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 .
Detailed Action
In amendments dated 7/26/26, Applicant amended claims 1, 4-511, 14-15, canceled claims 3 and 13, and added no new claims. Claims 1-2, 4-12, and 14-20 are presented for examination.
Examiner notes, in specification paragraph 0056 discusses “U.S. Application No. 11409589” titled “Methods and Systems for Determining Stopping Point” as describing techniques for training a machine learning classifier. Said US Application number is actually U.S. Patent Number 11,409,589, which was filed on 10/22/2020 as Application No. 17/077,681.
Rejections under 35 U.S.C. 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 1-2, 4-12, and 14-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to mental processes without significantly more. Independent claims 1 and 11 each recites generating, via one or more processors, a prompt for input to the generative Al model based on prompt criteria defining an inquiry associated with a corpus of documents, wherein the prompt criteria include one or more component fields; generating, via the one or more processors, a classification of an initial set of documents from the corpus of documents by inputting the initial set of documents and the prompt to the generative Al model; evaluating, via the one or more processors, classification performance of the prompt based on ground truth data associated with the initial set of documents, wherein evaluating classification performance of the prompt comprises evaluating classification performance of the one or more component fields of the prompt criteria based on the ground truth data; based on the evaluation, generating, via the one or more processors, one or more modified prompt criteria; generating, via the one or more processors, one or more modified prompts respectively associated with the one or more modified prompt criteria; generating, via the one or more processors, one or more respective classifications of the initial set of documents associated with each of the one or more modified prompts by inputting the initial set of documents and each of the one or more modified prompts to the generative Al model; evaluating, via the one or more processors, classification performance of the one or more modified prompts based on the ground truth data; and based on the evaluation of the one or more modified prompts, selecting, via the one or more processors, a preferred prompt from among the prompt and the one or more modified prompts. Generating a prompt and a modified prompt are recited broadly and are mental processes accomplishable in the human mind or on paper. Generating a classification of a set of documents by inputting the documents and a prompt or a modified prompt into an AI model is merely applying the AI model and is not more significant than a mental process per Recentive Analytics v. Fox Broadcasting Corp. (134 F.4th 1205, 2025 U.S.P.Q.2d 628). Evaluating the classification performance of a prompt or a modified prompt and selecting a preferred prompt are each evaluating and are mental processes. Each claim recites an additional element of providing, via the one or more processors, an indication of preferred prompt criteria associated with the preferred prompt, which is an output step and insignificant extra-solution activity. Claim 11 recites one or more processors and one or more non-transitory memories, which are each generic components of a computer. Examiner notes specification paragraph 0003 describes how attorneys deploy machine learning models in an eDiscovery process to identify documents responsive to an inquiry. Paragraphs 0004 and 0032 describe said deploying of machine learning models can be cumbersome and inefficient due to different attorneys deploying models in different ways creating conflicts in the models and training a classifier on thousands of documents as cumbersome and inefficient. Paragraph 0033-0034 begin discussing techniques to address the shortcomings mentioned in paragraphs 0004 and 0032. The claim steps do not recite a particular improvement in any technology or function of a computer per MPEP 2106.04(d) and do not recite any unconventional steps in the invention per MPEP 2106.05(a). Therefore, the recited mental processes are not integrated into a practical application. Taking the claims as a whole, the output step is recited broadly and amounts to sending data across a network per specification paragraphs 0045-0046 and 0054 and figures 1 and 10, and sending data is routine and conventional per the list of such activities in MPEP 2106.05(d) part II. The one or more processors and one or more non-transitory memories are each still generic components of a computer. Thus the claims do not include additional elements that are sufficient to amount to significantly more than the recited mental processes.
Claims 2 and 12 each recites evaluating, via the one or more processors, classification performance of one or more respective component fields of each of the one or more modified prompt criteria based on the ground truth data, and evaluating classification performance is recited broadly and is a mental process. Claims 4 and 14 each recites based on the evaluation, selecting, via the one or more processors, one or more first preferred component fields from among the one or more component fields of the prompt criteria and one or more second preferred component fields from among the one or more respective component fields of each of the one or more modified prompt criteria, and selecting fields is evaluating and a mental process. Claims 5 and 15 each recites generating, via the one or more processors, one or more modified component fields each corresponding to a component field of the one or more component fields by inputting the prompt and the classification performance of the one or more modified component fields to the generative AI model, and generating modified components by inputting a prompt and a classification performance into an AI model is applying the AI model and is not more significant than a mental process per Recentive Analytics v. Fox Broadcasting Corp. (134 F.4th 1205, 2025 U.S.P.Q.2d 628); and generating, via the one or more processors, the modified prompt criteria based on the one or more modified component fields, and generating modified prompt criteria is recited broadly and a mental process accomplishable in the human mind or on paper.
Claims 6 and 16 each recites wherein the classification performance of the prompt includes one or more of: one or more respective indications of one or more misclassifications of documents from the initial set of documents, or one or more respective indications of one or more low-confidence classifications of documents from the initial set of documents, and the recited indications are each data and are mental processes accomplishable in the human mind or on paper. Claims 7 and 17 each recites determining, via the one or more processors, one or more component fields of the prompt criteria associated with at least one of the one or more misclassifications of documents, and determining is recited broadly and is a mental process; and modifying, via the one or more processors and by the generative AI model, the one or more component fields to generate the one or more modified component fields, and applying the AI model and is not more significant than a mental process per Recentive Analytics v. Fox Broadcasting Corp. (134 F.4th 1205, 2025 U.S.P.Q.2d 628). Claims 8 and 18 each recites determining, via the one or more processors, one or more component fields of the prompt criteria associated with at least one of the one or more low-confidence classifications of documents, and determining associated prompt criteria is evaluating and a mental process; and modifying, via the one or more processors and by the generative AI model, the one or more component fields to generate the one or more modified component fields, and applying the AI model and is not more significant than a mental process per Recentive Analytics v. Fox Broadcasting Corp. (134 F.4th 1205, 2025 U.S.P.Q.2d 628).
Claims 9 and 19 each recites generating, via the one or more processors, the prompt criteria by inputting one or more of: a review protocol, a complaint, a request for production, one or more of key documents, one or more background documents, to a generative AI model, and inputting criteria into an AI model is a mental process accomplishable in the human mind or on paper. Claims 10 and 20 each recites obtaining, via the one or more processors, a preliminary set of documents associated with the inquiry and the corpus of documents, wherein the preliminary set of documents include at least one of: (i) one or more key documents or (ii) one or more background documents, which is a data gathering step and amounts to receiving data across a network per specification paragraphs 0045-0046 and 0054 and figures 1 and 10, and is routine and conventional per the list of such activities in MPEP 2106.05(d) part II; and generating, via the one or more processors, the initial prompt criteria by inputting the preliminary set of documents to the generative AI model, and inputting criteria into an AI model is a mental process accomplishable in the human mind or on paper.
Relevant Prior Art
During his search for prior art, Examiner found the following references to be relevant to Applicant's claimed invention. Each reference is listed on the Notice of References form included in this office action:
Trautmann, Deitrich, "Large Language Model Prompt Chaining for Long Legal Document Classification," teaches use of prompt chaining in classification of long legal documents, does not teach prompt criteria including component fields, ground truth or evaluating classification performance based on ground truth (section 1 Introduction and sections 3-5 Data through Prompt Chaining, pages 1-4); and
Kelly et al (US 20240289560) teaches classifying a subset of documents, using the classification to refine the document subset, generating text fields using an LLM and a prompt document, does not teach modifying prompt criteria or evaluating classification performance based on ground truth (paragraphs 0005, 0025, 0170-0181 figure 8).
Responses to Applicant’s Remarks
Regarding objections to claims 1 and 11 for antecedent basis of “the evaluation,” in view of amendments reciting “the evaluation of the one or more modified prompts,” these objections are withdrawn. Regarding rejections of claims 1-2, 9, 11-12, and 19 under 35 U/.S/C/ 103 by Smith in view of Rankin, Applicant’s amendments overcome Smith’s and Rankin’s teachings, in particular evaluating classification performance of the component fields. Regarding rejections of claims 1-20 under 35 U.S.C. 101 for reciting mental processes without significantly more, Applicant’s arguments have been considered but are not persuasive. On pages 8-11 Applicant lists the amended limitations in claims 1 and 11, discusses Ex Parte Desjardins and Example 48, and asserts these amendments integrate the claims into a practical application. Examiner disagrees as the amended limitations (“generating, via one or more processors, a prompt for input to the generative AI model based on prompt criteria defining an inquiry associated with a corpus of documents, wherein the prompt criteria include one or more component fields; generating, via the one or more processors, a classification of an initial set of documents from the corpus of documents by inputting the initial set of documents and the prompt to the generative AI model; and evaluating, via the one or more processors, classification performance of the prompt based on ground truth data associated with the initial set of documents, wherein evaluating classification performance of the prompt comprises evaluating classification performance of the one or more component fields of the prompt criteria based on the ground truth data;”) are still recited broadly and lack details showing how the invention addresses the drawbacks discussed in specification paragraphs 0004 and 0032, namely machine learning models may be deployed in different ways by different lawyers and a machine learning classifier may require manual review of thousands of documents to become trained enough for use in eDiscovery tasks. For example, these claim limitations broadly recite generating a prompt input based on prompt criteria with no details on how the prompt is generated, generating a classification of an initial set of documents by applying a generative AI model with the generated prompt, and evaluating classification performance of the prompt (and later in the claims, classification performance of the modified prompts) without reciting details on how the invention evaluates said classification performance. As Applicant stated on page 9 of his Remarks, “Under MPEP § 2106.04(d)(1), a claim integrates a judicial exception into a practical application when it improves the functioning of a computer or improves another technology or technical field. To evaluate such an improvement, (1) the specification should provide sufficient details such that one of ordinary skill in the art would recognize the claimed invention as providing an improvement, and (2) the claim itself must reflect the disclosed improvement.” In both Ex Parte Desjardins and Example 48, the claims reflected the improvement, but the claims still do not provide details of the invention that show an improvement, for example how the invention evaluates classification performances and the one or more modified prompts to select a prompt as a “preferred prompt.” Examiner notes support in the specification for evaluating classification performance at least in paragraphs 0052-0056, 0110-0111, and 0117-0124 for figure 12.
Conclusion
THIS ACTION IS MADE FINAL. Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Inquiry
Any inquiry concerning this communication or earlier communications from the examiner should be directed to BRUCE M MOSER whose telephone number is (571)270-1718. The examiner can normally be reached M-F 9a-5p.
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, Boris Gorney can be reached at 571 270-5626. 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.
/BRUCE M MOSER/Primary Examiner, Art Unit 2154
9/12/26
Prosecution Insights