DETAILED ACTION
Response to Amendment
1. The amendment filed on 5/21/2026 has been entered. Claims 1, 9, 12, 15 and 22 have been amended. No new claims have been added or cancelled. Accordingly, claims 1-17, 22, 23, and 28 are pending in this office action
Notice of Pre-AIA or AIA Status
2. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Allowable Subject Matter
3. Claims 6-12 objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Claim Rejections - 35 USC § 102
4. 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 the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claim(s) 1, 2, 16, 17 and 22 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by US 2022/0179840 (hereinafter Chatterjee).
As for claim 1 Chatterjee discloses: 1. A computer-implemented method for determining bias of machine learning models (See paragraphs 0002-0004 and 0021 note the system determines bias from the remote machine but can also determine bias with respect to the local machine), comprising: forming a training dataset comprising input data samples provided to a remote machine learning model developed using a machine learning process, and corresponding output data samples obtained from the remote machine learning model (See paragraphs 0026, 0029 and 0044 note the data samples are taken from trusted and untrusted and the system trains from the samples); training a local machine learning model which approximates the remote machine learning model using a machine learning process and the training dataset (See paragraphs 006, 0024, 0038 note the system can be trained on both the trusted and untrusted data); and interrogating (204) the trained local machine learning model to determine whether the remote machine learning model is biased with respect to one or more biasing data parameters (See paragraphs 0019-0024 note the bias correction is trained).
As for claim 2 the rejection of claim 1 is incorporated and further discloses: wherein interrogating the local machine learning model comprises: augmenting the training dataset with the one or more biasing data parameters as input features, to create an augmented training dataset (See paragraphs 0007 and 0037, note an updated training dataset based on the features and the bias is an augmented training dataset); retraining the local machine learning model using the machine learning process and the augmented training dataset (See paragraphs 0026, 0037 note the initial training is used to retrain the system based on the dynamically built trusted dataset);
determining an importance of the one or more biasing data parameters to the retrained local machine learning model; and based on the determined importance of the one or more biasing data parameters to the retrained local machine learning model, determining whether the remote machine learning model is biased with respect to the one or more biasing data parameters (See paragraphs 0033 and 0066).
Claims 16 and 17 are apparatus claims substantially corresponding to the method of claims 1 and 2 and are thus rejected for the same reasons as set forth in the rejection of claims 1 and 2.
As for claim 22 the rejection of claim 16 is incorporated and further Chatterjee discloses: wherein the apparatus is further caused to perform one or more actions to mitigate bias with respect to the one or more biasing parameters, the one or more actions further comprise determining an importance of input data features to the further machine learning model, based on the importance of the one or more input data features to the further machine learning model, removing one or more input data features from the training dataset to obtain an unbiased training dataset, and retraining the local machine learning model using the machine learning process and the unbiased training dataset (See paragraphs 006, 0024, 0038 note the system can be trained on both the trusted and untrusted data).
Claim Rejections - 35 USC § 103
5. 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.
Claim(s) 3-4 is/are rejected under 35 U.S.C. 103 as being unpatentable over Chatterjee as applied to claim 1 above, and further in view of US 2023/0229735 (hereinafter Jain).
As for claim 3 the rejection of claim 1 is incorporated and further Jain renders obvious: wherein the importance of the one or more biasing parameters is quantified by respective Shapley values associated with the one or more biasing data parameters (See paragraphs 0097). It would have been obvious to an artisan of ordinary skill in the pertinent at the time the instantly claimed invention was filed to have incorporated the teaching of Jain into the system of Chatterjee. The modification would have been obvious because the two references are concerned with the solution to problem of processing data based on bias (See Jain and Chatterjee abstract), therefore there is an implicit motivation to combine these references (i.e. motivation from the references themselves). In other words, the ordinary skilled artisan, during his/her quest for a solution to the cited problem, would look to the cited references at the time the invention was made. Consequently, the ordinary skilled artisan would have been motivated to combine the cited references since Jain’s teaching would enable users of the Chatterjee system to have more efficient processing.
As for claim 4 the rejection of claim 2 is incorporated and further Jain renders obvious: wherein whether the remote machine learning model is biased with respect to the one or more biasing data parameters is further based on a comparison of an accuracy of the local machine learning model to an accuracy of the retrained local machine learning model (See paragraphs 0006, 0057 and 0084 note the system can compare intermediate/local data to determine accuracy). It would have been obvious to an artisan of ordinary skill in the pertinent at the time the instantly claimed invention was filed to have incorporated the teaching of Jain into the system of Chatterjee. The modification would have been obvious because the two references are concerned with the solution to problem of processing data based on bias (See Jain and Chatterjee abstract), therefore there is an implicit motivation to combine these references (i.e. motivation from the references themselves). In other words, the ordinary skilled artisan, during his/her quest for a solution to the cited problem, would look to the cited references at the time the invention was made. Consequently, the ordinary skilled artisan would have been motivated to combine the cited references since Jain’s teaching would enable users of the Chatterjee system to have more efficient processing.
6. Claim(s) 5, 13-15, 23 and 28 is/are rejected under 35 U.S.C. 103 as being unpatentable over Chatterjee as applied to claim 1 or 16 above, and further in view of US 2024/0095588 (hereinafter Vandikas).
As for claim 5 the rejection of claim 1 is incorporated and further Vandikas discloses:
further comprising performing one or more actions to mitigate bias with respect
to the one or more biasing parameters (See paragraph 0029 note the system will retrain to mitigate bias). It would have been obvious to an artisan of ordinary skill in the pertinent at the time the instantly claimed invention was filed to have incorporated the teaching of Vandikas into the system of Chatterjee. The modification would have been obvious because the two references are concerned with the solution to problem of processing data based on bias (See Vandikas and Chatterjee abstract), therefore there is an implicit motivation to combine these references (i.e. motivation from the references themselves). In other words, the ordinary skilled artisan, during his/her quest for a solution to the cited problem, would look to the cited references at the time the invention was made. Consequently, the ordinary skilled artisan would have been motivated to combine the cited references since Vandikas’s teaching would enable users of the Chatterjee system to have more efficient processing.
As for claims 13 and 28 the rejection of claims 1 and 16 are incorporated respectfully and further Vandikas discloses: wherein the method is performed by a network node in a communications network (See paragraphs 0022 note network node). It would have been obvious to an artisan of ordinary skill in the pertinent at the time the instantly claimed invention was filed to have incorporated the teaching of Vandikas into the system of Chatterjee. The modification would have been obvious because the two references are concerned with the solution to problem of processing data based on bias (See Vandikas and Chatterjee abstract), therefore there is an implicit motivation to combine these references (i.e. motivation from the references themselves). In other words, the ordinary skilled artisan, during his/her quest for a solution to the cited problem, would look to the cited references at the time the invention was made. Consequently, the ordinary skilled artisan would have been motivated to combine the cited references since Vandikas’s teaching would enable users of the Chatterjee system to have more efficient processing.
As for claim 14 the rejection of claim 13 is incorporated and further Vandikas discloses: wherein the input data samples and corresponding output data samples are received from one or more wireless devices coupled to the communications network (See paragraphs 0022 note wireless). It would have been obvious to an artisan of ordinary skill in the pertinent at the time the instantly claimed invention was filed to have incorporated the teaching of Vandikas into the system of Chatterjee. The modification would have been obvious because the two references are concerned with the solution to problem of processing data based on bias (See Vandikas and Chatterjee abstract), therefore there is an implicit motivation to combine these references (i.e. motivation from the references themselves). In other words, the ordinary skilled artisan, during his/her quest for a solution to the cited problem, would look to the cited references at the time the invention was made. Consequently, the ordinary skilled artisan would have been motivated to combine the cited references since Vandikas’s teaching would enable users of the Chatterjee system to have more efficient processing.
As for claim 15 the rejection of claim 1 is incorporated and further Vandikas discloses:
wherein one or more of the following applies: the remote machine learning model provides
an output on a basis of which a radio access network operation is performed; the remote machine learning model provides an output on the basis of which an action is performed in a smart factory; the remote machine learning model provides an output on the basis of which an autonomous vehicle operation is performed; and the remote mode provides an output on the basis of which a medical procedure is performed (See paragraphs 0022 note radio). It would have been obvious to an artisan of ordinary skill in the pertinent at the time the instantly claimed invention was filed to have incorporated the teaching of Vandikas into the system of Chatterjee. The modification would have been obvious because the two references are concerned with the solution to problem of processing data based on bias (See Vandikas and Chatterjee abstract), therefore there is an implicit motivation to combine these references (i.e. motivation from the references themselves). In other words, the ordinary skilled artisan, during his/her quest for a solution to the cited problem, would look to the cited references at the time the invention was made. Consequently, the ordinary skilled artisan would have been motivated to combine the cited references since Vandikas’s teaching would enable users of the Chatterjee system to have more efficient processing.
As for claim 23 the rejection of claim 22 is incorporated and further Vandikas discloses: wherein the apparatus is caused to train the further machine learning model, determine an importance of input data features to the further machine learning model, and remove one or more input data features from the training dataset iteratively (See paragraphs 0052 and 0067 note the system performs iterative removal). It would have been obvious to an artisan of ordinary skill in the pertinent at the time the instantly claimed invention was filed to have incorporated the teaching of Vandikas into the system of Chatterjee. The modification would have been obvious because the two references are concerned with the solution to problem of processing data based on bias (See Vandikas and Chatterjee abstract), therefore there is an implicit motivation to combine these references (i.e. motivation from the references themselves). In other words, the ordinary skilled artisan, during his/her quest for a solution to the cited problem, would look to the cited references at the time the invention was made. Consequently, the ordinary skilled artisan would have been motivated to combine the cited references since Vandikas’s teaching would enable users of the Chatterjee system to have more efficient processing.
Response to Arguments
7. Applicant's arguments filed on 5/21/2026 have been fully considered but they are not persuasive.
Applicant argues:
However, Chatterjee does not teach "forming a training dataset comprising input data samples provided to a remote machine learning model developed using a machine learning process, and corresponding output data samples obtained from the remote machine learning model." Instead, Chatterjee merely discusses that it obtains a "training dataset from database 116" and that the "training dataset may be sampled to generate or determine a trusted dataset including a small portion of the training dataset, such that a remaining dataset exists The data samples in the remaining dataset may be determined to be untrusted dataset, that is, dataset including biased labels." Chatterjee may discuss that the training dataset includes a trusted and untrusted dataset, as asserted by the Office, but there is no indication in Chatterjee that the trusted and untrusted dataset include i) "input data samples provided to a remote machine learning model" or ii) "corresponding output data samples obtained from the remote machine learning model," as recited in claims 1 and 16.
Examiner responds:
Examiner is not persuaded. Examiner is entitled to give claim limitations their broadest reasonable interpretation in light of the specification. Interpretation of Claims-Broadest Reasonable Interpretation: During patent examination, the pending claims must be ‘given the broadest reasonable interpretation consistent with the specification.’ Applicant always has the opportunity to amend the claims during prosecution and broad interpretation by the examiner reduces the possibility that the claim, once issued, will be interpreted more broadly than is justified. In re Prater, 162 USPQ 541,550-51 (CCPA 1969). In this case the remaining dataset is remote. A full reading of the claim limitations also read in parts “forming a training dataset comprising input data samples provided to a remote machine learning model developed using a machine learning process” , and corresponding output data samples obtained from the remote machine learning model. Examiner recommends clarifying how the remote learning model is developed using a machine learning process.
Applicant argues:
Moreover, Chatterjee does not teach "training a local machine learning model which approximates the remote machine learning model using a machine learning process and the training dataset." In short, Chatterjee does not use the trusted dataset and/or untrusted dataset to approximate a remote machine learning model. Instead, Chatterjee merely discusses performing different operations based on an updated remaining dataset (e.g., the untrusted dataset). For example, paragraph [0006] discusses how a "computing device may perform operations (e.g., train a machine learning model) using the trusted dataset and the updated remaining dataset," paragraph [0024] discusses how a "Database 116 can be a remote storage device, such as a cloud-based server, a memory device on another application server, a networked computer, or any other suitable remote storage," and paragraph [0038] discusses how a "bias correction computing device 102 may generate a classification model, M trained (e.g., built) based on the untrusted dataset." However, none of the model trained using the trusted dataset, the remote database, or the generated classification model based on the untrusted data disclosure "training a local machine learning model which approximates the remote machine learning model," as recited in claims 1 and 16.
Examiner responds:
Examiner is not persuaded. Examiner is entitled to give claim limitations their broadest reasonable interpretation in light of the specification. Interpretation of Claims-Broadest Reasonable Interpretation: During patent examination, the pending claims must be ‘given the broadest reasonable interpretation consistent with the specification.’ Applicant always has the opportunity to amend the claims during prosecution and broad interpretation by the examiner reduces the possibility that the claim, once issued, will be interpreted more broadly than is justified. In re Prater, 162 USPQ 541,550-51 (CCPA 1969). In this case approximations are never exact. As applicant admits the bias correction and classification are ways of approximating the learning process and dataset. Examiner recommends clarifying how the instant claims approximate a learning model.
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 ELIYAH STONE HARPER whose telephone number is (571)272-0759. The examiner can normally be reached on Monday-Friday 10:00 am - 6:00 pm.
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, Sanjiv Shah can be reached on (571) 272-4098. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300. Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/Eliyah S. Harper/Primary Examiner, Art Unit 2166 July 10, 2026
Prosecution Insights