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 .
Response to Amendment
In response to the Office Action mailed August 6, 2026, applicant submitted an amendment filed on October 19, 2026, in which the applicant amended and requested reconsideration.
Response to Arguments
Applicants argue that the prior art cited fails to teach the claims as amended. Applicants’ arguments are persuasive, but are moot in view of new grounds of rejection.
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) 1-5 is/are rejected under 35 U.S.C. 103 as being unpatentable over by Chandrasekhar et al. (PGPUB 2020/0151201), hereinafter referenced as Chandrasekhar in view of Manikani et al. (PGPUB 2024/0428138), hereinafter referenced as Manikani.
Regarding claim 1, Chandrasekhar discloses a computer-implemented method for training a domain-specific translation model for product naming conventions, the computer-implemented method comprising:
Prompting standardized product descriptors from a canonical reference ontology (using machine learning models and standardized product taxonomies to normalize product information received from different sources that use inconsistent terminology and taxonomies; p. 0026, 0048, 0057 with fig. 2A, element 208, Fig. 2B, element 212, 220, Fig. 2C, elements 222-228, fig. 1B extraction and inference module 106, fig. 3B product classification tree and fig. 3C attribute lists);
receiving an individual one of the plurality of standardized product descriptors based at least in part on the arbitrary item descriptor (generating normalized product information based on the input product record; p. 0026, 0048, 0057 with fig. 2A, element 208 and fig. 2C)
training, by the computing device, the domain-specific translation model to generate a plurality of mappings of proprietary descriptors to the plurality of standardized product descriptors from the canonical reference ontology (fig. 2A, element 208 and fig. 2C), but does not specifically receiving, by the computing device and from the generative AI model a plurality of synthetic input-output pairs, wherein an input of each of the synthetic input-output pairs comprises a synthetic proprietary descriptor and an output of each of the synthetic input-output pairs comprises an individual one of the plurality of standardized product descriptors and training, by the computing device, the domain-specific translation model to generate a plurality of mappings of proprietary descriptors to the plurality of standardized product descriptors from the canonical reference ontology based on the plurality of synthetic input-output pairs.
Manikani discloses a method comprising:
prompting, by a computing device, a generative AI model with a plurality of standardized product descriptors (prompting a LLM to generate synthetic data instances and corresponding labels from predefined classes; p. 0018-0023, 0040-0041 with fig. 3, fig. 2, element 202 and fig. 1, elements 127-129)
receiving, by the computing device and from the generative AI model a plurality of synthetic input-output pairs, wherein an input of each of the synthetic input-output pairs comprises a synthetic proprietary descriptor and an output of each of the synthetic input-output pairs comprises an individual one of the plurality of standardized product descriptors (prompting the teaching LLM to generate synthetic data, wherein a teaching LLM receiving prompts that can instruct it to generate synthetic instances and corresponding labels distributed across predefined emission classes and domain knowledge base containing reference documents associated with a particular knowledge domain; p. 0027-0031, 0034-0036, 0040, 0071-0085 with fig. 2, element 202-204, figure 1, elements 112, 122, 126-129, fig. 5 and fig. 3 prompting the LLM to generate synthetic instances); and
training, by the computing device, the domain-specific translation model to generate a plurality of mappings of proprietary descriptors to the plurality of standardized product descriptors from the canonical reference ontology based on the plurality of synthetic input-output pairs (p. 0040-0042 with fig. 1, element 114 training engine and element 116 ML model, fig. 2, elements 204-206 with subsequent model training operations), to reduce need for manually labeled product records while increasing the diversity of training examples.
Therefore, it would have been obvious to one of ordinary skill of the art, before the effective filing date of the claimed invention, to modify the method as described above, to provide additional labeled examples representing different product description variations, to improve the training datasets coverage of variations in terminology.
Regarding claim 2, it is interpreted and rejected for similar reasons as set forth above. In addition, Manikani discloses a method further comprising:
validating, by the computing device, the plurality of mappings based on the plurality of synthetic input-output pairs with labeled real-world data (testing dataset is used to evaluate model predictions wherein the dataset can include real enterprise data; p. 0018-0023 with fig. 1, elements 130, 135-135 and fig. 2); and
dividing, by the computing device, the plurality of synthetic input-output pairs and the labeled real-world data into a first group for training the domain-specific translation model and a second group for validating the domain-specific translation model (dividing synthetic data into training and testing datasets and evaluating the model on previously unseen text instances and calculating a model validation score; p. 0060-0064 with figs. 1-2).
Regarding claim 3, it is interpreted and rejected for similar reasons as set forth above. In addition, Manikani discloses a method further comprising:
receiving, by the computing device, a newly generated plurality of synthetic input-output pairs and user feedback on suggested mappings and training, by the computing device, the domain specific translation model to generate a plurality of updated mappings of proprietary descriptors to the plurality of standardized product descriptors from the canonical reference ontology based on the newly generated plurality of synthetic input-output pairs and the user feedback (generating example synthetic data using an LLM, presenting the generated synthetic data to a user, receiving feedback concerning the synthetic data, generating additional synthetic data based on the feedback and using the additional synthetic data to support training classifier; p. 0031-0033 with figs. 1-3).
Regarding claim 4, it is interpreted and rejected for similar reasons as set forth above. In addition, Chandrasekhar discloses a method further comprising:
receiving, by the computing device, real-world data comprising a plurality of additional standardized product descriptors (receiving product records from external data sources and product description taxonomy representative values and equivalent values; abstract with p. 0053); and
preparing, by the computing device, the real-world data for training the domain-specific translation model (sampling, clustering and organizing product records; p. 0050-0061, 0066-0069). In addition, Manikani teaches receiving labeled real world data (real enterprise data; p. 0018-0023 with fig. 1, elements 130, 135-135 with fig. 2) comprising training, by the computing device, the domain specific translation model to generate a plurality of additional mappings of proprietary descriptors to the plurality of standardized product descriptors from the canonical reference ontology based on the labeled real-world data (real world testing instances and corresponding classification labels and training a field ML model using labeled examples; p. 0031-0042 with figs. 1-3).
Regarding claim 5, it is interpreted and rejected for similar reasons as set forth above. In addition, Chandrasekhar discloses a method further comprising:
providing, by the computing device, an arbitrary item descriptor to the domain-
specific translation model and receiving, by the computing device and from the domain-specific translation model, an individual one of the plurality of standardized product descriptors based at least in part on the arbitrary item descriptor (providing an input product record to a machine learning model and receiving a normalized product record based on a standardized product taxonomy. In particular, applying one or more machine learning models to an input product record to generate an output normalized product record according to a product description taxonomy, receiving product records from external data sources that use different, inconsistent product description taxonomies; p. 0026, 0048, 0057 with fig. 2A, element 208, Fig. 2B, element 220, Fig. 2C, elements 222-228 and fig. 1B, extraction and inference module 106).
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). 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 JAKIEDA R JACKSON whose telephone number is (571)272-7619. The examiner can normally be reached Mon - Fri 6:30a-2:30p.
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, Daniel Washburn can be reached at 571.272.5551. 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.
/JAKIEDA R JACKSON/Primary Examiner, Art Unit 2657