DETAILED ACTION
This Office Action is sent in response to Applicant’s Communication received 11/29/2023 for application number 18/522,866.
Claims 1-20 are pending.
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 .
Information Disclosure Statement
The information disclosure statement filed 6/24/2026 fails to comply with 37 CFR 1.98(a)(2), which requires a legible copy of each cited foreign patent document; each non-patent literature publication or that portion which caused it to be listed; and all other information or that portion which caused it to be listed. NPL entry 3 does not have a legible copy of the publication so that publication could not be considered.
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. Independent claim 1 (representative of independent claims 15 and 18) recites:
A system comprising: at least one memory that stores instructions; and one or more processors configured by the instructions to perform operations comprising: accessing process metadata of an automated process associated with a process design application, the process metadata comprising a plurality of attributes; generating prompt data by adding, to the process metadata, an instruction comprising at least one request to classify the plurality of attributes according to a predetermined classification scheme; providing the prompt data to a machine learning model to obtain output comprising, for each of the plurality of attributes and based on the predetermined classification scheme, a classification result; and in response to obtaining the output from the machine learning model: storing the classification results in association with the automated process, and causing presentation of one or more of the classification results in a user interface.
(2A, prong 1) The underlined portions of the claim recite an abstract idea, specifically a mental process: a human can mentally generate a prompt asking to classify attributes according to a classification scheme (i.e. a human can ask a second human to classify fields of a dataset as being PII or not), and can mentally determine the classifications of each attribute based on the scheme (i.e. the second human can mentally determine if a field is PII or not). Applicant’s specification explicitly states that human can manually categorize attributes (see spec. para. 0002-03 as published).
(2A, prong 2) This judicial exception is not integrated into a practical application. The claims recite the additional elements of [a] generic computer components, [b] accessing process metadata comprising attributes, [c] providing prompt data to a ML model to get output, [d] storing a classification result, and [e] displaying classification results in a UI. Additional element [a] is a mere instruction to apply the exception because it merely adds generic computer hardware to the abstract idea after-the-fact. Additional elements [b], [d], and [e] are insignificant extra-solution activity because the limitations amount to mere necessary data gathering and outputting for the abstract idea. Finally, additional element [c] is a mere instruction to apply the exception because it merely states the idea of a solution (of asking a ML model to classify attributes and getting the classification back) without how the solution is accomplished (i.e. how the ML model operates to take the prompt, attributes, and classification scheme, and outputs a classification for each attribute). Even when all the additional elements are considered in ordered combination with the recited abstract idea, the claim as a whole does not integrate the abstract idea into a practical application because the additional elements only add mere instructions to apply the exception and insignificant extra-solution activity to the abstract idea.
(2B) The claim(s) does/do not include additional elements that are sufficient to amount to significantly more than the judicial exception. Additional elements [a] and [c] are mere instructions to apply the exception, as explained above. Additional elements [b], [d], and [e] are well-understood, routine, and conventional activity, analogous to storing and retrieving information in memory for [b] and [d], see MPEP 2106.05(d) citing Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015), and presenting offers and gathering statistics for [e], see MPEP 2106.05(d) citing OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1362-63, 115 USPQ2d 1090, 1092-93 (Fed. Cir. 2015). Even when all of the additional elements are considered in ordered combination with the recited abstract idea, the claim as a whole does not amount to significantly more than the abstract idea itself because the additional elements only add mere instructions to apply the exception and insignificant extra-solution activity that is well-understood, routine, and conventional to the abstract idea.
For dependent claim 2, this claim recites the metadata includes, for each attribute, a name and data type. This adds to additional element [b] and is insignificant extra-solution activity that is well-understood, routine, and conventional activity as explained above.
For dependent claims 3-4, 16, and 19, these claims recite the classification request is for identifying personal data attributes, and adding a personal data indicator to the metadata. (2A, prong 1) this additional element adds to the mental process; a human can mentally determine a request to classify attributes as personal data and mark, on pen and paper, attributes that are personal data.
For dependent claim 5, this claim recites:
The system of claim 3, wherein the machine learning model is a first machine learning model, the operations further comprising: accessing a first data set comprising the plurality of attributes of the automated process; identifying, in the first data set, a target attribute from among the plurality of attributes, wherein the personal data indicator corresponding to the target attribute indicates that the target attribute is a personal data attribute; performing a data filtering operation to remove the target attribute from the first data set to obtain a second data set, the second data set being a training data set for a second machine learning model; and training the second machine learning model on the training data set.
(2A, prong 1). The underlined portions of the claim add to the mental process. A human can look at a dataset, identify fields marked as personal data, and redact the fields. (2A, prong 2) The additional element [f] of training a second ML model with the filtered data set does not integrate the abstract idea into a practical application because it is a mere instruction to apply the exception because the limitation merely states the idea of a solution (that a ML model is trained on filtered data) and not how the solution is accomplished (i.e. any technical details or limitation on how the ML model is trained). (2B) Additional element [f] does not amount to significantly more than the abstract idea itself because it is a mere instruction to apply the exception, as explained above. Even when all of the additional elements are considered in ordered combination with the recited abstract idea, the claim as a whole does not amount to significantly more than the abstract idea itself because the additional elements only add mere instructions to apply the exception and insignificant extra-solution activity that is well-understood, routine, and conventional to the abstract idea.
For dependent claims 6-9, 17, and 20, these claims recite:
6 / 17/ 20
. The system of claim 1, wherein the at least one request to classify each of the plurality of attributes comprises a second request to identify a category of each attribute from among a plurality of candidate data categories, the classification result for each attribute comprising the category of the attribute as identified by the machine learning model.
7. The system of claim 6, wherein the machine learning model is a first machine learning model, the operations further comprising: accessing a first data set comprising the plurality of attributes of the automated process; performing a data enrichment operation to add the categories identified by the first machine learning model to the first data set to obtain a second data set, the second data set being a training data set for a second machine learning model; and training the second machine learning model on the training data set.
8. The system of claim 7, wherein the at least one request to classify each of the plurality of attributes further comprises a first request to identify whether each attribute is a personal data attribute, the classification result for each attribute comprises a personal data indicator that indicates whether the attribute is a personal data attribute, the operations further comprising: identifying, in the first data set, a target attribute from among the plurality of attributes, wherein the personal data indicator corresponding to the target attribute indicates that the target attribute is a personal data attribute; and performing a data filtering operation to remove the target attribute from the first data set such that the training data set does not include the target attribute.
9. The system of claim 6, wherein the candidate data categories are included in the prompt data, the candidate data categories comprising at least one of: numerical, text, categorical, or date.
The (2A, prong 1). The underlined portions of the claim add to the mental process. A human can classify attributes of a dataset into different categories, including numerical, text, and personal data, add the categories to the dataset, and filter out personal data. (2A, prong 2) The claims recite the additional elements of [f] training a second ML model with the filtered data set, and [g] accessing the dataset. The additional element [f] does not integrate the abstract idea into a practical application because it is a mere instruction to apply the exception because the limitation merely states the idea of a solution (that a ML model is trained on filtered data) and not how the solution is accomplished (i.e. any technical details or limitation on how the ML model is trained). Additional element [g] is insignificant extra-solution activity because the limitations amount to mere necessary data gathering and outputting for the abstract idea, and therefore does not integrate the abstract idea into a practical application. (2B) Additional element [f] does not amount to significantly more than the abstract idea itself because it is a mere instruction to apply the exception, as explained above. Even when all of the additional elements are considered in ordered combination with the recited abstract idea, the claim as a whole does not amount to significantly more than the abstract idea itself because the additional elements only add mere instructions to apply the exception and insignificant extra-solution activity that is well-understood, routine, and conventional to the abstract idea. Additional elements [g] is well-understood, routine, and conventional activity, analogous to storing and retrieving information in memory, see MPEP 2106.05(d) citing Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015). Even when all of the additional elements are considered in ordered combination with the recited abstract idea, the claim as a whole does not amount to significantly more than the abstract idea itself because the additional elements only add mere instructions to apply the exception and insignificant extra-solution activity that is well-understood, routine, and conventional to the abstract idea.
For dependent claims 10-11, these claims recite the additional element [f] displaying the classification of an attribute with an adjustment element, and allowing a user to change the classification. (2A, prong 2) Additional element [f] does not integrate the abstract idea into a practical application because it is mere necessary data gathering and outputting for the abstract idea. (2B) Additional element [f] does not amount to significantly more than the abstract idea itself because it is well-understood, routine, and conventional activity, analogous to presenting offers and gathering statistics, see MPEP 2106.05(d) citing OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1362-63, 115 USPQ2d 1090, 1092-93 (Fed. Cir. 2015). Even when all of the additional elements are considered in ordered combination with the recited abstract idea, the claim as a whole does not amount to significantly more than the abstract idea itself because the additional elements only add mere instructions to apply the exception and insignificant extra-solution activity that is well-understood, routine, and conventional to the abstract idea.
For dependent claims 12-14, these claims recite:
12. The system of claim 1, wherein the instruction identifies a data structure for returning the classification results, the output being structured according to the identified data structure, and the operations further comprising: providing the process design application with programmatic access to the output to integrate the classification results into the process metadata of the automated process.
13. The system of claim 12, wherein the at least one request to classify each of the plurality of attributes comprises a first request to identify whether each attribute is a personal data attribute, the classification result for each attribute comprises a personal data indicator that indicates whether the attribute is a personal data attribute, and the personal data indicators are integrated into the process metadata of the automated process.
14. The system of claim 12, wherein the at least one request to classify each of the plurality of attributes comprises a first request to identify whether each attribute is a personal data attribute, the classification result for each attribute comprises a personal data indicator that indicates whether the attribute is a personal data attribute, and the personal data indicators are used to perform a data filtering operation with respect to one or more of the attributes in the process metadata.
(2A, prong 1) the underlined portions of the claim recite a mental process: a human can manually categorize and mark data as being personal data. (2A, prong 2) the additional element of [f] providing an application with access to the classification output to add the classification to the data does not integrate the abstract idea into a practical application because it is insignificant extra-solution activity that acts as mere necessary data gathering for the abstract idea. (2B) additional element [f] does not amount to significantly more than the abstract idea itself because it is well-understood, routine, and conventional activity, analogous to storing and retrieving information in memory, see MPEP 2106.05(d) citing Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015). Even when all of the additional elements are considered in ordered combination with the recited abstract idea, the claim as a whole does not amount to significantly more than the abstract idea itself because the additional elements only add mere instructions to apply the exception and insignificant extra-solution activity that is well-understood, routine, and conventional to the abstract idea.
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 (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 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Sublett et al. (US 2023/0418978 A1) in view of Zhang et al., Utilising a Large Language Model to Annotate Subject Metadata (NPL [U], see Notice of References Cited) and Broyles et al. (US 2021/0406716 A1).
In reference to claim 1, Sublett teaches a system comprising: at least one memory that stores instructions; and one or more processors configured by the instructions to perform operations (fig. 2) comprising: accessing process metadata of an automated process associated with a process design application, the process metadata comprising a plurality of attributes (document data comprises a plurality of fields, or attributes, para. 0017-19); generating … at least one request to classify the plurality of attributes according to a predetermined classification scheme (request made to classify fields as PHI, para. 0017-22); providing … data to a machine learning model to obtain output comprising, for each of the plurality of attributes and based on the predetermined classification scheme, a classification result (PHI rules are used to identify and classify fields, para. 0021-22); and in response to obtaining the output from the machine learning model: storing the classification results in association with the automated process (result stored so PHI term can be replaced, para. 0022-23).
However, Sublett does not explicitly teach generating prompt data by adding, to the process metadata, an instruction comprising at least one request to classify the plurality of attributes according to a predetermined classification scheme; providing the prompt data to a machine learning model.
Zhang teaches generating prompt data by adding, to the process metadata, an instruction comprising at least one request to classify the plurality of attributes according to a predetermined classification scheme; providing the prompt data to a machine learning model (prompt generated with instructions, classification rules, and target record, which is then submitted to a LLM that can categorize the data, pages 3, 6-9).
It would have been obvious to one of ordinary skill in art, having the teachings of Sublett and Zhang before the earliest effective filing date, to modify the classification rules of Sublett to include the prompt and LLM of Zhang.
One of ordinary skill in the art would have been motivated to modify the classification rules of Sublett to include the prompt and LLM of Zhang because it can annotate metadata more easily than classical ML approaches and humans (Zhang, page 4).
However, Sublett and Zhang do not explicitly teach causing presentation of one or more of the classification results in a user interface.
Broyles teaches causing presentation of one or more of the classification results in a user interface (para. 0052-53, 0066).
It would have been obvious to one of ordinary skill in art, having the teachings of Sublett, Zhang, and Broyles before the earliest effective filing date, to modify the system of Sublett to include the interface of Broyles.
One of ordinary skill in the art would have been motivated to modify the system of Sublett to include the interface of Broyles because it allows users to verify and correct classifications (Broyles, para. 0066).
In reference to claim 2, Broyles further teaches the system of claim 1, wherein the process metadata comprises, for each of the plurality of attributes, an attribute name and an attribute data type (field category, which is a name, para. 0038, and field data type, para. 0029, 0033).
In reference to claim 3, Sublett further teaches the system of claim 1, wherein the at least one request to classify each of the plurality of attributes comprises a first request to identify whether each attribute is a personal data attribute, and the classification result for each attribute comprises a personal data indicator that indicates whether the attribute is a personal data attribute (PHI rules are used to identify and classify fields as personal information, para. 0021-22).
In reference to claim 4, Sublett further teaches the system of claim 3, the operations further comprising: performing a data enrichment operation to add one or more of the personal data indicators to the process metadata (PHI classification is stored so the field can be redacted, para. 0021-23).
In reference to claim 5, Sublett further teaches the system of claim 3, wherein the machine learning model is a first machine learning model, the operations further comprising: accessing a first data set comprising the plurality of attributes of the automated process; identifying, in the first data set, a target attribute from among the plurality of attributes, wherein the personal data indicator corresponding to the target attribute indicates that the target attribute is a personal data attribute (personal health information is identified, para. 0021-22); performing a data filtering operation to remove the target attribute from the first data set to obtain a second data set (PHI is redacted, para. 0023), the second data set being a training data set for a second machine learning model; and training the second machine learning model on the training data set (the redacted document is then used for training another ML model, para. 0024-26, 0005-10).
In reference to claim 6, Sublett further teaches the system of claim 1, wherein the at least one request to classify each of the plurality of attributes comprises a second request to identify a category of each attribute from among a plurality of candidate data categories, the classification result for each attribute comprising the category of the attribute as identified by the machine learning model (request is to categorize fields as personal health information or not PHI, which are a plurality of categories, and classification is the category of PHI or not, para. 0017-22).
In reference to claim 7, Sublett further teaches the system of claim 6, wherein the machine learning model is a first machine learning model, the operations further comprising: accessing a first data set comprising the plurality of attributes of the automated process; performing a data enrichment operation to add the categories identified by the first machine learning model to the first data set to obtain a second data set, the second data set being a training data set for a second machine learning model; and training the second machine learning model on the training data set (data redaction, which is enrichment, is used to create a second data set, para. 0021-23, and the second data set is used to train another ML model, para. 0024-26, 0005-10).
In reference to claim 8, Sublett further teaches the system of claim 7, wherein the at least one request to classify each of the plurality of attributes further comprises a first request to identify whether each attribute is a personal data attribute, the classification result for each attribute comprises a personal data indicator that indicates whether the attribute is a personal data attribute, the operations further comprising: identifying, in the first data set, a target attribute from among the plurality of attributes, wherein the personal data indicator corresponding to the target attribute indicates that the target attribute is a personal data attribute; and performing a data filtering operation to remove the target attribute from the first data set such that the training data set does not include the target attribute (data redaction, which is enrichment, is used to create a second data set, para. 0021-23, and the second data set is used to train another ML model, para. 0024-26, 0005-10).
In reference to claim 9, Zhang further teaches the system of claim 6, wherein the candidate data categories are included in the prompt data, the candidate data categories comprising at least one of: numerical, text, categorical, or date (categorical classifications are in prompt, pages 6-8).
In reference to claim 10, Broyles further teaches the system of claim 1, wherein the user interface is a graphical user interface of the process design application, the operations further comprising: receiving, via the user interface, the process metadata from a user device associated with a user of the process design application (user can upload documents that include metadata, para. 0016-18); causing presentation, in the user interface, of a classification adjustment element in association with the classification result obtained for a first attribute from among the plurality of attributes; detecting user selection of the classification adjustment element; and adjusting the classification result obtained for the first attribute based on the user selection of the classification adjustment element (classification results are displayed and user can change / correct classification, para. 0052-53, 0066).
In reference to claim 11, Broyles further teaches the system of claim 10, the operations further comprising: receiving, via the user interface, an additional user selection of a replacement classification for the first attribute, wherein the classification result obtained for the first attribute is adjusted by replacing the classification result with the replacement classification (classification results are displayed and user can change / correct classification, para. 0052-53, 0066).
In reference to claim 12, Sublett further teaches the system of claim 1, wherein the instruction identifies a data structure for returning the classification results, the output being structured according to the identified data structure (classification result is structured with PHI terms and locations, so that the replacement can replace PHI with contextually relevant terms at the correct location, para. 0016-19), and the operations further comprising: providing the process design application with programmatic access to the output to integrate the classification results into the process metadata of the automated process (redaction program can access the output in order to redact document so the document may be used as training data, para. 0016-19).
In reference to claim 13, Sublett further teaches the system of claim 12, wherein the at least one request to classify each of the plurality of attributes comprises a first request to identify whether each attribute is a personal data attribute, the classification result for each attribute comprises a personal data indicator that indicates whether the attribute is a personal data attribute, and the personal data indicators are integrated into the process metadata of the automated process (classification is for personal health information, and redaction program can accesses the output in order to redact document, para. 0016-19).
In reference to claim 14, Sublett further teaches the system of claim 12, wherein the at least one request to classify each of the plurality of attributes comprises a first request to identify whether each attribute is a personal data attribute, the classification result for each attribute comprises a personal data indicator that indicates whether the attribute is a personal data attribute, and the personal data indicators are used to perform a data filtering operation with respect to one or more of the attributes in the process metadata (classification is for personal health information, and redaction program can accesses the output in order to redact document, para. 0016-19).
In reference to claim 15, this claim is directed to a method associated with the system claimed in claim 1 and is therefore rejected under a similar rationale.
In reference to claim 16, this claim is directed to a method associated with the system claimed in claim 3 and is therefore rejected under a similar rationale.
In reference to claim 17, this claim is directed to a method associated with the system claimed in claim 6 and is therefore rejected under a similar rationale.
In reference to claim 18, this claim is directed to a non-transitory computer-readable medium associated with the system claimed in claim 1 and is therefore rejected under a similar rationale.
In reference to claim 19, this claim is directed to a method associated with the system claimed in claim 3 and is therefore rejected under a similar rationale.
In reference to claim 20, this claim is directed to a method associated with the system claimed in claim 6 and is therefore rejected under a similar rationale.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. The other cited references generally teach background information on ML for categorizing data / detecting PII in data.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Andrew T. Chiusano whose telephone number is (571)272-5231. The examiner can normally be reached M-F, 10am-6pm.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Tamara Kyle can be reached at 571-272-4241. 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.
/ANDREW T CHIUSANO/ Primary Examiner, Art Unit 2144
Prosecution Insights