DETAILED ACTION
1. The present application, filed on or after March 13, 2013, is being examined under the first inventor to file provisions of the AIA .
This is a CIP patent application with a claim of priority to a provisional application filed January 25, 2022 and the parent application, Application No. 17/824,688, filed May 25, 2022.
Response to Amendment
2.. An RCE with accompanying Amendment was filed May 26, 2026 (hereinafter “Amendment”) and has been entered into the record and fully considered. The Amendment was filed in response to a Final Rejection dated August 5, 2025, which was subsequently withdrawn, and a new Final Rejection was mailed January 16, 2026.
Despite the Amendment to the Claims and Applicant’s remarks, the Rejections set forth in the Non-Final Rejection are hereby maintained; although, the Rejection under §103 is based on NEW GROUNDS necessitated by the Amendment.
An explanation of the maintained Rejections and a response to Applicant’s arguments are set forth below. Please see the “Conclusion” section of this Action below for important information regarding responding to this Action.
Claims 1 – 4, 6 – 8, 10, 12 – 14, 16 – 17, and 19 – 25 are pending and examined herein.
Claims 5, 9, 11, 15, and 18 were cancelled in a previous response.
OFFICE NOTE: Interviews are always welcome at any stage of prosecution. Please use the AIR form for scheduling an interview if such is desired. The link for the AIR form is found at the end of this Action.
STATUS OF THE CLAIMS:
The independent claims were amended in substantially identical/similar fashion, making it unnecessary to address each Claim. That is, independent Claims 1, 14, and 20 were amended so as to be structurally and functionally virtually equivalent and of a similar scope.
Therefore, the explanation of the maintained Rejections with respect to Claim 1 is considered explanatory of the rejection as a whole.
With regard to the Amendment:
In light of the Amendment, the previous OBJECTION TO THE CLAIMS is hereby WITHDRAWN.
Claim 1 was amended as follows:
PNG
media_image1.png
678
671
media_image1.png
Greyscale
PNG
media_image2.png
68
663
media_image2.png
Greyscale
Summary of the Amendment and Broadest Reasonable Interpretation:
Claim terminology is to be given its plain and ordinary meaning to a person of ordinary skill in the art, consistent with the specification. This is true, unless the terms are given a special meaning. See MPEP §2111.01
Here, no special meaning is detected. As noted in the Amendment, the
changes to Claim 1 relate generally to:
correcting the previous objection to the language of the independent claims, and
iteratively/adusting weight values for “behaviors” (e.g. features, indicators, variables) that are indicative of either fraud or non-fraud.
Weighted values are assigned to each type of one or more flagged fraudulent behavior and non-fraudulent behavior. See at least 0005.
With regard to §101:
Respectfully, the Amendment does not advance prosecution substantially.
Thus, the amendments to the Claim do not alter the analysis set for the Non-Final Rejection regarding §101. The only changes are summarized above.
The changes to the claim relevant to the previous objection merely clarify the language of the Claim – i.e. “addresses” in lieu of “address information address.”
The amendment relating to “determining relationships” merely clarifies that the “relationship” relates to a likelihood that the behaviors are associated with fraud or non-fraud. Furthermore, it is clear from the previous limitation that “flagged” behaviors are associated with fraud.
All of these are high level and generic concepts and very common in machine learning. Moreover, iteratively adjusting weights is also a common and high level, generic common. A first weight and a second weight merely indicates that different features are to be given different weights in the prediction and score generation process. This is extremely well known. This is what weights are for.
Iterative adjustment of weights or coefficients is extremely common in improving the accuracy of machine learning models. This process is a core part of training and fine-tuning models, and it’s used across nearly all modern ML and deep learning workflows. Machine learning models are defined by parameters (weights and biases) that determine how inputs are transformed into outputs. These parameters are typically initialized randomly or with small values, and then updated repeatedly to minimize a loss or cost function that measures prediction error.
Thus, the changes to the claim are high-level, abstract, and conceptual. No specific details about “how” the weights are adjusted and for which features/behaviors. There is nothing concrete, substantive, or specific about these new recitations.
No special functionality is recited. No new computerized components are recited.
These limitations recite results or “outcome” of computer processing without specifying “how” a technical problem is solved. That is, the solution of a technical problem is not reflected in the Claim.
Taking the claim elements separately, the function performed by the computer elements at each step of the process is purely typical of obtaining training data, deriving “behaviors” or attributes or features that are indicative of possible fraud, labelling those behaviors, and iteratively adjusting weights applied to the respective features. And then using the thus trained machine learning model to generate a risk score are among the most basic functions of a computer. Without greater specificity as to “how” certain functions solve a technical problem, the currently recited limitations can be achieved by any general purpose computer without special programming. In short, each step does no more than require a generic computer to perform generic computer functions. Considered as an ordered combination, the computer components of the Claim add nothing that is not already present when the steps are considered separately.
Claim 1 does not, for example, purport to improve the functioning of the computer elements nor does the claim reflect how an improvement in any other technology or technical field is achieved. Thus, Claim 1 remains in the category of “apply it.”
Accordingly, the Rejection is maintained.
With regard to §103:
It is respectfully submitted that the primary reference to Lagneaux teaches the features are were added by way of the Amendment. Lagneaux is a reference owned by the U.S. Postal Service. It clearly relates to address fraud and to behaviors/indicators/features of detecting fraudulent activities relatively to the delivery of mail to an address. This reference therefore applies directly to the amended claim.
As noted above, the iterative adjusting of weights is commonplace. It is taught in Lagneaux as adjusting “parameters,” which is considered to constitute the recited term “weight.” Adjusting such parameters is taught at 0052, 0102, and 0107. Furthermore, the secondary teaching reference to Goshen also teaches the adjustment of weights. See Goshen: 0051 and 0063.
However, out of an abundance of caution, Applicant’s own U.S. Patent Publication No. to Love et al. is cited for its teachings relative to adjusting weights of a ML model for greater accuracy in detecting fraud.
Thus, the Claims are rejected on the following NEW GROUNDS:
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 of this title, 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 – 8, 10, 12 – 14, 16 – 17, and 19 – 25 are rejected under 35 U.S.C. §103 as being unpatentable over U.S. Patent Publication No. 2021/0110343 to Lagneaux et al. (hereinafter “Lagneaux”) in view of U.S. Patent Publication No. 2021/0158356 to Goshen (hereinafter “Goshen”) and further in view of U.S. Patent Publication No. 2014/0222631 to Love et al. (hereinafter “Love”).
Applicant should be well-acquainted with the teachings of Love; therefore, no detailed explanation of the Rejection is deemed necessary. Love relates to the detection of fraud in terms of a “suspect entity.” See title and 0002 – 0004. One of the features or attributes of fraud used by Love is a physical address or mailing address. (See at least 0019 and 0021). The ML model of Love uses the adjustment of weights to improve accuracy:
“[0049] Finally, at step 328, the final linked nodes and networks are stored at system 200 for subsequent access by data users 230. For purposes of being stored and indexed in the storage device 250, each network, node, and link may be assigned an identifier. Further, along with each candidate network reviewed by the data user, the data user may enter a perceived value score that ranks how accurately the score associated with the candidate network reflects an actual level of risk (or a perceived level of risk). The perceived score information entered by the data user in step 740 (FIG. 7) may then be subsequently used to improve the accuracy of scoring or identification of candidate networks. For example, the perceived score information entered by the data user may be fed to learning algorithm such as the neural decision engine discussed herein, and in conjunction with the stored score value for the candidate network, an error signal can be generated that reflects the magnitude of the difference between the scored risk and the perceived score, which may then be fed forward to adjust the scoring algorithm or network weights. In this manner, the system automatically adjusts for the scoring of candidate networks that more closely match real-world end-user conditions.” (Emphasis Added)
Therefore, it would have been prima facie obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the combined machine learning based address risk system of Lagneaux in view of Goshen to add the weight adjusting teachings of Love. The motivation to make this modification comes from Lagneaux. It teaches, as illustrated above, that “parameters” can be adjusted to improve accuracy in the predictions of fraud. It would greatly enhance the efficiency and accuracy of the combined system of Lagneaux in view of Goshen to add the weight adjustment features of Love.
As to Claim 10, previously discussed, Lagneaux teaches that “counts” are to be used in adjusting the weights. (See at least 0070 and 0095 and Tables 2 and 5).
Therefore, the existing Rejection under §103 must be maintained.
Response to Arguments
3. Applicant's arguments set forth in the Remarks section of the Amendment have been fully considered but they are not persuasive.
With regard to section 101 rejection, Applicant argues as follows:
PNG
media_image3.png
532
685
media_image3.png
Greyscale
The basic weakness of these arguments – assuming arguendo they are true and supported in the specification – is that these features are not reflected in the Claim – not in general, not at all. The Claim is devoid of teachings as to “how” the weights are adjusted. For example, there is no reflection in the claim of how the error is minimized in training, nor how optimization is achieved, nor how the behaviors are identified to be increased or decreased in weightage.
These, and likely other, details need to be reflected in the Claim to bring it into the realm of eligibility.
Such is not yet the case. The Rejection is maintained.
With regard to the §103 Rejection, Applicant arguments are moot in view of the new grounds of Rejection.
The amended features are taught by the current combination of references as noted above. Thus, Applicant’s arguments are moot.
The Rejection must be maintained.
Conclusion
4. Applicant should carefully consider the following in connection with this Office Action:
A. Search and Prior Art
The search conducted in connection with this Office Action, as well as any previous Actions, encompassed the inventive concepts as defined in the Applicant’s specification. That is, the search(es) included concepts and features which are defined by the pending claims but also pertinent to significant although unclaimed subject matter. Accordingly, such search(es) were directed to the defined invention as well as the general state of the art, including references which are in the same field of endeavor as the present application as well as related fields (e.g. using labeled training data to train ML models to detect fraud). Indeed, there is a plethora of prior art in these fields.
Therefore, in addition to prior art references cited and applied in connection with this and any previous Office Actions, the following prior art is also made of record but not relied upon in the current rejection:
U.S. Patent Publication No. 2022/0172211 to Muthuswamy et al. This reference relates to the concept of adjusting weights of a ML model.
Non-Patent Literature to Su et al., “Credit Risk Prediction Based on DenseNet-BC of Fusion Focal Loss and Static Restart SGD,” IEEE SmartWorld, Ubiquitious Intelligence and Computing, 2021
B. Responding to this Office Action
In view of the foregoing explanation of the scope of searches conducted in connection with the examination of this application, in preparing any response to this Action, Applicant is encouraged to carefully review the entire disclosures of the above-cited, unapplied references, as well as any previously cited references. It is likely that one or more such references disclose or suggest features which Applicant may seek to claim. Moreover, for the same reasons, Applicant is encouraged to review the entire disclosures of the references applied in the foregoing rejections and not just the sections mentioned.
C. Interviews and Compact Prosecution
The Office strongly encourages interviews as an important aspect of compact prosecution. Statistics and studies have shown that prosecution can be greatly advanced by way of interviews. Indeed, in many instances, during the course of one or more interviews, the Examiner and Applicant may reach an agreement on eligible and allowable subject matter that is supported by the specification.
Interviews are especially welcomed by this examiner at any stage of the prosecution process. Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool (e.g. TEAMS). To facilitate the scheduling of an interview, the Examiner requests the use of the AIR form as follows:
USPTO Automated Interview Request http://www.uspto.gov/interviewpractice.
Other forms of interview requests filed in this application may result in a delay in scheduling the interview because of the time required to appear on the Examiner's docket. Thus, the use of the AIR form is strongly encouraged.
D. Communicating with the Office
Any inquiry concerning this communication or earlier communications from the examiner should be directed to WILLIAM BUNKER whose telephone number is (571)272-0017. The examiner can normally be reached on M - F 8:30AM - 5:30PM, Pacific.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Abhishek Vyas, can be reached at 571-270-1836. Information regarding the status of an application, whether published or unpublished, may be obtained from the “Patent Center” system. For more information about the Patent Center system, see https://patentcenter.uspto.gov/
/William (Bill) Bunker/
U.S. Patent Examiner
AU 3691
(571) 272-0017 - office
william.bunker@uspto.gov
June 24, 2026
/ABHISHEK VYAS/Supervisory Patent Examiner, Art Unit 3691
Prosecution Insights