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 .
This Non-Final Office Action is in response to the application 17/806,530 filed on 02/23/2026.
Status of claims:
Claims 1-2, 8-9, and 15-16 are amended in this Office Action.
Claims 1-20 are pending in this Office Action.
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and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114.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 02/10/2026 has been entered.
Response to Arguments
Objections to drawings
Applicant indicated in the remarks filed on 01/30/2026 (page 9) that the applicant filed FIG. 2 in accordance with 37 CFR 1.121(d). However, the examiner is unable to locate the newly submitted drawing, therefore the objection made to the drawing is maintained.
Rejection of claims under 35 USC § 101
Applicant’s arguments filed on 01/30/2026 (pages 10-13) regarding claim
rejections under 35 U.S.C 101 and the amendments submitted have been fully
considered. The rejections made under 35 U.S.C 101 in the previous office action are
now withdrawn after considering the applicant’s remarks and amendments.
Rejection of claims under 35 U.S.C. §103
Applicant’s arguments filed on 01/30/2026 (pages 13-15) regarding claim
rejections under 35 U.S.C 103 have been fully considered. However, after further
examination, new grounds of rejection are presented necessitated by applicant’s
amendments.
Drawings
The drawings are objected to as failing to comply with 37 CFR 1.84(p)(4) because reference element “300” has been used to designate both “a data facet generation system” of fig.2 and “an operational flowchart illustrating the steps of a method” of fig. 3. The examiner submits that element 300 in fig.2 should be corrected to element 200.
Corrected drawing sheets in compliance with 37 CFR 1.121(d) are required in reply to the Office action to avoid abandonment of the application. Any amended replacement drawing sheet should include all of the figures appearing on the immediate prior version of the sheet, even if only one figure is being amended. Each drawing sheet submitted after the filing date of an application must be labeled in the top margin as either “Replacement Sheet” or “New Sheet” pursuant to 37 CFR 1.121(d). If the changes are not accepted by the examiner, the applicant will be notified and informed of any required corrective action in the next Office action. The objection to the drawings will not be held in abeyance.
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.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-5, 8-12, and 15-19 are rejected under 35 U.S.C. 103 as being unpatentable over Orsini (US PGPUB 20210224238) “Orsini” in view of Modarresi et al. (US PGPUB 20190286747) “Modarresi”, Kandel et al. (US Patent 10824606) “Kandel”, and Ryan et al. (US PGPUB 20190379589) “Ryan”.
Regarding claim 1, Orsini teaches a method of data facet generation, executable by a processor, comprising: receiving data associated with a dataset, wherein the received data includes one or more data entries having one or more elements ([0077]: Operations of the Contract Forest system may include 802 receiving data representing information stored at a node of a distributed network. For example, the data received can be the source data that is located via the IPCF address (210) shown in FIG. 2. In this example, the received data is the Data File-A 212. In some examples, the information is PII information and/or is cryptographically secured, digitally signed, and/or authenticated. Data can contain occurrence of data fields,e.g., “kW”, “kWh”… Examiner’s note: Thus, data is received with entries having one or more elements such as occurrence of data fields); associating the one or more elements with one or more data types ([0077]: Operations of the Contract Forest system may include analyzing received data in a semantic manner to identify one or more data types reflected in received data… Examiner’s note: Thus, the received data can associate with a data types); generating one or more data facets for each of the data entries with the received data based on the associated data type ([0077]: The system analyzes received data in a semantic manner to identify one or more data types reflected in received data. The semantic manner can identify the received data by determining a word graph of one or more sequences of data fields embedded in the received data. In the example above, the semantic manner identifies the received data type to represent energy data because of the occurrence of data fields (e.g., “kW”, “kWh”) in the received data… Examiner’s note: Thus, the identified data type can be equivalent to a data facet); and generating one or more transformations for the data facet associated with the dataset ([0078] Operations of the Contract Forest system may include selecting one or more transforms based on the one or more data types of the received data and data types used by the one or more transforms. The Node-A can select the transforms that are related to the one or more data types of the received data. For example, transforms related to energy data are selected when the received data is identified as energy data, transforms related to financial data are selected when the received data is identified as financial data, etc.).
Orsini does not explicitly teach generating one or more transformations for the data facet corresponding to a machine learning task associated with the dataset and displaying a generated recommendation of one or more transformations on a user interface; storing the data facets and one or more transformations as metadata; and applying the one or more transformations via a generated code based on the metadata and machine learning task wherein the generated code comprises a variance explanation selected based on an optimal transformation associated with the machine learning task.
Modarresi teaches generating one or more transformations for the data facet corresponding to a machine learning task associated with the dataset ([0050] The latent classes generated by the clustering module are then employed by a conversion module to convert the categorical data into numerical data . Each cluster represents a respective latent class, which are assigned a respective numerical value by the conversion module. These numerical values (one for each cluster) are then used by the conversion module to convert the categorical variables of the categorical data into numerical data based on membership of classes within respective clusters…Fig. 3 & 5, [0057]: The numerical data is then processed using a machine learning module and a result of the processing is output… Examiner’s note: Thus, a cluster of data can be equivalent to a data facet. The system converts categorical data into numerical data that belong to each cluster can be equivalent to generating one or more transformations for the data facet and the numerical data is subsequently used to process in a machine learning module); storing the data facets and one or more transformations as metadata (Fig. 2 & [0036]: The service provider system 102 in this example is configured to employ a data transformation module 122 to transform the categorical data 108 into numerical data 124 that is acceptable for processing by the model 120 as part of machine learning... [0038]The data transformation module 122 also employs a clustering module 126 to generate latent classes from the numerical data by forming clusters 128, e.g., using K-means clustering, Silhouette model, and so forth… [0050] The latent classes 204 generated by the clustering module 126 are then employed by a conversion module 218 to convert the categorical data 108 into numerical data 124… Examiner’s note: Data relating to cluster/data facet and transformed data such as numerical data can be stored within the service provider system and both data can be considered as metadata because it can be used to describe an input data); and applying the one or more transformations via a generated code based on the metadata and machine learning task ([0036]: The machine learning module 118 and model 120 are typically configured to accept numerical data… Accordingly, the service provider system 102 in this example is configured to employ a data transformation module 122 to transform the categorical data 108 into numerical data 124 that is acceptable for processing by the model 120 as part of machine learning… [0050] The latent classes 204 generated by the clustering module 126 are then employed by a conversion module 218 to convert the categorical data 108 into numerical data 124 (block 308). Continuing with the above example, each cluster 128 represents a respective latent class 204, which are assigned a respective numerical value by the conversion module 218. These numerical values (one for each cluster) are then used by the conversion module 218 to convert the categorical variables of the categorical data 108 into numerical data 124 based on membership of classes within respective clusters… Examiner’s note: The transformed data such as numerical data is applied to a machine learning model that can based on its’ cluster/ data facet and format to be accepted be the machine learning model). 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 incorporate the Modarresi teachings in the Orsini system. Skilled artisan would have been motivated to incorporate transforming data corresponding to a machine learning task taught by Modarresi in the Orsini system to improve operation of a computing device to support efficient and accurate use of categorical data as part of machine learning and may do so in real time, as recognized by Modarresi ([0019]). This close relation between both of the references highly suggests an expectation of success.
Orsini in view of Modarresi does not explicitly teach displaying a generated recommendation of one or more transformations on a user interface; and wherein the generated code comprises a variance explanation selected based on an optimal transformation associated with the machine learning task.
Kandel teaches displaying a generated recommendation of one or more transformations on a user interface (Col 7 line 27-47: FIG. 3A shows architecture of a client application for interacting with a data preprocessing system for developing transformation scripts for preprocessing data, according to an embodiment… The user interface presents information describing the dataset to the user and provides various widgets for allowing users to interact with the data. The information describing the dataset includes a textual representation of the data as well as charts describing the data. The user interface presents transformation recommendations to the user and receives selections of transformation recommendations. The user interface also allows users to enter transformations manually via a text box… Examiner’s note: Thus, the system interacts with a user by providing transformation recommendations of data to the user). 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 incorporate the Kandel teachings in the Orsini and Modarresi system. Skilled artisan would have been motivated to incorporate providing transformation recommendations to the user taught by Kandel in the Orsini and Modarresi system to improve user experience with the system and improve the automation of the system by providing the user with transformation recommendations, rather than enforcing the user to manually select the optimal transformations. This close relation between the references highly suggests an expectation of success.
Orsini in view of Modarresi and Kandel does not explicitly teach wherein the generated code comprises a variance explanation selected based on an optimal transformation associated with the machine learning task.
Ryan teaches generated code comprises a variance explanation selected based on an optimal transformation associated with the machine learning task (Fig. 9 & [0098]: Key Performance Indicators (KPIs) include Accuracy, confusion matrix (False Positive Rate, False Negative rate), or functions of these…[0100] FIG. 9 is a flowchart showing a method 130 of training to select a single best transformation. Every data transformation is evaluated with the same hyper-parameters given to the machine learning algorithm and the best transformation is chosen for the classification. Note that each training pipeline can be performed in parallel… [0101] The method 130 includes preparing the training data (step 132) and copying the training data into data streams (step 134). In parallel, the method 130 includes performing transformation #1-4 (blocks 136-1 through 136-4), training the machine learning algorithm (blocks 138-1 through 138-4), and validating and saving the model KPIs (blocks 140-1 through 140-4)… Examiner’s note: The system determines the optimal transformation out of multiple transformations to feed into the machine learning model wherein values related to the transformations are created and saved such as KPI of the processed transformation. Thus, KPI can correspond to a variance explanation selected based on transformation associated with the machine learning task). 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 incorporate the Ryan teachings in the Orsini, Modarresi, and Kandel system. Skilled artisan would have been motivated to incorporate providing evaluation or explanation for a transformation of data in a machine learning model taught by Ryan in the Orsini, Modarresi, and Kandel system to provide further information or analysis for the recommended transformation, thus can improve user experience with data transformations. This close relation between the references highly suggests an expectation of success.
Regarding claim 2, Orsini in view of Modarresi, Kandel and Ryan teaches all of the limitations of claim 1. Orsini in view of Modarresi does not explicitly teach providing a recommendation to a user based on the generated transformation, wherein the provided recommendation includes generated computer code corresponding to an optimal transformation associated with the machine learning task.
Kandel teaches providing a recommendation to a user based on the generated transformation, wherein the provided recommendation includes generated computer code corresponding to an optimal transformation associated with the machine learning task (Col 7 line 27-47: FIG. 3A shows architecture of a client application for interacting with a data preprocessing system for developing transformation scripts for preprocessing data, according to an embodiment… The user interface presents information describing the dataset to the user and provides various widgets for allowing users to interact with the data. The information describing the dataset includes a textual representation of the data as well as charts describing the data. The user interface presents transformation recommendations to the user and receives selections of transformation recommendations. The user interface also allows users to enter transformations manually via a text box… Examiner’s note: Thus, the system interacts with a user by providing transformation recommendations of data to the user wherein the data can be in a textual representation which can be equivalent to a computer code ). Please refer to claim 1 for the motivational statement.
Regarding claim 3, Orsini in view of Modarresi, Kandel and Ryan teaches all of the limitations of claim 2. Orsini further teaches wherein the optimal transformation is determined based on historic transformation data and metadata corresponding to the dataset ([0083]: Analyzing the received data in a semantic manner comprises determining one or more sequences of data fields (e.g., energy fields such as “kW”, “kWh”, etc., medical fields such as “weight”, “height”, etc., financial fields such as “currency”, “money”, etc.) embedded in the received data using a machine learning system trained on historical data stored at the node. The machine learning system can be trained to identify the received data based on historical data with a known data type (e.g., known energy data is used to train the machine learning system by identifying most likely data fields associated with a particular data type. … Examiner’s note: Thus, the data fields can be equivalent to metadata and historical data can be equivalent to historic transformation data and both can be used in the process to perform a data transformation).
Orsini does not explicitly teach wherein the optimal transformation is determined based on the machine learning task.
Modarresi teaches wherein the optimal transformation is determined based on the machine learning task ([0036] The machine learning module and model are typically configured to accept numerical data. However, in practice, categorical data is often also included as part of data to be processed by the model , e.g., user data in this example. The user data , for instance, may include numerical data such as “age” as well as categorical data such as “city of residence.” Accordingly, the service provider system in this example is configured to employ a data transformation module to transform the categorical data into numerical data that is acceptable for processing by the model as part of machine learning… Examiner’s note: Thus, the system transforms data to a desired format based on the machine learning task). Please refer to claim 1 for the motivational statement.
Regarding claim 4, Orsini in view of Modarresi, Kandel and Ryan teaches all of the limitations of claim 3. Orsini further teaches wherein the metadata corresponds to one or more from among a business application, a user profile, and an internal code repository ([0083]: Analyzing the received data in a semantic manner comprises determining one or more sequences of data fields (e.g., energy fields such as “kW”, “kWh”, etc., medical fields such as “weight”, “height”, etc., financial fields such as “currency”, “money”, etc.) embedded in the received data using a machine learning system trained on historical data stored at the node… Examiner’s note: Thus, the system can at least determine data fields or metadata that relate to business application. For example, fields or metadata such as “kW”, “kWh” is detected to be in a business application of energy).
Regarding claim 5, Orsini in view of Modarresi, Kandel and Ryan teaches all of the limitations of claim 3. Orsini further teaches generating a ranked list of historic
transformations from the historic transformation data based on matching a similarity between the one or more data facets and the metadata ([0083]: The machine learning system can be trained to identify the received data based on historical data with a known data type (e.g., known energy data is used to train the machine learning system by identifying most likely data fields associated with a particular data type)… Examiner’s note: Thus, the system can depend on a ranked historical data such as identifying most likely data fields associated with a particular data type to identify data type for the received data).
Regarding claim 8, note the rejections of claim 1. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Regarding claim 9, note the rejections of claim 2. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Regarding claim 10, note the rejections of claim 3. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Regarding claim 11, note the rejections of claim 4. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Regarding claim 12, note the rejections of claim 5. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Regarding claim 15, note the rejections of claim 1. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Regarding claim 16, note the rejections of claim 2. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Regarding claim 17, note the rejections of claim 3. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Regarding claim 18, note the rejections of claim 4. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Regarding claim 19, note the rejections of claim 5. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Claims 6, 13, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Orsini (US PGPUB 20210224238) “Orsini” in view of Modarresi et al. (US PGPUB 20190286747) “Modarresi”, Kandel et al. (US Patent 10824606) “Kandel”, and Ryan et al. (US PGPUB 20190379589) “Ryan” and Du et al. (US PGPUB 20220405314) “Du”.
Regarding claim 6, Orsini in view of Modarresi, Kandel and Ryan teaches all of the limitations of claim 1. Orsini in view of Modarresi, Kandel and Ryan does not explicitly teach receiving a natural language input from the user, wherein the natural language input corresponds to a selection of a transformation from among the one or more generated transformations.
Du teaches receiving a natural language input from the user ([0035] A natural language query or input may be provided via an application 120 operating on the user device 110. In this regard, the user device 110, via an application 120, might allow a user to input, select, or otherwise provide a natural language query), wherein the natural language input corresponds to a selection of a transformation from among the one or more generated transformations ([0106]: An intent probability that is higher indicating a stronger similarity to the natural language query may be selected, while a popularity probability may be selected when the intent probability is lower. As such, the data visualization engine adjusts to provide recommendations more relevant to an input user query or recommendations reflecting a higher popularity. For example, using this score, exact matches to the user query are generally ranked the highest, followed by recommendations that contain trending data, design, and insight intents. Accordingly, in cases that a user is specific on the intent, a high intent probability will exist and more likely result in a desired data visualization. On the other hand, in cases that a user is not specific in intent, the user will more likely be presented with a popular visualization… Examiner’s note: Thus, a natural language input can correspond to a selection of data among a group of data). 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 incorporate the Du teachings in the Orsini in view of Modarresi, Kandel and Ryan system. Skilled artisan would have been motivated to incorporate inputting a natural language select taught by Du in the Orsini in view of Modarresi, Kandel and Ryan system to improve user experience and easier to use for a wider range of people. This close relation between both of the references highly suggests an expectation of success.
Regarding claim 13, note the rejections of claim 6. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Regarding claim 20, note the rejections of claim 6. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Claims 7 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Orsini (US PGPUB 20210224238) “Orsini” in view of Modarresi et al. (US PGPUB 20190286747) “Modarresi”, Kandel et al. (US Patent 10824606) “Kandel”, and Ryan et al. (US PGPUB 20190379589) “Ryan” and Silberman et al. (US PGPUB 20170330058) “Silberman”.
Regarding claim 7, Orsini in view of Modarresi, Kandel and Ryan teaches all of the limitations of claim 1. Orsini in view of Modarresi, Kandel and Ryan does not explicitly teach debiasing the received data based on modifying weight values associated with each of the elements of the data facet.
Silberman teaches debiasing the received data based on modifying weight values associated with each of the elements of the data facet ([0061]: A weight may be assigned to each event in a timeline… [0062]: Individual weights may be grouped into categories based on a corresponding reference. The categories may include particular demographics, a particular location (e.g., same zip code, adjacent zip codes, same city, same state, or the like), particular financial events, and the like…[0068]: Each category may be selected and the weight of the applicant's event may be compared with a corresponding weight of the reference. If the weight of the applicant's event differs from the weight of the reference by more than a threshold amount, then the category is flagged as potentially causing bias in decisions…[0071]: The weights assigned to each category and sub-category may be repeatedly modified over time to further reduce bias… Examiner’s note: Thus, a dataset can be assigned with weights according to the data it contained and biases can be corrected or reduced by modifying the weights). 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 incorporate the Silberman teachings in the Orsini in view of Modarresi, Kandel and Ryan system. Skilled artisan would have been motivated to incorporate applying weights to data and adjust bias based on weights taught by Silberman in the Orsini in view of Modarresi, Kandel and Ryan system to effectively reduce bias in datasets, thus improves the quality of data and improves the accuracy of the results in a system. This close relation between both of the references highly suggests an expectation of success.
Regarding claim 14, note the rejections of claim 7. The instant claims recite substantially same limitations as the above-rejected claims and are therefore rejected under the same prior-art teachings.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CAO DANG VUONG whose telephone number is (571)272-1812. The examiner can normally be reached M-F 7:30-5 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, Kavita Stanley can be reached at (571) 272-8352. 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.
/C.D.V./ Examiner, Art Unit 2153 04/17/2026
/KAVITA STANLEY/ Supervisory Patent Examiner, Art Unit 2153
Prosecution Insights