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. Claims 1-20 are presented for examination. This Office action is Non-Final . Information Disclosure Statement The information disclosure statement s (IDS) filed on 09/20/2023, 10/29/2024 and 09/11/2025 have been considered by the Examiner and made of record in the application file. Claim Objections Claims 11, 15, 16, 18 and 20 are objected to because of the following informalities: in the first limitation of the claims it appears to be missing a semicolon --;-- Appropriate correction is required. Claim Rejections - 35 USC § 101 35 U.S.C. 101 reads as follows: Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title. Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Step 1 — Statutory Category Independent claims 1, 10, and 17 recite a method, computer program product, and system, respectively, and therefore fall within the statutory categories of process, manufacture, and machine. Eligibility analysis is required because the claims may be directed to a judicial exception. Step 2A — Prong One The claims recite an abstract idea. The claims recite operations including: identifying pairs of categories for potential merging testing merge strategies evaluating predictive accuracy selecting a merger based on accuracy impact merging categories into a hybrid category These steps constitute mathematical concepts and mental processes. Specifically, the claims involve: evaluating data relationships performing comparisons optimizing model parameters selecting results based on performance metrics Such activities are forms of mathematical analysis and data evaluation, which courts have consistently held to be abstract ideas (e.g., Alice Corp. v. CLS Bank, Electric Power Group v. Alstom, SAP America v. InvestPic ). The claims manipulate abstract information (categories, predictors, accuracy values) rather than physical objects or technological processes. Step 2A — Prong Two The claims do not integrate the abstract idea into a practical application . The additional elements recited include: a processor computer-readable media program instructions a generic “engine” These components merely perform their conventional functions of executing software instructions and storing data. The claims do not: improve the functioning of a computer itself recite specialized hardware control a physical device or process effect a transformation of matter solve a technological problem rooted in computer technology Improving the predictive accuracy of a model is an improvement in an abstract mathematical construct, not a technological improvement to computer functionality. Accordingly, the claims fail to integrate the abstract idea into a practical application. Step 2B — Inventive Concept The claims do not recite “significantly more” than the abstract idea . The claims do not include additional elements that amount to significantly more than the abstract idea. The recited components are well-understood, routine, and conventional computer elements performing their ordinary functions: processors executing instructions storage media storing data generic software modules There is no unconventional technical solution or specific implementation beyond applying the abstract idea on a generic computer. The dependent claims (2–9, 11–16, 18–20) also do not add eligibility-conferring subject matter. They merely further limit the abstract data-analysis process recited in the independent claims by specifying particular rules, criteria, or variations for selecting and merging categories, including: identifying “idle” or low-importance categories handling zero-count categories distinguishing ordinal vs. non-ordinal predictors selecting merges based on degree of accuracy change iterative application of the merging procedure These limitations amount to additional data evaluation, comparison, and optimization steps, which remain within the abstract idea (mathematical concepts / mental processes). The dependent claims do not recite: technological improvements to computer functionality specialized hardware physical-world interaction a transformation of matter any other practical application beyond the abstract analysis The dependent claims merely refine the abstract idea by specifying particular criteria or rules for selecting and evaluating category mergers and therefore do not integrate the judicial exception into a practical application or add significantly more. Therefore, claims 1-20 represents an abstract idea. Claim Rejections - 35 USC § 103 The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action: A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made. Claim s 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Chang et al. (US 11,250,042 B2) hereinafter “Chang”, in view of Mac Manus et al. (US 2022/0027782 A1) hereinafter “Manus”, further in view of Achin et al. (US 10,366,346 B2) hereinafter “Achin” . With respect to claims 1, 10 and 17, the Chang reference discloses a computer-implemented method, computer program product and computer system a processor and one or more computer readable storage media, and program instructions collectively stored on the one or more computer readable storage media, the program instructions executable by the processor to cause the processor to perform operations [ see col. 6, lines 11-18, disclosing the computing system 300 can be implemented with a suitable combination of hardware and/or software that generally includes one or more suitably configured or programmed hardware processors, and may be, e.g., a general-purpose computer or computer cluster with CPUs/GPUs, memory, input/output devices and/or a network interface ] comprising: identifying, by a post-modeling category merging engine, a plurality of valid pairs [ see col. 6, lines 18-25, disclosing t he computing systems 200, 300 for training the conflation and alignment system 212 and using it to merge new data into the taxonomy 202 are simply different subsets of components (functional blocks and data structures) all residing and executing on the same physical machine or group of machines (e.g., using the same hardware processors to execute overlapping, but distinct sets of instructions) ] associated with a categorical predictor , the plurality of valid pairs representing potential mergers of categories associated with a categorical predictor of a predictive model ; and merging , by the post-modeling category merging engine based on the testing, a valid pair in the plurality of valid pairs [ see col. 11, lines 34-36, disclosing t he neural network model provides a completely automated approach to classifying the merge actions for conflation candidates 224 ] to form a hybrid category . (emphasis added) Chang discloses the computer-implemented method, computer program product and computer system, as referenced above. Chang is silent of it compris ing : identifying, by a post-modeling category merging engine , a plurality of valid pairs associated with a categorical predicto r, the plurality of valid pairs representing potential mergers of categories associated with a categorical predictor of a predictive model ; testing, by the post-modeling category merging engine, a merge strategy for the plurality of valid pairs to determine a merger that minimizes a loss in accuracy of the predictive model ; and merging, by the post-modeling category merging engine based on the testing , a valid pair in the plurality of valid pairs to form a hybrid category . (emphasis added) However , Manus discloses it comprises: identifying, by a post-modeling category merging engine , a plurality of valid pairs associated with a categorical predictor, the plurality of valid pairs representing potential mergers of categories associated with a categorical predictor of a predictive model [ see Abstract, disclosing performing predictive data analysis that utilize at least one of categorical level merging, mutual-information-based feature filtering, feature-correlation-based feature filtering to generate training data feature value arrangements, as well as training and using categorical input machine learning models trained using the training data feature value arrangements ] ; and merging , by the post-modeling category merging engine based on the testing , a valid pair in the plurality of valid pairs to form a hybrid category [ see ¶0033, disclosing p er-level predictive correlation measures may be determined for both initial categorical levels as well as merged categorical levels determined by combining groups of two or more initial categorical levels ] . (emphasis added) Therefore, it would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains to modify the taxonomy enricher as disclosed by Chang with the categorical input machine learning models as disclosed by Manus. Doing so have enhanced Chang by allow it to efficiently and reliably perform predictive data analysis using categorical input data. The combination of Chang and Manus discloses the computer-implemented method, computer program product and computer system, as referenced above. The combination does not discloses testing, by the post-modeling category merging engine, a merge strategy for the plurality of valid pairs to determine a merger that minimizes a loss in accuracy of the predictive model . (emphasis added) However, Achin discloses testing, by the post-modeling category merging engine, a merge strategy for the plurality of valid pairs to determine a merger that minimizes a loss in accuracy of the predictive mode l [ see col. 2, lines 1-8, disclosing t he observations are generally partitioned into at least one “training” dataset and at least one “test” dataset ; A data analyst then selects a statistical-learning procedure and executes that procedure on the training dataset to generate a predictive model ; The analyst then tests the generated model on the test dataset to determine how well the model predicts the value(s) of the target(s), relative to actual observations of the target(s) ; also, see col. 2, lines 51-54, disclosing reducing (e.g., minimizing) any loss of accuracy associated with moving from a source model to a second-order model (and, in some cases, for generating second-order models with greater accuracy than their source models) ] . Therefore, it would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains to modify the combination with the statistical learning methods as taught by Achin. Doing so would have enhanced the combination by developing and selecting predictive models that minimizes loss of accuracy. With respect to claims 2, 11 and 18, as modified the combination of Chang, Manus and Achi discloses the computer-implemented method, computer program product and computer system of claims 1, 10 and 17, as referenced above. The combination further discloses it comprises: identifying a plurality of idle categories representing underutilized categories associated with the categorical predictor [ Manus, see ¶0035, disclosing adjusted categorical levels include initial categorical levels whose individual per-level prediction correlation measure satisfies a per-level prediction correlation threshold as well as merged categorical levels each associated with one or more low-correlation categorical levels whose merged per-level prediction correlation measure satisfies a per-level prediction correlation threshold, where the one or more low-correlation categorical levels associated with a particular merged categorical level are each deemed to not be sufficiently statistically significant in predicting the target feature in accordance with the per-level predictive correlation measures of the noted low-correlation categorical levels ] ; merging the plurality of idle categories [ Manus, see ¶0099, disclosing after generating the adjusted categorical levels (which may or may not include the merged categorical level for the low-correlation subset of initial categorical levels depending on whether the merged per-level correlation measure for the merged categorical level satisfies the per-level predictive correlation threshold), the predictive data analysis computing entity 106 generates a set of categorically refined features based at least in part on the noted adjusted categorical levels ] ; and generating, based on a plurality of category importance values associated with a plurality of non-idle categories, the plurality of valid pairs [ Achin, see col. 61, lines 33-38, disclosing the system 100 may present evaluated predictive models and their associated model-specific predictive values to the user to the user (e.g., at step 446 of the method 400) ; In some embodiments, the system 100 calculate and/or display the feature importance values only for a subset of the predictive models (e.g., the top models) ] . With respect to claims 3, 12 and 19 , as modified the combination of Chang, Manus and Achi discloses the computer-implemented method, computer program product and computer system of claims 2, 11 and 18, as referenced above. The combination does not explicitly disclose identifying a category as idle responsive to a determination that the category includes a zero count in a training dataset associated with the predictive model . However, it would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains to modify the low-correlation categorical levels of Manus as zero-frequency categories as merge candidates as it is a routine preprocessing step known in the art. See KSR International Co. v. Teleflex Inc. , 82 USPQ2d 1385 (U.S. 2007). With respect to claims 4 and 13 , as modified the combination of Chang, Manus and Achi discloses the computer-implemented method and computer program product of claims 2 and 11, as referenced above. The combination does not explicitly discloses it comprises: determining whether the categorical predictor is ordinal; and identifying as a valid pair, responsive to a determination that the categorical predictor is ordinal, two adjacent categories having a category importance value below a predetermined threshold. However, it would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains to modify the categorical levels of Manus to be handled in either an ordered or unordered manner, additionally, choosing adjacent categories for ordinal predictors is known in the art. See KSR International Co. v. Teleflex Inc. , 82 USPQ2d 1385 (U.S. 2007). With respect to claims 5 and 14 , as modified the combination of Chang, Manus and Achi discloses the computer-implemented method and computer program product of claims 4 and 13, as referenced above. The combination does not explicitly discloses identifying as a valid pair, responsive to a determination that the categorical predictor is not ordinal, any two categories having a category importance value below a predetermined threshold [ see claims 4 and 13, it would have been obvious to handle the categorical levels in an unordered (i.e., not ordinal) manner as well ] . With respect to claims 6 and 11 , as modified the combination of Chang, Manus and Achi discloses the computer-implemented method and computer program product of claims 1 and 10, as referenced above. The combination further discloses it comprises: computing a plurality of model accuracy changes for a plurality of categories associated with the categorical predictor [ Achin, see col. 2, lines 1-8, disclosing t he observations are generally partitioned into at least one “training” dataset and at least one “test” dataset. A data analyst then selects a statistical-learning procedure and executes that procedure on the training dataset to generate a predictive model ; The analyst then tests the generated model on the test dataset to determine how well the model predicts the value(s) of the target(s), relative to actual observations of the target(s) ; also, see col. 2, lines 51-54, disclosing reducing (e.g., minimizing) any loss of accuracy associated with moving from a source model to a second-order model (and, in some cases, for generating second-order models with greater accuracy than their source models) ] ; sorting the plurality of model accuracy changes based on change magnitude [ Achin, see col. 61, lines 5-10, disclosing for each of the fitted predictive models: (c1) determining a first accuracy score representing an accuracy with which the fitted model generates predictions for data in the initial dataset; (c2) determining a second accuracy score representing an accuracy with which the fitted model generates predictions for data in the modified dataset in which the multi-model predictive value of the feature has been reduced; and (c3) determining a model-specific predictive value of the feature based on the first and second accuracy scores of the fitted model; and (d) determining, based on the model-specific predictive values of the feature, that the multi-model predictive value of the feature is low; performing feature engineering on the initial dataset based on the multi-model predictive value of the feature, including pruning the feature having the low multi-model predictive value from the initial dataset, thereby generating a pruned dataset; and building a predictive model for the prediction problem ] ; identifying a minimum model accuracy change in the sorted plurality of model accuracy changes [ Achin, see col. 61, lines 5-10, disclosing t he system 100 may classify a model as a “top” model if the accuracy of the model is greater than a threshold accuracy, if the model has one of the N highest accuracy values among the fitted models, if the model does not have one of the M lowest accuracy values among the fitted models, etc ] ; and determining to merge two categories associated with the minimum model accuracy change [ Manus, see ¶0035, disclosing adjusted categorical levels include initial categorical levels whose individual per-level prediction correlation measure satisfies a per-level prediction correlation threshold as well as merged categorical levels each associated with one or more low-correlation categorical levels whose merged per-level prediction correlation measure satisfies a per-level prediction correlation threshold, where the one or more low-correlation categorical levels associated with a particular merged categorical level are each deemed to not be sufficiently statistically significant in predicting the target feature in accordance with the per-level predictive correlation measures of the noted low-correlation categorical levels ] . With respect to claim 7, as modified the combination of Chang, Manus and Achi discloses the computer-implemented method of claim 6, as referenced above. The combination further discloses wherein the minimum model accuracy change is associated with a minimal decrease in accuracy for the predictive model [ Achin, see col. 2, lines 51-54, disclosing reducing (e.g., minimizing) any loss of accuracy associated with moving from a source model to a second-order model (and, in some cases, for generating second-order models with greater accuracy than their source models) ] . With respect to claim 8, as modified the combination of Chang, Manus and Achi discloses the computer-implemented method of claim 6, as referenced above. The combination further discloses wherein the minimum model accuracy change is associated with a maximum increase in accuracy for the predictive model [ Achin, see col. 61, lines 5-10, disclosing t he system 100 may classify a model as a “top” model if the accuracy of the model is greater than a threshold accuracy, if the model has one of the N highest accuracy values among the fitted models, if the model does not have one of the M lowest accuracy values among the fitted models, etc ] . With respect to claims 9, 16 and 20 , as modified the combination of Chang, Manus and Achi discloses the computer-implemented method, computer program product and computer system of claims 1, 10 and 17, as referenced above. The combination further discloses it comprises: identifying, by the post-modeling category merging engine, a plurality of hybrid valid pairs associated with the hybrid category of the categorical predictor [ Manus, see Abstract, disclosing performing predictive data analysis that utilize at least one of categorical level merging, mutual-information-based feature filtering, feature-correlation-based feature filtering to generate training data feature value arrangements, as well as training and using categorical input machine learning models trained using the training data feature value arrangements ] ; testing, by the post-modeling category merging engine, another merge strategy for the plurality of hybrid valid pairs to determine another merger that minimizes the loss in accuracy of the predictive model [ Achin, see col. 2, lines 1-8, disclosing t he observations are generally partitioned into at least one “training” dataset and at least one “test” dataset ; A data analyst then selects a statistical-learning procedure and executes that procedure on the training dataset to generate a predictive model ; The analyst then tests the generated model on the test dataset to determine how well the model predicts the value(s) of the target(s), relative to actual observations of the target(s) ; also, see col. 2, lines 51-54, disclosing reducing (e.g., minimizing) any loss of accuracy associated with moving from a source model to a second-order model (and, in some cases, for generating second-order models with greater accuracy than their source models) ] ; and merging, by the post-modeling category merging engine based on the testing, a valid hybrid pair in the plurality of hybrid valid pairs to form another hybrid category [ Manus, see ¶0033, disclosing p er-level predictive correlation measures may be determined for both initial categorical levels as well as merged categorical levels determined by combining groups of two or more initial categorical levels ] . Prior Art Made of Record The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. Yan discloses causal model offline evaluation and learning for personalized decision systems. Ayyadevara discloses reinforcement learning machine learning models for intervention recommendation. Elser et al. discloses database systems and user interfaces for processing discrete data items with statistical models associated with continuous processes. Wang et al. discloses weight of evidence based feature engineering and machine learning. Valipour et al. discloses automated dataset drift detection. Ebrard et al. discloses enhanced mechanisms for predictive estimation in an enterprise environment. Lee et al. discloses interactive interfaces for machine learning model evaluations. Dirac et al. discloses efficient duplicate detection for machine learning data sets. Backhed et al. discloses identification of a person having risk for developing type 2 diabetes. Chu et al. discloses decision tree insight discovery. Conclusions/Points of Contacts An y inquiry concerning this communication or earlier communications from the examiner should be directed to FILLIN "Examiner name" \* MERGEFORMAT JORGE A CASANOVA whose telephone number is FILLIN "Phone number" \* MERGEFORMAT (571)270-3563 . The examiner can normally be reached FILLIN "Work Schedule?" \* MERGEFORMAT M-F: 9 a.m. to 6 p.m. (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, FILLIN "SPE Name?" \* MERGEFORMAT Aleksandr Kerzhner can be reached at FILLIN "SPE Phone?" \* MERGEFORMAT (571) 270-1760 . 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. /JORGE A CASANOVA/ Primary Examiner, Art Unit 2165