predetermined threshold value at a given false positive tolerance level, indicates whether or not a merchant has been breached”. The prior art, as a whole, teaches the deactivation happening in response to one score being less than another score.
Further, the applicant submits tertiary prior art Lee does not disclose feedback being indicative of a confirmed lack of fraud. Rather, the applicant submits the prior art only teaches verification of records rather than confirmation of lack of fraud. However, the claims are interpreted under broadest reasonable interpretation. One of ordinary skill could interpret confirmation and verification to be synonymous. Further, fraud detection is already taught by prior art Agrwal, and the rejection is based on the combination of references, not one reference individually. Hence, it is the combination of references that teaches the limitations described by the applicant. One cannot show nonobviousness by attacking the reference individually where the rejections are based on a combination of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
Additionally, the applicant submits tertiary prior art Lee does not teach feedback being uploaded subsequent to the user system receiving the fraud detection message. Rather, the applicant submits the prior art only teaches feedback without disclosing a fraud detection message. However, the fraud detection message is already taught by paragraph 40 of Agrawal which is cited previously in the rejection. Further, the rejection states, “One of ordinary skill would have recognized that incorporating feedback on the accuracy of fraud detection would be beneficial in training a machine learning model”. It is the combination of references that teaches the limitations described by the applicant. One cannot show nonobviousness by attacking the reference individually where the rejections are based on a combination of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986). Therefore, the applicant’s arguments are unpersuasive, and the rejection is maintained.
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.
Claim 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) with no practical application and without significantly more.
Claims 1-10 are method claims. Claims 11- 20 are systems claims. Thus, each claim on its face is directed to one of the statutory categories of 35 USC 101. However, claims 1-20 are rejected under 35 USC 101 because the claimed invention is directed to an abstract idea without significantly more.
The independent claims (1, 11, and 20) recite a method and systems to detect fraudulent merchant activities. These claim elements are being interpreted as concepts performed in the human mind (including observation, evaluation, judgement, and opinion). Using data to determine if transactions are fraudulent can equivalently be done by pen and paper and vulnerabilities could be detected through human observation and evaluation of data. Therefore, these steps fall under the abstract idea of mental processes or concepts performed in the human mind.
The instant application fails to integrate the judicial exception into a practical application because the instant application merely recites an “apply it” (or an equivalent) with the judicial exception or merely includes instructions to implement an abstract idea. The instant application is directed towards a method and systems to implement the identified abstract idea of receiving information, processing information, and displaying the result of the analysis (i.e. gathering, processing and inputting data to detect fraud and the like) on a generically claimed computer structure. The claims do not include additional elements that amount to significantly more than the judicial exception. The independent claims recite the additional elements: “at least one processor”, “Machine Learning Model”, “A non-transitory computer readable storage medium… computer processing system”, and “A server computer system… and one or more processors”. These claim elements are recited at a high level of generality such that it amounts to no more than mere instructions to apply the exceptions using a general computer environment. The machines merely act as a modality to implement the abstract idea and are not indicative of integration into a practical application (i.e. the additional elements are simply used as a tool to perform the abstract idea), see MPEP 2106.05(f).
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed in Step 2A Prong Two analysis, the additional elements in the claims amount to no more than mere instructions to apply the exception using generic computer components. The same analysis applies here in 2B and does not provide an inventive concept.
Regarding the dependent claims:
Claim 2 limits the scope by including an acceptable combination of machine learning models. As previously stated, applying the additional element of a machine learning model is not indicative of a practical application, see MPEP 2106.05(f). Claim 12 mirrors claim 2 and is analyzed in the same manner as above
Claim 3-10 and 13-19 describe no new abstract ideas or new additional elements and do not impact analysis under 35 USC 101
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-4, 6-14, and 16-20 are rejected under 35 U.S.C. 103 as being unpatentable over Agrawal (US-20210279731-A1) in view of Ares (US 20200234154 A1) in further view of Lee (US 20220237618 A1).
Regarding claims 1, 11 and 20, (substantially similar in scope and language) Agrawal teaches:
A method for detecting fraudulent activities at a server computer system, , by at least one processor, a first set of different from the first set, [(Para 0006) “receiving, with at least one processor, transaction data associated with a plurality of transactions between at least one financial device holder and at least one merchant”, (Abstract) “The method also includes generating a first model input dataset associated with the at least one merchant and a second model input dataset associated with the at least one previously identified data-breach incident”] the two sets of data generated for a user , by the at least one processor the first set of , by the at least one processor the second set of [The limitations recite multiple data sets; (Para 0005) “receiving fraudulent transaction data, and generating a first model input dataset and a second model input dataset]
, by the at least one processor, the set of … data input signals into a standardized set of … data input signals and the set of … data input signals into a standardized set of…data input signals [The limitations recite the conversion of data into a “standardized” set; (Para 0057) “Additionally, two units may be in communication with each other even though the data transmitted may be modified, processed, relayed, and/or routed between the first and second unit… As another example, a first unit may be in communication with a second unit if an intermediary unit processes data from one unit and transmits processed data to the second unit. It will be appreciated that numerous other arrangements are possible”, (Para 0019) “generating, with at least one processor and based at least partly on the transaction data and the fraudulent transaction data, a first model input dataset associated with the at least one merchant and a second model input dataset associated with at least one merchant of the at least one previously identified data-breach incident”, (Para 0068) “The modeling and detection server 112 receives a number of data sources to build model input datasets, e.g., profiles, to train machine learning prediction models”]
inputting, by at least one processor, the standardized set of structure data input signals into a first machine learning model[[,]] and [(Para 0019) “generating, with at least one processor and based at least partly on the transaction data and the fraudulent transaction data, a first model input dataset associated with the at least one merchant and a second model input dataset associated with at least one merchant of the at least one previously identified data-breach incident; training, with at least one processor and based at least partly on a comparison of the first model input dataset to the second model input dataset”, (Para 0068) “feature datasets of merchant transaction activity may be input to the predictive models to determine likelihoods of merchant breach”] generating, by at least one processor, a single fraud score by combining a first fraud score generated by the first machine learning model with a second fraud score generated by the second machine learning model; [(Para 0072) “The ensemble scoring algorithm 222 may combine the individual scores of two or more models”]
subsequent by the at least one processor, an account associated with the subsequent by the at least one processor, the ; and [(Para 0070) “The alert may also be a notification that the financial device holder's 102 financial device 104 has been deactivated, e.g., suspended from engaging in financial transactions”, (Para 0073) “The generated scores may be used for, and individually trigger, a number of computer-driven services 225 for resolving merchant breach. For example, the scores may be communicated through an application programming interface (API) 226 to a security personnel that is tasked with alerting merchants”. (Para 0072) “Each model 218, 220 may output a score that, when compared to a predetermined threshold value at a given false positive tolerance level, indicates whether or not a merchant has been breached.”, (Para0070) “This communication may include additional action options for the financial device holder 122, such as … freeze future transactions,”]
subsequent , by the at least one processor, one or more remediative actions against the account [(Para 0072) “the combined score of both models may be compared to a composite threshold, which may indicate whether or not a merchant has been breached” (Para 0006) “The method further includes generating, with at least one processor, a communication configured to cause at least one action to be taken in response to the determination of the at least one breached merchant.”]
wherein initiating the one or more remediative actions comprises: sending, by the at least one processor, a fraud detection message to a user system, associated with the account with data indicative of the single fraud score, [(Para 0073) “For example, the scores may be communicated through an application programming interface (API) 226 to a security personnel that is tasked with alerting merchants”]
and annotating, by the at least one processor, the set of structured numeric data input signals and the set of unstructured text data input signals with the feedback data as annotated service request data. [(Para 0071) “Data sources include: (i) historic breach data 202, including historical support requests from issuer institutions… fraudulent transaction data, and/or the like; (ii) transaction authorization request data 204, including transaction type, transaction amount, transaction description, merchant identifier, merchant location, transaction time, portable financial device identifier, PAN information, automatic ratings from security systems estimating the riskiness of the transaction, and/or the like…”, see also (Para 0075, Para 0068, Figure 2)]
While Agrawal teaches the limitations set forth above, the use of two sets of data in the machine models, and processing data into a standardized format, it does not explicitly teach:
…a set of structured numeric data input signals
…a set of unstructured text data input signals
…structured numeric…
…unstructured text…
receiving, from the user system, feedback data indicative of a confirmed lack of fraud associated with subsequent to the user system receiving the fraud detection message,
However, Ares teaches:
…a set of structured numeric data input signals… a set of unstructured text data input signals; [(Para 0009) “a method of navigating structured and unstructured data using a relational computer model” (Para 0114) “When data corresponding to a node is unstructured (not readily identifiable as associated with a certain data table), the analytics server may also use artificial intelligence and machine-learning techniques to revise the nodal network and identify a node for the collected data… In some configurations, the AI model may incorporate other machine learning techniques, such as gradient boosting, support vector machines, deep neural networks, and logistic regression”, (Para 0312) “The user may designate the data type (e.g., the data type, illustrated in row 3, can be either string or numeric).”]
structured numeric… unstructured text… [(Para 0009) “a method of navigating structured and unstructured data using a relational computer model”, (Para 0312) “The user may designate the data type (e.g., the data type, illustrated in row 3, can be either string or numeric).”]
Therefore, it would be obvious for one of ordinary skill in the art before the effective filing date to combine the use of machine learning to detect fraud taught by Agrawal, with the techniques using structured and unstructured data in Ares. Ares discusses an example of a data source being paid services data or financial data. Combining the data techniques taught by Ares with the transaction data taught by Agrawal would yield predictable results. Furthermore, Agrawal cites the use of gradient boosting, and neural networks as an example of two machine learning models. These models are also cited in Ares and are associated with structured and unstructured data.
While Agrawal in view of Ares teach a fraud detection method using different types of data, they do not explicitly teach:
receiving, from the user system, feedback data indicative of a confirmed lack of fraud associated with subsequent to the user system receiving the fraud detection message,
However, Lee teaches:
receiving, from the user system, feedback data indicative of a confirmed lack of fraud associated with subsequent to the user system receiving the fraud detection message, [(Para 0040) “ As described herein, the association model may be a supervised machine learning model and/or may include a supervised machine learning model that is configured to learn to detect associations based on feedback (e.g., from a user and/or a user device… The account management system may provide the prediction to the user device to prompt a user of the user device to provide the feedback (e.g., via a user input) that verifies an accuracy of the prediction (e.g., verifies whether the prediction of an association between a mitigation record and an event record is accurate)”]
Therefore, it would be obvious for one of ordinary skill in the art before the effective filing date to combine the use of machine learning to detect fraud taught by Agrawal in view of Ares, with the use of feedback from users taught by Lee. One of ordinary skill would have recognized that incorporating feedback on the accuracy of fraud detection would be beneficial in training a machine learning model.
Regarding Claims 2 and 12, The combination of Agrawal, Ares, and Lee teach the limitations set forth above, Agrawal further teaches:
The method of claim 1, wherein the first machine learning model comprises an XGBoost machine learning model, and the second machine learning model comprises a Neural Network machine learning model. [(Para 0008) “The first machine-learning prediction model may be a fully connected neural network, and the second machine-learning prediction model may be a gradient boosted decision tree.”]
Regarding Claims 3 and 13, The combination of Agrawal, Ares, and Lee teach the limitations set forth above, Agrawal further teaches:
The method of claim 1, further comprising: accessing, by the at least one processor, a third set of time-based event data associated with time-based events occurred at the server computer system, [(Para 0010) “the transaction data includes, for each transaction of the plurality of transactions, an authorization request, a portable financial device identifier, and at least one of the following: transaction amount, transaction time, transaction type, merchant identifier, merchant type, or any combination thereof.”] the time-based events comprising application programming interface (API) events having associated timing data; [(Para 0071) “transaction authorization request data 204, including transaction type, transaction amount, transaction description, merchant identifier, merchant location, transaction time… fraud report data 208, including financial device identifier(s), report date, report time, merchant identifier, transaction data, and/or the like”, (Para 0073) “For example, the API may be used to display a breach investigation platform that visually represents the merchants (e.g., in a sortable table) with respect to case status, merchant region, detection date, merchant level, merchant type, predictive model breach score, breach date, total number of PANs associated with transactions with the merchant in a prior time period (e.g., 180 days),”]
encoding, by the at least one processor, the third set of time-based event data into a set of time-based event input signals; [(Para 0071) “The method 200 includes a model training process 201, in which data from multiple data sources are used to create prediction model input feature datasets, e.g., model training profiles. Data sources include: (i) historic breach data 202… time periods of past breaches…transaction time…The modeling data 212, taken from the data sources, may be used to create a number of features, e.g., feature vectors, to form model input datasets, e.g., training profiles.”]
converting, by the at least one processor, the set of time-based event input signals into a standardized set of time-based event input signals; [(Para 0057) “Additionally, two units may be in communication with each other even though the data transmitted may be modified, processed, relayed, and/or routed between the first and second unit… As another example, a first unit may be in communication with a second unit if an intermediary unit processes data from one unit and transmits processed data to the second unit. It will be appreciated that numerous other arrangements are possible”]
inputting, by the at least one processor, the standardized set of time-based event input signals into a third machine learning model, wherein the third machine learning model is trained to detect fraudulent activities based on the standardized set of time-based event input signals; [(Para 0019) “generating, with at least one processor and based at least partly on the transaction data and the fraudulent transaction data, a first model input dataset associated with the at least one merchant and a second model input dataset associated with at least one merchant of the at least one previously identified data-breach incident; training, with at least one processor and based at least partly on a comparison of the first model input dataset to the second model input dataset”, (Para 0068) “feature datasets of merchant transaction activity may be input to the predictive models to determine likelihoods of merchant breach.” (Para 0072) “For example, a fully connected deep neural network 218 may be ensembled with gradient boosted decision trees”]
generating, by the at least one processor, the single fraud score [(Para 0072) “Once trained, the machine learning prediction models 218, 220, either individually or as an ensemble scoring algorithm 222, may be used to assign a breach score to individual merchants… The ensemble scoring algorithm 222 may combine the individual scores of two or more models”]
Regarding claim 4 and 14, The combination of Agrawal, Ares, and Lee teach the limitations set forth above, Agrawal further teaches:
wherein the numeric data associated with the set of structured numeric data input signals comprises one or more of an amount associated with the user system service request, a total number of service requests associated with the user system, a number of declines associated with the user system, and a number of fraud detections associated with the user system, and wherein the text data associated with the set of unstructured data input signals comprises one or more of an email address, a first name associated with the user system, a last name associated with the user system, an internet protocol address of the user system, and a country of the user system. [(Para 0006) “The transaction data includes… and at least one of the following: transaction amount, transaction time, transaction type, merchant identifier, merchant type, or any combination thereof”.]
Regarding Claims 6 and 16, The combination of Agrawal, Ares, and Lee teach the limitations set forth above, Agrawal further teaches:
wherein initiating the one or more remediative actions further comprises: declining, by the at least one processor, the user request; deactivating, by the at least one processor, the account associated with the user system; or a combination thereof. [(Para 0006) “The method further includes generating, with at least one processor, a communication configured to cause at least one action to be taken in response to the determination of the at least one breached merchant... deactivating at least one portable financial device used for at least one prior transaction”.]
Regarding Claims 7 and 17, The combination of Agrawal, Ares, and Lee teach the limitations set forth above, Agrawal further teaches:
wherein initiating the one or more remediative actions further comprises: storing, by the at least one processor, the annotated service request data with a set of annotated training data in a training data store. [(Para 0075) “The transaction data may be stored in a transaction database and used for further fraud analysis.” (Para 0068) “Historic transaction data from known data breach events may be used to train the prediction models and to evaluate the performance of the prediction models”, (Figure 2)]
Regarding Claims 8 and 18, The combination of Agrawal, Ares, and Lee teach the limitations set forth above, and Agrwal further teaches:
using, at least in part on the set of annotated training data [(Para 0068) “Historic transaction data from known data breach events may be used to train the prediction models”, (Para 0068) “Once trained, feature datasets of merchant transaction activity may be input to the predictive models to determine likelihoods of merchant breach”]
and deploying, by the at least one processor, the first retrained machine learning model to detect fraudulent activities based on a further standardized set of structured data input signals, and the second retrained machine learning model to detect fraudulent activities based on a further standardized set of unstructured data input signals. [(Para 0072) “Once trained, the machine learning prediction models 218, 220, either individually or as an ensemble scoring algorithm 222, may be used to assign a breach score to individual merchants”]
While Agrawal in view of Ares teaches training the machine learning model with annotated training data, it does not explicitly teach:
periodically retraining, by the at least one processor, the first models
However, Lee teaches:
periodically retraining, by the at least one processor, the first models [(Para 0043) “As further shown in FIG. 1E, and by reference number 155, the account management system may retrain the association model based on the feedback”]
Therefore, it would be obvious for one of ordinary skill in the art before the effective filing date to perform retraining of a machine learning model taught by Lee, on the machine learning model taught by Agrawal. The continual retraining of machine learning models based on the outputs and feedback of the model is obvious to one of ordinary skill, and the retraining of the models taught in Agrawal would yield predictable results.
Regarding Claims 9 and 19, The combination of Agrawal, Ares, and Lee teach the limitations set forth above
While Agrawal in view of Ares teaches generating a single fraud score, it does not explicitly teach:
wherein the single fraud score is generated in real time or near real time, and detecting that satisfies
However, Lee teaches:
wherein the single fraud score is generated in real time or near real time, and detecting satisfies [(Para 0031) “The account management system may monitor the record log for newly received mitigation records (e.g., by monitoring for transaction records associated with a refund or transaction records that have a negative value). Based on identifying or detecting a mitigation record, the account management system, using the association model, may compare one or more attributes (e.g., return attributes) associated with the mitigation record with one or more corresponding attributes (e.g., charge attributes) in the event records”]
Therefore, it would be obvious for one of ordinary skill in the art before the effective filing date to combine the near real time analysis taught by Lee, with the machine learning models taught by Agrawal in view of Ares. Monitoring new transactions and inputting real time data into the machine learning models taught by Agrawal would provide the most up to date information to determine scores yielding predictable results.
Regarding Claim 10, The combination of Agrawal, Ares, and Lee teach the limitations set forth above, Agrawal further teaches:
wherein the user [(Para 0075) (Para 0006) “The transaction data includes, for each transaction of the plurality of transactions, an authorization request, a portable financial device identifier, and at least one of the following: transaction amount, transaction time, transaction type, merchant identifier, merchant type, or any combination thereof”]
Claims 5 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over Agrawal (US-20210279731-A1) in view of Ares (US-20200234154-A1) in view of Lee (US 20220237618 A1) in further view of Edwards (US-20210182830-A1).
Regarding Claims 5 and 15, The combination of Agrawal, Ares, and Lee teaches the limitations set forth above
While the combination of Agrawal, Ares, and Lee teaches generating a single score by combining two machine learning model scores, it does not explicitly teach:
Calculating, by the at least one processor, an average of the first fraud score and the second fraud score.
However, Edwards teaches
calculating, by the at least one processor, an average of the first fraud score and the second fraud score. [(Para 0034) “For example, the trained machine learning model may determine multiple confidence scores…the multiple confidence scores satisfy a confidence threshold, such as based on an average of the multiple confidence scores, a weighted average of the multiple confidence scores, a combination of the multiple confidence scores, and/or the like.”]
Therefore, it would be obvious for one of ordinary skill in the art before the effective filing date to calculate the combination score taught by the combination of Agrawal, Ares, and Lee with the use of averaging taught by Edwards. Simply substituting a composite score/threshold taught by Agrawal, with the average score/threshold taught by Edwards would yield predictable results.
Conclusion
THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 Examiner Benjamin Truong, whose telephone number is 703-756-5883. The examiner can normally be reached on Monday-Friday from 9 am to 5 pm (EST).
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Nathan Uber SPE can be reached on 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.
/B.L.T. /Examiner, Art Unit 3626
/SANGEETA BAHL/Primary Examiner, Art Unit 3626
Prosecution Insights