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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on April 6, 2026, has been entered.
Acknowledgments
This Action is in response to the request for continued examination and the amendment filed on April 6, 2026. Claims 1-15 and 18-23 are currently pending and have been fully examined. Claims 16-17 have been cancelled by Applicant and claims 22-23 are newly added.
Response to Arguments
With respect to the claim objections, the amendment overcomes the issues, but new issues arise.
With respect to the 101 rejection, as discussed during the Interview conducted on March 24, 2026, and considering the recent court case of Desjardins, the amendment to the claim to recite training a learning model, overcomes the rejection and the rejection is withdrawn.
With respect to the 103 rejections, Applicant argues, on page 16 of remarks, that cited references do not disclose the recited features that the information includes characteristics of the actions on the predetermined service and includes location information, date and time information, and usage information. The examiner respectfully notes that the argued claim recitation has previously been recited in the claims and has been noted by examiner as non-functional descriptive material in previous office action. The examiner respectfully notes that the claim recitation “information that includes characteristics of the first action including a first location information, a first date and time information, and a first usage information,” merely describes information, without any ties between the information and the claimed functions and does not gain patentable weight. If applicant wants the recited “information” to be given patentable weight, the claims are required to be amended such that the effect of recited “information” such as, for example, time, date and location, on the claimed functions is explicitly and clearly shown in the claim. For example, the claim should be amended to show how information such as time, date and location affect the “training a learning model” function.
The examiner further notes that, although not argued by Applicant, new grounds of rejection with respect to claim amendments and newly added claims have been provided in this document.
Claim Objections
With respect to claim 5, the term “predetermined authentication” needs to be replaced with “second authentication, because the other instances of the term have been replaced throughout the claims.
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 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.
This application currently names joint inventors. In considering patentability of the claims, the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
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 set forth in Graham v. John Deere Co., 383 U.S. 1, 148 USPQ 459 (1966), that are applied 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.
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
Claims 1-4, 9-12 and 14-15, 19-21 and 23 are rejected under 35 U.S.C. 103 as being unpatentable over Streit (US Patent Publication No. 2021/0141896,) in view of Jass (US Patent Publication No. 2022/0116390 ,) further in view of Khan (US Patent Publication No. 2022/0083916.)
With respect to claims 1, 14 and 15, Streit teach:
train a learning model by adjusting a plurality of parameters of the learning model based on training data… (machine learning model is trained using training data sets: [0008], [0065]-[0067], [0100], claim 2)
wherein the training data comprises an input portion including first information that includes characteristics of the first action … (classifier (learning model) has an input layer and an output layer: FIGS. 4A-4D, training data includes input and output:[0014], [0050], input includes user behavioral data (i.e., actions): [0043], [0052], [0060])
wherein the training data comprises an output portion indicating that the first action is valid. (output indicates whether input is a valid match: [0043]-[0044] the model output indicates that input meets a validation threshold: [0070]-[0074])
authenticating the target user according to a first authentication method… (authentication function: FIG. 2A, [0248], FIG. 9, [0185])
The examiner notes that the claim recitation: “the information including a first location information, a first date and time information, and a first usage information,” indicates non-functional descriptive material and therefore does not further limit the scope of the claim.
In addition, with respect to claim 1, Streit teaches:
a learning model evaluation system, comprising at least one processor, ([0330])
In addition, with respect to claim 15, Streit teaches:
a non-transitory computer-readable information storage medium for storing a program ([0335])
Streit does not explicitly teach; however, Jess teaches:
a user terminal comprising an NFC unit or a camera; (computing device 950-a: [0061])
wherein the first authentication method is a login authentication to a predetermined service, (FIG. 9, login screen 905: [0061])
authenticate the target user according to the second authentication method, wherein the second authentication method comprises possession authentication to confirm the target user is in physical possession of a tangible object by using the NFC unit… (Multifactor authentication verifies user using two or more factors, such as possession factor of a security token or NFC dongle: [0061], [0074])
Jass does not explicitly teach:
…to acquire an individual number on the tangible object or the camera to image the tangible object and perform optical character recognition to recognize the individual number;
However, the claim recitation indicates intended use of the NFC unit and does not further limit the scope of the claim.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate multi-factor authentication where the first authentication is a login and a second authentication is based on a possession factor determined using NFC, as taught by Jass, into the learning model accuracy evaluation system of Streit based on a first authentication method (biometrics), in order to ensure model accuracy is evaluated based on an authentic user data.
Streit and Jass do not explicitly teach; however, Khan teaches:
train a learning model based on training data that indicates that a first action by a past user who has executed a second authentication is valid, (training the learning model based on a trusted dataset: [0030]-[0034,] associated with a pre-authorised used (who has executed a second authentication) [0062], [0210]-[0211])
Khan does not explicitly teach:
“…such that when the first action is input into the learning model, the learning model outputs that the first action is valid,”
However, claim recitation indicates intended use of the learning model, with an expected result, and therefore does not further limit the scope of the claim.
acquire a second information that includes characteristics of a second action of the target user on the predetermined service after the target user has been authenticated according to the second authentication method, (trusted data set (second information) is acquired: [0017], associated with a pre-authorised user: [0062], [0210]-[0211] )
input the second information into the learning model, and acquire an output from the learning model indicating whether the second action of the target user in the predetermined service is fraudulent or valid; (input the trusted data set to the learning model: [0020] and calculate model performance metrics using the trusted data set as input: [0017], [0020], [0063], [0081], [0094], claim 1)
evaluate an accuracy of the learning model by comparing the output of the learning model to the second information. (determine potential model drift (drop in model accuracy): [0017], [0021], [0063], [0081]-[0082], [0094], [0121], [0123])
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate calculating learning model accuracy using trusted data sets as input, as taught by Khan, into the learning model accuracy evaluation system of Streit and Jass, in order to detect drifts in learning model accuracy. (Khan: Abstract, [0017])
With respect to claim 2, Streit Jass and Khan teach the limitations of claim 1.
Moreover, Khan teaches:
acquire a plurality of pieces of the second information, (trusted data set is updated with updated parameters: [0031], [0033], [0064])
acquire the output corresponding to each of the plurality of pieces of the second information, (updated trusted data set is used for training the model: [0033]-[0034], [0043], [0173])
evaluate the accuracy of the learning model based on the output corresponding to each of the plurality of pieces of the second information. (evaluate model drifts based on training model using updated parameters: [0055], [0069], [0071], [0081])
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate calculating learning model accuracy using trusted data sets as input, as taught by Khan, into the learning model accuracy evaluation system of Streit and Jass, in order to detect drifts in learning model accuracy. (Khan: Abstract, [0017])
With respect to claim 3, Streit Jass and Khan teach the limitations of claim 1.
Moreover, Streit teaches:
retrain the learning model by using a latest action in the predetermined service to generate new training data when the accuracy of the learning model becomes less than a predetermined accuracy. (retrain model using results (as training data) until desired accuracy is achieved [0175], [0178], [0297])
The examiner notes that claim recitation “to generate new training data” indicates intended use of the retraining, and therefore does not further limit the scope of the claim.
With respect to claim 4, Streit Jass and Khan teach the limitations of claim 1.
Moreover, Streit teaches:
acquire confirmed information relating to a confirmed user action for which the confirmed user action has been confirmed as being fraudulent or not fraudulent , (known or unknown result: [0074]-[0077])
evaluate the accuracy of the learning model based on the second information and the confirmed information. (results are used to determine model accuracy: [0178], [0205], [0216])
With respect to claim 9, Streit Jass and Khan teach the limitations of claim 1.
Moreover, Streit teaches:
retrain, based on the second information, the learning model (retrain model using results (as training data) [0175], [0178], [0297])
such that the second action of the target user is estimated to be valid. (prediction by the classification network (i.e., model): [0050], determine whether authentication data meets a validation threshold (i.e. estimated validity: [0070]-[0074] )
The examiner notes that the claim recitation: “…such that the action of the authenticated user is estimated to be valid,” indicates an intended result of retraining and therefore does not further limit the scope of the claim.
With respect to claim 10, Streit Jass and Khan teach the limitations of claim 9.
Moreover, Streit teaches:
wherein the learning model is a supervised learning model, (the model is trained (i.e., supervised): [0045])
wherein the at least one processor is configured to train the learning model by creating first training data indicating that the second action of the target user is valid based on the second information, and training the learning model based on the first training data. ([0065]-[0067], [0088]-0092])
With respect to claim 11, Streit, Jass, and Khan teach the limitations of claim 10.
Moreover, Streit teaches:
acquire unauthenticated information relating to a third action of an unauthenticated user who is yet to execute the second authentication method , (enroll new user: [0065])
create second training data indicating that the third action of the unauthenticated user is valid or fraudulent based on the unauthenticated information, and to train the learning model based on the second training data. (additional training data for new user: [0065], update the model with new user data: [0091])
The examiner notes that the claim recitation “…who is yet to execute the second authentication method,” indicates non-functional descriptive material and therefore does not further limit the scope of the claim. In addition, the user is “yet to be executed…” is a function that has not been performed.
The examiner further notes that the claim recitation “to train the learning model…” indicates intended use and therefore does not further limit the scope of the claim because the function “training” is not positively recited.
With respect to claim 12, Streit Jass and Khan teach the limitations of claim 11.
Moreover, Streit teaches:
acquire an output from the learning model based on the unauthenticated information, and to create the second training data based on the output. (the model is incrementally retrained: [0087]-[0092])
With respect to claim 19, Streit Jass and Khan teach the limitations of claim 1.
Moreover, Streit teaches:
wherein if the accuracy of the learning model is more than a threshold, a notification is sent… (FIG. 11, [0228])
The examiner notes that the claim recitation “indicating that the accuracy of the learning model is high,” merely indicates content of the notification which is non-functional descriptive material and therefore does not further limit the scope of the claim.
With respect to claim 20, Streit Jass and Khan teach the limitations of claim 1.
Moreover, Streit teaches:
wherein if the accuracy of the learning model is less than a threshold, a notification is sent…(FIG. 11, [0228])
The examiner notes that the claim recitation “indicating that the accuracy of the learning model is low,” merely indicates content of the notification which is non-functional descriptive material and therefore does not further limit the scope of the claim.
With respect to claim 21, Streit Jass and Khan teach the limitations of claim 1.
Moreover, Streit teach:
wherein the accuracy indicates a probability of a desired result being obtained as an output from the learning model; ([0076]-[0077])
Streit Jass and Khan do not explicitly teach:
wherein the accuracy is measured as: a precision rate, a correct answer rate, a reproducibility rate, a false positive rate, a log loss, or an area under the curve.
However, the claim recitation indicates non-functional descriptive material that merely describes data and does not further limit the scope of the claim.
Claim 23 (New): The learning model evaluation system according to claim 3, wherein the at least one processor is configured to:
With respect to claim 23, Streit Jass and Khan teach the limitations of claim 3.
Moreover, Streit teach:
retrain the learning model by adjusting the plurality of parameters of the learning model based on the new training data, (retrain model using results (as training data) until desired accuracy is achieved [0175], [0178], [0297])
The examiner notes that claim recitation “wherein the new training data indicates that a fourth action by an authenticated user is valid” indicates non-functional descriptive material that merely describes data and therefore does not further limit the scope of the claim.
Claims 5-8, 13 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Streit, in view of Jass and Khan, further in view of Zarakas (US Patent Publication No. 2021/0383394)
With respect to claim 5, Streit Jass and Khan teach the limitations of claim 1.
Streit Jass and Khan do not explicitly teach:
wherein the predetermined authentication is possession authentication for confirming whether the target user possesses a predetermined card through use of the user terminal.
However, Zarakas teach:
wherein the predetermined authentication is possession authentication for confirming whether the user possesses a predetermined card through use of the user terminal, (authenticating user using a transaction card: [0014]-[0015], [0026], [0031])
The examiner notes that the claim recitation “…the predetermined authentication is possession authentication for confirming…” indicate non-functional descriptive material which does not further limit the scope of the claim. In addition, the recitation “to confirm…” indicates intended use of the authentication and does not further limit the scope of the claim.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate authentication using transaction cards in a machine learning environment, as taught by Zarakas, into the learning model accuracy evaluation system of Streit Jass and Khan in order to verify model accuracy based on transaction card information. (Zarakas : Abstract, [0038])
With respect to claim 6, Streit, Jass Khan and Zarakas teach the limitations of claim 5.
Moreover, Zarakas teach:
wherein each of a first card and a second card, each of which is the predetermined card, is usable in the predetermined service by the target user, ([0031], [0034]-[0038])
The examiner notes that the claim recitation “each of a first card and a second card, each of which is the predetermined card, is usable in the predetermined service by the authenticated user…” indicates non-functional descriptive material that describes intended use of the cards and therefore does not further limit the scope of the claim.
acquire the second information corresponding to the first card, (training learning model using card data: FIG. 2, [0044]-[0049])
acquire the output corresponding to the first card based on the information corresponding to the first card, (training learning model using card data: FIG. 2, [0044]-[0049])
evaluate the accuracy of the learning model based on the output corresponding to the first card. (FIG. 2, [0049]-[0051])
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate authentication using transaction cards in a machine learning environment, as taught by Zarakas, into the learning model accuracy evaluation system of Streit Jass and Khan in order to verify model accuracy based on transaction card information. (Zarakas : Abstract, [0038])
With respect to claim 7, Streit, Jass Khan and Zarakas teach the limitations of claim 6.
Moreover, Zarakas teach:
compare first name information relating to a name of the first card and second name information relating to a name of the second card, (cross-validation the learning model over features set which includes user data on the first card and second card: FIG. 2, [0053]-[0055])
acquire the second information corresponding to the second card when a result of the comparison is a predetermined result, (using a trusted card to authenticate another card: [0031], [0034], [0058]-[0060])
acquire the output corresponding to the second card based on the second information corresponding to the second card, (FIG. 2, [0044]-[0049])
evaluate the accuracy of the learning model based on the output corresponding to the first card and the output corresponding to the second card. (FIG. 2, [0049]-[0051])
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate authentication using transaction cards in a machine learning environment, as taught by Zarakas, into the learning model accuracy evaluation system of Streit Jass and Khan in order to verify model accuracy based on transaction card information. (Zarakas : Abstract, [0038])
With respect to claim 8, Streit, Jass Khan and Zarakas teach the limitations of claim 6.
Moreover, Zarakas teach:
wherein the second card is a card other than a card which supports the possession authentication, (each card has a different trust score: [0022])
wherein the second information corresponding to the second card is information relating to the second action of the second user who has used the second card on which the possession authentication has not been executed. (authentication user based on trust score of the card: [0059])
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate authentication using transaction cards in a machine learning environment, as taught by Zarakas, into the learning model accuracy evaluation system of Streit Jass and Khan in order to verify model accuracy based on transaction card information. (Zarakas : Abstract, [0038])
The examiner notes that the claim recitation “…wherein the second card is a card other than a card which supports the possession authentication…” indicate non-functional descriptive material which merely describes the card and does not further limit the scope of the claim.
The examiner further notes that the claim recitation “…wherein the information corresponding to the second card is information relating to the action of the authenticated user who has used the second card on which the possession authentication has not been executed…” indicate non-functional descriptive material which merely describes the data but does not affect the functions of the claim and therefore does not further limit the scope of the claim.
With respect to claim 13, Streit Jass and Khan teaches the limitations of claim 1.
Streit Jass and Khan do not explicitly teach; however Zarakas teach:
wherein the predetermined service is an electronic payment service usable from the user terminal, ([0075])
The examiner notes that the claim recitation “usable from the user terminal…” indicates intended use of the service and therefore does not further limit the scope of the claim.
wherein the second authentication method is authentication of the electronic payment service executed from the user terminal, ([0034])
wherein the second information is information relating to the second action of the target user in the electronic payment service, ([0033]-[0034])
wherein the learning model is a model for detecting fraud in the electronic payment service. ([0034], [0038], [0040], [0075])
The examiner notes that claim recitation “for detecting fraud…” indicates intended use of the learning model and therefore does not further limit the scope of the claim, because the “detecting” function is not performed.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate system of detecting fraud in payment services using machine learning models, as taught by Zarakas, into the learning model accuracy evaluation system of Streit Jass and Khan, in order to evaluate fraud detection models used in transactions. (Zarakas : Abstract, [0015])
With respect to claim 18, Streit, Jass Khan and Zarakas teach the limitations of claim 13
Moreover, Streit teaches:
restrict the target user's access to the predetermined service if the learning model detects fraud. ([0057], [0115], [0185]
Claim 22 is rejected under 35 U.S.C. 103 as being unpatentable over Streit, in view of Jass and Khan, further in view of Yokota (US Patent Publication No. 2022/0222582)
With respect to claim 22, Streit, Jass, and Khan teach the limitations of claim 1.
Streit, Jass, and Khan do not explicitly teach; however, Yokota teach:
calculate a correct answer rate…(a correct answer rate is calculated: [0104])
The examiner notes that the claim recitation “as a ratio of outputs of the learning model indicating valid from among the second information input into the learning model” indicates non-functional descriptive material that merely describes the rate and does not further limit the scope of thew claim.
evaluate the accuracy of the learning model based on the correct answer rate wherein the accuracy of the learning model is evaluated as higher when the correct answer rate is higher and the accuracy of the learning model is evaluated as lower when the correct answer rate is lower. (accuracy is calculated based on correct answer rate, when correct answer rate decreases (lower) accuracy decreases (lower): [0111], [0138])
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate learning model accuracy detection , as taught by Yokota, into the learning model accuracy evaluation system of Streit Jass and Khan, in order to prevent deterioration of learning model accuracy. (Yokota : Abstract, [0003]-[0004])
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Baba (US 2019/0281172,) teaches detecting learning model accuracy using correct answer rate.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to SIMA ASGARI whose telephone number is (571)272-2037. The examiner can normally be reached M-F 9am-6pm.
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, Patrick McAtee can be reached at (571)272-7575. 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.
/SIMA ASGARI/Examiner, Art Unit 3698
/PATRICK MCATEE/Supervisory Patent Examiner, Art Unit 3698
Prosecution Insights