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 .
Status of Claims
This office action is in response to arguments and amendments entered on February 10, 2026 for the patent application 18/543,458 originally filed on December 18, 2023. Claims 1 and 11 are amended. Claims 8, 9, 18, and 19 are canceled. Claims 1-7, 10-17, and 20 remain pending. The first office action of March 27, 2024, the second office action of August 12, 2024, the third office action of December 19, 2024, the fourth office action of July 25, 2025, and the fifth office action of November 25, 2025 are fully incorporated by reference into this office action.
Response to Amendment
Applicant’s amendments to the claims have been noted by the Examiner.
Applicant has canceled claim limitations previously rejected under 35 USC 112(a). Accordingly, the outstanding rejections to the claims under 35 USC 112(a) are withdrawn.
The cancelation of claim limitations previously rejected under 35 USC 112(a) obviates the issues rejected under 35 USC 112(b). Therefore, the outstanding 35 USC 112(b) rejections are withdrawn.
The Applicant’s amendments are not sufficient to overcome the outstanding 35 USC 101 rejections, for reasons set forth below.
Applicant’s amendments are sufficient to overcome the outstanding rejections under 35 USC 103. However, new rejections under 35 USC 103 are set forth below in response to findings of new prior art.
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 1-7, 10-17, and 20 are rejected under 35 U.S.C. § 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Claim 1 is directed to “an apparatus” (i.e. a machine) and claim 11 is directed to “a method” (i.e. a process), hence the claims are directed to one of the four statutory categories (i.e. process, machine, manufacture, or composition of matter). In other words, Step 1 of the subject-matter eligibility analysis is “Yes.”
However, the claims are drawn to an abstract idea of “mediating multiparty communication,” in the form of “mental processes,” in terms of processes that can be performed in the human mind (including an observation, evaluation, judgement or opinion) which are “performed on a computer” (per MPEP 2106(III)(C) “A Claim That Requires a Computer May Still Recite a Mental Process”).
Regardless, the claims are reasonably understood as “mental processes,” which require the following limitations (as in exemplary claim claim 11):
“receiving a discussion topic;
…receiving prompt generation training data, wherein the prompt generation training data correlates a plurality of discussion topics to a plurality of model prompts;
…generating a first prompt as a function of the discussion topic using a prompt generation machine learning model comprising:
receiving the prompt generation training data;
training, iteratively, the prompt generation machine learning model using the prompt generation training data; and
generating the first prompt as a function of the discussion topic using the trained prompt generation machine learning model;
…presenting to a plurality of users the first prompt…
…training a score generation machine learning model, wherein training the score generation machine learning model further comprises:
receiving a training dataset comprises model discussion data and model discussion topics correlated to model user understanding scores; and
training the score generation machine learning model using the received training dataset; and
…receiving… a discussion datum, wherein the discussion datum comprises a response by a user of the plurality of users to the first prompt; and
…generating a user understanding score as a function of the trained score generation machine learning model, wherein generating the user understanding score further comprises:
inputting the first prompt generated as a function of the prompt machine learning model and the discussion datum into the trained score generation machine learning model;
outputting, as a function of the trained score generation machine learning model, the user understanding score;
normalizing the user understanding score relative to a plurality of user understanding scores associated with other users participating in the discussion topic; and
generating a certainty score associated with the normalized user understanding score, wherein, when the certainty score fails to satisfy a predefined certainty score threshold, … prompt to acquire additional discussion data for evaluating the user.”
These limitations simply describe a process of data gathering and manipulation, which is partially analogous to “collecting information, analyzing it, and displaying certain results of the collection analysis” (i.e. Electric Power Group, LLC, v. Alstom, 830 F.3d 1350, 119 U.S.P.Q.2d 1739 (Fed. Cir. 2016)). Hence, these limitations are akin to an abstract idea which has been identified among non-limiting examples to be an abstract idea. In other words, Step 2A, Prong 1 of the subject-matter eligibility analysis is “Yes.”
Furthermore, the claims do not include additional elements that either alone or in combination are sufficient to claim a practical application because to the extent that, e.g., “an apparatus,” “a processor,” “a memory,” “a dedicated hardware unit,” and “a user device” are claimed, as these are merely claimed to add insignificant extra-solution activity to the judicial exception (e.g., data gathering) and/or do no more than generally link the use of a judicial exception to a particular technological environment or field of use. In other words, the claimed “mediating multiparty communication,” is not providing a practical application, thus Step 2A, Prong 2 of the subject-matter eligibility analysis is “No.”
Likewise, the claims do not include additional elements that either alone or in combination are sufficient to amount to significantly more than the judicial exception because to the extent that, e.g., “an apparatus,” “a processor,” “a memory,” “a dedicated hardware unit,” and “a user device” are claimed these are all generic, well-known, and conventional computing elements. As evidence that these are generic, well-known, and conventional computing elements, Applicant’s specification discloses them in a manner that indicates that the additional elements are sufficiently well-known that the specification does not need to describe the particulars of such additional elements to satisfy 35 U.S.C. § 112(a), per MPEP § 2106.07(a) III (a), which satisfies the Examiner’s evidentiary burden requirement per the Berkheimer memo.
Specifically, the Applicant’s claimed “apparatus” is described in paragraph [0009] as follows: “Referring now to FIG. 1, an exemplary embodiment of an apparatus 100 for synchronous learning is illustrated. Apparatus 100 may include a computing device. Apparatus 100 may include a processor. Processor may include, without limitation, any processor described in this disclosure. Processor may be included in a computing device. Computing device may include any computing device as described in this disclosure, including without limitation a microcontroller, microprocessor, digital signal processor (DSP) and/or system on a chip (SoC) as described in this disclosure. Computing device may include, be included in, and/or communicate with a mobile device such as a mobile telephone or smartphone. Computing device may include a single computing device operating independently, or may include two or more computing device operating in concert, in parallel, sequentially or the like; two or more computing devices may be included together in a single computing device or in two or more computing devices.” This element is reasonably interpreted as a generic computer which provides no details of anything beyond ubiquitous standard equipment. It is further noted that “a processor” and “a memory” are conventional components of all computing devices.
Applicant’s claimed “a dedicated hardware unit” is described in instant specification paragraph [0139] as “a hardware component, circuit, or the like, aside from a principal control circuit and/or processor performing method steps as described in this disclosure, that is specifically designated or selected to perform one or more specific tasks and/or processes described in reference to this figure, such as without limitation preconditioning and/or sanitization of training data and/or training a machine-learning algorithm and/or model. A dedicated hardware unit 336 may include, without limitation, a hardware unit that can perform iterative or massed calculations, such as matrix- based calculations to update or tune parameters, weights, coefficients, and/or biases of machine- learning models and/or neural networks, efficiently using pipelining, parallel processing, or the like; such a hardware unit may be optimized for such processes by, for instance, including dedicated circuitry for matrix and/or signal processing operations that includes, e.g., multiple arithmetic and/or logical circuit units such as multipliers and/or adders that can act simultaneously and/or in parallel or the like. Such dedicated hardware units 336 may include, without limitation, graphical processing units (GPUs), dedicated signal processing modules, FPGA or other reconfigurable hardware that has been configured to instantiate parallel processing units for one or more specific tasks, or the like, A computing device, processor, apparatus, or module may be configured to instruct one or more dedicated hardware units 336 to perform one or more operations described herein, such as evaluation of model and/or algorithm outputs, one-time or iterative updates to parameters, coefficients, weights, and/or biases, and/or any other operations such as vector and/or matrix operations as described in this disclosure” (emphasis added). While the dedicated hardware units are specialized processors utilized to perform parallel computations, they are still conventional hardware units that the invention is merely using to perform the computationally intensive artificial intelligence tasks. That is, it was common to use such dedicated hardware units for tasks like artificial intelligence processing at the time of filing, and the instant claimed invention provides no innovations in their usage.
Furthermore, the Applicant’s claimed “user device” is described in paragraph [0055] as follows: “apparatus 100 may present to a user (such as a student) prompt 132 by sending a signal to user device 140, where the signal configures user device 140 to communicate prompt 132 to the user. In some embodiments, user device 140 may communicate prompt 132 to a user through user interface 144.User device 140and/or user interface 144 may have one or more properties described with reference to instructor device 124 and instructor interface 128. For example, user device 140 may include a smartphone, tablet, smartwatch, computer, and the like.” That is, the user device is reasonably interpreted as a generic computing device such as a smartphone, tablet, smartwatch, or computer, which also provides no details of anything beyond ubiquitous standard equipment.
As such, the claimed limitations of “an apparatus,” “a processor,” “a memory,” “a dedicated hardware unit,” and “a user device” are reasonably understood as not providing anything significantly more. Therefore, Step 2B, of the subject-matter eligibility analysis is “No.”
In addition, dependent claims 2-7, 10, 12-17, and 20 do not provide a practical application and are insufficient to amount to significantly more than the judicial exception. As such, dependent claims 2-7, 10, 12-17, and 20 are also rejected under 35 U.S.C. § 101, based on their respective dependencies to independent claims 1 and 11.
Therefore, claims 1-7, 10-17, and 20 are rejected under 35 U.S.C. § 101 as being directed to non-statutory subject matter.
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.
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.
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1-7, 10-17, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Birchfield et al. (hereinafter “Birchfield,” US 2021/0150925) in view of Rodrigo Cavalin et al. (hereinafter “Rodrigo,” US 2023/0092274), and in further view of Mascarenhas (US 6,808,393).
Regarding claim 1, and substantially similar limitations in claim 11, Birchfield discloses an apparatus for artificial intelligence-mediated multiparty electronic communication (Birchfield Abstract, “receiving, from a transceiver, signals include audio and/or video information associated with the participant and one or more recipients, determining a first question and/or a first statement provided by the participant based on analyzing the audio information and/or the video information that was received, identifying a second question and/or a second statement provided by the one or more recipients based on the analyzing the audio information and/or the video information,” receiving communication between participant and recipients; Birchfield [0050], “The system may use natural language processing, machine learning, artificial intelligence, and related techniques to identify question types.”), the apparatus comprising:
at least a processor (Birchfield [0010], “Various embodiments of the present inventive concepts are directed towards an electronic device that includes a processor”); and
a memory communicatively connected to the at least processor, the memory containing instructions (Birchfield [0010], “a memory coupled to the processor, the memory including computer readable program code”) configuring the at least processor to:
receive a discussion topic (see Birchfield Figs. 2A, 2B, and 3; also Birchfield [0042], “Still referring to FIGS. 2A, 2B, and/or 3, named entities and related feature extraction may be performed, at block 214. Keywords, specialized terms, important topics, and/or related terms may be identified and extracted from the teacher's and/or students' speech,” collecting audio/video of the participants, and identifying topics discussed);
…
training, iteratively, the prompt generation machine learning model using the prompt generation training data (Birchfield [0050], “training data may be used as the input to train the machine learning algorithm. After training, the machine learning algorithm will analyze individual questions and output a question-type label. The question-labeling algorithm may improve over time as the training dataset increases in time with each new session. The updated training dataset may be used to retrain and refine the machine learning algorithm at a regular interval.”); and
generating the first prompt as a function of the discussion topic using the trained prompt generation machine learning model (Birchfield [0063], “A question generator 555 may be capable of generating questions that are relevant to the topic and/or needs of the session”);
present to a plurality of users the first prompt by transmitting to a plurality of user devices a signal, wherein the signal configures the plurality of user devices to display the first prompt (see Birchfield Fig. 17 and [0071], “Analyzed data may be sent to devices and applications where it can be rendered for teachers and students to provide guidance for improved interaction”);
receive from a user device of the plurality of user devices a discussion datum, wherein the discussion datum comprises a response by a user of the plurality of users to the first prompt (Birchfield [0042], “Statements from the teacher 212 or students 213 may be identified, at block 231,234. Voice inflection and/or related natural language processing techniques may be used to distinguish questions vs. statements in transcribed speech.”);
train a score generation machine learning model, wherein training the score generation machine learning model (Birchfield [0077], “these metrics may be analogous to the confidence and higher-order thinking metrics in the classroom check for understanding score. Practically, these may be calculated based on the types of assessment data and/or student responses that are housed in the LMS,” where the understanding score is generated using the LMS, which is equivalent to the score generation machine learning model; also Birchfield [0050], “The system may use natural language processing, machine learning, artificial intelligence, and related techniques”) further comprises:
receiving a training dataset comprises model discussion data and model discussion topics correlated to model user understanding scores (Birchfield [0050], “Initial training data sets may be generated by human domain experts who are trained to label questions with the correct question-type. This training data may be used as the input to train the machine learning algorithm. After training, the machine learning algorithm will analyze individual questions and output a question-type label. The question-labeling algorithm may improve over time as the training dataset increases in time with each new session. The updated training dataset may be used to retrain and refine the machine learning algorithm at a regular interval. The question-types discussed above are included as specific non-limiting examples to illustrate an application in the field of education. The specific question-types may include but are not limited to these samples. In addition, the specific question-types may be modified to suit the appropriate techniques in related application areas including business, medicine, social-work, and human-resources.”); and
training the score generation machine learning model using the received training dataset (Birchfield [0050], “The updated training dataset may be used to retrain and refine the machine learning algorithm at a regular interval.”); and
generate a user understanding score as a function of the trained score generation machine learning model (Birchfield [0043], “A Check for Understanding Score may be computed, at block 235. The system may analyze student questions and statements to evaluate their understanding of the topic”), wherein generating the user understanding score further comprises:
inputting the first prompt generated as a function of the prompt machine learning model and the discussion datum into the trained score generation machine learning model (Birchfield claim 4, “determining a check for understanding score based on the second question and/or the second statement from the one or more recipients”);
outputting, as a function of the trained score generation machine learning model, the user understanding score (Birchfield claim 4, “generating the output comprising the guidance to the participant based on the check for understanding score.”);
…
generating a certainty score associated with the normalized user understanding score, wherein, when the certainty score fails to satisfy a predefined certainty score threshold, the apparatus automatically generates a follow-up prompt to acquire additional discussion data for evaluating the user (Birchfield [0045], “questions 410 may be analyzed to extract data about the student's understanding… The presence of specific question types (e.g., Content-Specific) indicates that the student understands the topic and is investigating further. The system increases the proficiency score and confidence values… Student statements are analyzed in the context of a dialog, at block 403… it may be determined if a teacher or another student asked a question that the student is now answering. If yes, at block 405, is there a direct response, at block 411. Now that the student has answered the question, does a teacher or other student respond directly to the student's statement? If no, at block 406, a decrease confidence Metric may be performed, at block 407a. If the system has ambiguous data, the confidence metric is lowered in response,” a confidence value is a certainty score; also Birchfield [0078], “It may be determined if the confidence is above a threshold, at block 1020. If the confidence is not above a threshold, at block 1021, the LMS for learning assessments may be queried, at block 1050a-b. If the confidence is low, the system may recommend gathering additional assessment data,” if the confidence value fails to satisfy a threshold value, the system gathers additional assessment data).
Birchfield does not explicitly teach every limitation of receive prompt generation training data, wherein the prompt generation training data correlates a plurality of discussion topics to a plurality of model prompts; generate a first prompt as a function of the discussion topic using a prompt generation machine learning model comprising: receiving the prompt generation training data.
However, Rodrigo discloses receive prompt generation training data, wherein the prompt generation training data correlates a plurality of discussion topics to a plurality of model prompts; generate a first prompt as a function of the discussion topic using a prompt generation machine learning model comprising: receiving the prompt generation training data (Rodrigo [0005], “The method can also include inputting the received topic and the extracted utterances to a trained machine learning model, the trained machine learning model generating example utterances for the new intent. The extracted utterances can include a question associated with the candidate intent, and the example utterances generated by the trained machine learning model can include a question associated with the topic”; also Rodrigo [0054], “example utterances generated by the trained machine learning model can include a question associated with the received topic”).
Rodrigo is analogous to Birchfield, as both are drawn to the art of machine learning. It would be obvious to try by one of ordinary skill in the art at the time of filing to have modified the method as taught by Birchfield, to include receive prompt generation training data, wherein the prompt generation training data correlates a plurality of discussion topics to a plurality of model prompts; generate a first prompt as a function of the discussion topic using a prompt generation machine learning model comprising: receiving the prompt generation training data, as taught by Rodrigo, since it combines prior art elements for training machine learning models according to known methods to yield predictable results. Doing so is a predictable solution that one of ordinary skill in the art could have pursued with a reasonable expectation of success.
Birchfield in view of Rodrigo does not teach every limitation of normalizing the user understanding score relative to a plurality of user understanding scores associated with other users participating in the discussion topic.
However, Mascarenhas discloses normalizing the user understanding score relative to a plurality of user understanding scores associated with other users participating in the discussion topic (Mascarenhas col. 6 lines 16-39, “Population means for number of moves and total time are calculated based on multiple responses collected for each assessment exercise, for example, from several hundred users. Each user's scores in each of these parameters (such as thinking time and search linearity) are then normalized and fitted to a population distribution. "High" scores are above the median value, while "Low" scores are below the median value,” assessment is made of the user’s understanding, then the score is normalized with a population distribution).
Mascarenhas is analogous to Birchfield in view of Rodrigo, as both are drawn to the art of education. It would be obvious to try by one of ordinary skill in the art before the effective filing date of the claimed invention to have modified the method as taught by Birchfield in view of Rodrigo, to include normalizing the user understanding score relative to a plurality of user understanding scores associated with other users participating in the discussion topic, as taught by Mascarenhas, in order to determine a user’s learning style (Mascarenhas col. 6 lines 16-39). Doing so is a predictable solution that one of ordinary skill in the art could have pursued with a reasonable expectation of success.
Regarding claim 2, and substantially similar limitations in claim 12, Birchfield in view of Rodrigo and Mascarenhas discloses wherein receiving the discussion topic comprises: receiving from an instructor device speech data; transcribing the speech data using an automatic speech recognition system to create transcribed speech data; and interpreting the transcribed speech data using a language model to determine the discussion topic (Birchfield [0040], “one or more persons 200 participate in a teaching and learning session. Real time audio from the session may be captured by a device or multiple devices in the room, to obtain an audio recording stream 201. Real time video from the session may be captured by a device or multiple devices in the room, to obtain a video recording stream 202. Automatic speech recognition may be used to transcribe each speaker's audio into text to provide speech to text transcription”; also Birchfield [0042], “named entities and related feature extraction may be performed, at block 214. Keywords, specialized terms, important topics, and/or related terms may be identified and extracted from the teacher's and/or students' speech. Summary statistics that include the presence and frequency of the terms that are used to establish the context of the lesson may be generated. For example, summary statistics may include what topics the lesson covers or if the topic is being introduced or reviewed.”).
Regarding claim 3, and substantially similar limitations in claim 13, Birchfield in view of Rodrigo and Mascarenhas discloses wherein the score generation machine learning model is trained on a training dataset including model discussion data and model prompts, associated with model user understanding scores; wherein the discussion datum and the first prompt are input into the score generation machine learning model (Birchfield [0050], “The system may use natural language processing, machine learning, artificial intelligence, and related techniques to identify question types. Initial training data sets may be generated by human domain experts who are trained to label questions with the correct question-type. This training data may be used as the input to train the machine learning algorithm. After training, the machine learning algorithm will analyze individual questions and output a question-type label. The question-labeling algorithm may improve over time as the training dataset increases in time with each new session. The updated training dataset may be used to retrain and refine the machine learning algorithm at a regular interval.”; also Birchfield [0077], “Conceptually, these metrics may be analogous to the confidence and higher-order thinking metrics in the classroom check for understanding score. Practically, these may be calculated based on the types of assessment data and/or student responses that are housed in the LMS,” past student responses influence calculation of the understanding scores).
Regarding claim 4, and substantially similar limitations in claim 14, Birchfield in view of Rodrigo and Mascarenhas discloses wherein presenting the first prompt to the user comprises sending a signal to a user device, wherein the signal includes the first prompt and configures the user device to display the first prompt, wherein receiving the discussion datum comprises receiving the discussion datum from the user device (Birchfield Fig. 2A and [0004], “determining a first question and/or a first statement provided by the participant based on analyzing the audio information and/or the video information that was received, identifying a second question and/or a second statement provided by the one or more recipients based on the analyzing the audio information and/or the video information”).
Regarding claim 5, and substantially similar limitations in claim 15, Birchfield in view of Rodrigo and Mascarenhas discloses wherein the memory contains instructions configuring the at least a processor to generate the user understanding score as a function of a grading threshold (Birchfield [0078], “the combined score and metrics 1005 may be determined by combining the classroom check for understanding score 235 and the enhanced LMS assessment data score 1004 to generate a composite score. It may be determined if the confidence is above a threshold, at block 1020. If the confidence is not above a threshold, at block 1021, the LMS for learning assessments may be queried, at block 1050a-b. If the confidence is low, the system may recommend gathering additional assessment data for the student by querying the LMS for related learning assessments. If the confidence is above a threshold, at block 1022, then it may be determined if the higher-order thinking score is above a threshold,… a check may be conducted as to whether the proficiency score is above the threshold,” the user understanding score is based on thresholds for confidence score, Higher-Order Thinking Score, and proficiency score, any or all of which may be considered a grading threshold).
Regarding claim 6, and substantially similar limitations in claim 16, Birchfield in view of Rodrigo and Mascarenhas does not explicitly teach communicating the user understanding score to an instructor. However, the Applicant’s decision to communicate the user understanding score to an instructor is an obvious design choice. Applicant has not disclosed that communicating the user understanding score to the instructor solves any stated problem or is for any particular purpose. Moreover, it appears that whether or not the user understanding score is communicated to an instructor in the apparatus and methods of Birchfield in view of Rodrigo or the Applicant would perform equally well. Therefore, it would have been prima facie obvious to modify Birchfield in view of Rodrigo to communicate the user understanding score to an instructor as specified in claim 6 and 16, because such a modification would have been considered a mere design consideration which fails to patentably distinguish over the prior art of Birchfield in view of Rodrigo.
Regarding claim 7, and substantially similar limitations in claim 17, Birchfield in view of Rodrigo and Mascarenhas discloses wherein generating the user understanding score comprises: recording audio of a user response to the first prompt to create an audio discussion datum; and transcribing the audio discussion datum using an automatic speech recognition system to create a transcribed audio discussion datum (Birchfield [0040], “one or more persons 200 participate in a teaching and learning session. Real time audio from the session may be captured by a device or multiple devices in the room, to obtain an audio recording stream 201. Real time video from the session may be captured by a device or multiple devices in the room, to obtain a video recording stream 202. Automatic speech recognition may be used to transcribe each speaker's audio into text to provide speech to text transcription”).
Birchfield in view of Rodrigo and Mascarenhas does not explicitly teach displaying the transcribed audio discussion datum to the instructor alongside the user understanding score. However, the Applicant’s decision to display the transcribed audio discussion datum to the instructor alongside the user understanding score is an obvious design choice. Applicant has not disclosed that communicating the user understanding score to the instructor solves any stated problem or is for any particular purpose. Moreover, it appears that whether or not the transcribed audio discussion datum is displayed to the instructor in the apparatus and methods of Birchfield in view of Rodrigo and Mascarenhas or the Applicant would perform equally well. Therefore, it would have been prima facie obvious to modify Birchfield in view of Rodrigo to communicate the user understanding score to an instructor as specified in claim 7 and 17, because such a modification would have been considered a mere design consideration which fails to patentably distinguish over the prior art of Birchfield in view of Rodrigo and Mascarenhas.
Regarding claim 10, and substantially similar limitations in claim 20, Birchfield in view of Rodrigo and Mascarenhas discloses wherein presenting to the user the first prompt is done using a chatbot (Birchfield [0106], “the Persuasion Intelligence Engine (PIE) may generate real time guidance for persuasive communication, at block 1350, and render two types of guidance, as will now be discussed. Recommended text for communication may be provided, at block 1351. The PIE makes specific recommendations for sentences, phrases, or words that can be inserted into a phone conversation, email, chat window or other communication channel. In some embodiments, such as a fully- or semi-automated chatbot, the PIE can automatically generate and broadcast statements directly to persuadees.”).
Response to Arguments
The Applicant’s arguments filed on February 10, 2026 have been fully considered and are addressed below.
Claim Rejections Under 35 USC 101: Step 2A, Prong one
The Applicant respectfully argues, “the claimed limitations cannot reasonably be construed as a mental process. While the Examiner characterizes the claims at a high level as ‘mediating multiparty communication,’ such a characterization improperly abstracts away express technical limitations recited in claim 1 and divorces the claim from its actual computational implementation. As explained in MPEP § 2111 and § 2106, the broadest reasonable interpretation must be consistent with the specification and the understanding of one of ordinary skill in the art, and does not permit reducing a claim to an oversimplified formulation that could theoretically be performed in the human mind only after stripping away its meaningful technical constraints.”
The Examiner respectfully disagrees. MPEP 2106.04(a)(2)(III)(C) states that a claim that requires a computer may still recite a mental process. The MPEP provides guidance that “examiners should review the specification to determine if the claimed invention is described as a concept that is performed in the human mind and applicant is merely claiming that concept performed 1) on a generic computer, or 2) in a computer environment, or 3) is merely using a computer as a tool to perform the concept. In these situations, the claim is considered to recite a mental process.” In the present case, the claimed invention is described as the abstract concept of “mediating multiparty communication” and applicant is merely claiming that concept performed on a generic computer or computer environment, and is merely using a computer as a tool to perform the concept.
The claims were considered as a whole without simplification or disregarding any limitations, as can be seen in the body of the rejection above, where each limitation is listed and each additional element considered separately. The claimed hardware elements of “an apparatus,” “a processor,” “a memory,” “a dedicated hardware unit,” and “a user device” are all recited as generic computer components. Further, these generic computer components only perform conventional operations such as receiving data, generating data, transmitting data, training a machine learning model, presenting data, receiving input, displaying output, and normalizing data. Each of these operations is claimed in a general sense, without improvements or unconventional limitations. Therefore, the instant claims still fall within the abstract subgrouping of “mental processes.”
The Applicant further respectfully argues that the “normalization step expressly requires computing percentile-based or scaled scores derived from outputs of the score-generation machine-learning model for a plurality of users, which necessitates population-level aggregation and statistical processing that cannot be practically performed by human cognition or with pencil and paper.”
The Examiner respectfully disagrees. The claimed normalization step only requires “normalizing the user understanding score relative to a plurality of user understanding scores associated with other users participating in the discussion topic.” Simply normalizing a score relative to another score, as the claim requires, is a basic mathematical calculation that can be performed by a human without aid of a computing device.
The Applicant also respectfully argues that “claim 1 further requires generating a certainty score associated with the normalized user understanding score and automatically generating a follow-up prompt when the certainty score fails to satisfy a predefined certainty score threshold. This limitation introduces conditional, machine-driven control logic that governs subsequent system behavior based on quantitative confidence determinations produced by a machine-learning model. The human mind does not generate certainty scores in the form required by machine-learning systems, does not compare such scores to predefined thresholds, and cannot automatically trigger subsequent data-acquisition operations in response to such threshold evaluations.”
The Examiner respectfully disagrees. Conditional behavior based on comparing scores to threshold scores is based on basic mathematical calculations that can be performed by a human without aid of a computing device. Furthermore, the limitations that require a computer (controlling computer behavior, formatting data in a way computers can read it, or running a machine-learning model) only require conventional processes performed by a generic computer (see MPEP 2106.04(a)(2)(III)(C), a claim that requires a computer may still recite a mental process).
The Applicant respectfully argues, “the limitations of claim 1 are analogous to those of Example 39, which the USPTO has identified as a representative example of claims that do not recite a mental process. As explained in Example 39, claims that recite multi-stage training and processing of machine-learning models using successive datasets derived from prior model outputs are not practically performed in the human mind and therefore fall outside the mental-process exception.”
The Examiner respectfully disagrees. Example 39 recites more than simple multi-stage training and processing of machine learning models. Example 39 was determined eligible because limitations such as “collecting a set of digital facial images,” applying “mirroring, rotating, smoothing, or contrast reduction” to the digital facial images, creating a first training set including non-facial images, and creating a second training set including non-facial images that were incorrectly detected as facial images could not be practically performed in the human mind. It is the storage of digital facial images, transformations performed on them, and training of machine learning models using unconventional combinations of different types of digital images that renders Example 39 eligible. Simply performing conventional training and processing of machine learning models, as is recited in the instant claims, is still a mental process (see MPEP 2106.04(a)(2)(III)(C), a claim that requires a computer may still recite a mental process).
Claim Rejections Under 35 USC 101: Step 2A, Prong Two
The Applicant respectfully argues, “Claim 1 provides a technical solution to this problem by requiring a specific computational architecture in which user understanding scores generated by a trained score-generation machine-learning model are normalized relative to a plurality of other users participating in the same discussion topic, such that the resulting score is determined as a percentile-based or scaled value derived from outputs of the model for multiple users. This population-level normalization improves the robustness and reliability of automated evaluation by contextualizing individual responses within a multi-user dataset, a task that cannot practically be performed in the human mind.”
The Examiner respectfully disagrees. As discussed previously, the claims only recite conventional operations performed by a generic computer. The apparatus consists of a processor and a memory, without any unconventional or specific elements, and therefore is a generic computer. Furthermore, the operations performed by the generic computer are standard operations for using a generic machine learning model, including receiving training data, iteratively training the model with data, generating prompts and scores, and normalizing output. None of these features improve upon the technology of machine learning, or the computer itself. Therefore, the claims are still directed to a judicial exception without significantly more.
The Applicant further respectfully argues, “Claim 1 further integrates the alleged exception into a practical application by requiring the generation of a certainty score associated with the user understanding score and the automatic generation of a follow-up prompt when the certainty score fails to satisfy a predefined certainty score threshold. This limitation introduces machine-driven control logic that dynamically governs subsequent data acquisition based on quantitative confidence determinations produced by the machine-learning model itself. Rather than merely evaluating information, the claim requires the system to adapt its operation in response to model uncertainty, thereby improving the technical functioning of the AI-mediated communication system.”
The Examiner respectfully disagrees. Generating a certainty score and comparing it with a threshold to make a decision is not an improvement on technology, because the claims fail to show the specifics of how the certainty score is generated and why it is an improvement on technology.
The Applicant respectfully argues that “the proper inquiry is whether the claim as a whole meaningfully limits the alleged exception and applies it in a specific technological manner. Even if certain underlying components, such as processors or machine-learning models, are known in isolation, claim 1 integrates them in a particular arrangement, population-level normalization combined with certainty-threshold-based automatic re-prompting, that imposes concrete operational constraints on the system and improves its reliability.”
The Examiner respectfully disagrees. The Applicant argues that the elements of the claims are laid out in “a particular arrangement,” and therefore constitutes a practical application. MPEP 2106.04(d)(I) identifies a number of considerations relevant to the evaluation of whether the claimed additional elements amount to an inventive concept. Of particular relevance is the final paragraph of MPEP 2106.04(d)(I), which states, “a specific way of achieving a result is not a stand-alone consideration in Step 2A Prong Two. However, the specificity of the claim limitations is relevant to the evaluation of several considerations including the use of a particular machine, particular transformation and whether the limitations are mere instructions to apply an exception. See MPEP §§ 2106.05(b), 2106.05(c), and 2106.05(f). For example, in Parker v. Flook, 437 U.S. 584, 198 USPQ 193 (1978), the Supreme Court noted that the "patent application does not purport to explain how to select the appropriate margin of safety, the weighting factor, or any of the other variables" in the claimed mathematical formula, "[n]or does it purport to contain any disclosure relating to the chemical processes at work, the monitoring of process variables, or the means of setting off an alarm or adjusting an alarm system." 437 U.S. at 586, 198 USPQ at 195. The Court found this failure to explain any specifics of how to use the claimed formula informative when deciding that the additional elements in the claim were insignificant post-solution activity and thus not meaningful enough to render the claim eligible. 437 U.S. at 589-90, 198 USPQ at 197.” The Examiner notes that “particular machine” and “particular transformation” are considerations, but “particular arrangement” is not. Further, the examples of relevant court cases fail to explain any specifics of how to use the claimed elements and were determined not meaningful enough to render the claims eligible. These relevant court cases are similar in scope to the instant claims, which recite determining a “certainty score” and “normalizing the user understanding score,” but fails to explain any specifics of how these are accomplished. Therefore, the claims do not integrate the judicial exception into a practical application.
The Applicant also respectfully argues, “Similar to BASCOM, amended claim 1 recites additional elements that, when considered individually and as an ordered combination, amount to significantly more than any alleged judicial exception because they define a non-generic, technical machine-learning implementation for reliably evaluating multiparty electronic communications, rather than merely stating a desired result and invoking generic ‘AI.’”
The Examiner respectfully disagrees. The facts of the instant case are very different from the facts of BASCOM. In BASCOM (according to MPEP 2106.05(I)(B)), the claims were found eligible because “the installation of a filtering tool at a specific location, remote from the end-users, with customizable filtering features specific to each end user” was a non-conventional and non-generic arrangement of the additional elements. The instant claims, in contrast, have the additional elements of “an apparatus,” “a processor,” “a memory,” “a dedicated hardware unit,” and “a user device,” which are arranged conventionally. Further, there is nothing unconventional about receiving training data, generating prompts, iteratively training the model, outputting data, and normalizing data, as recited in the claims. Few details are provided as to how these operations are performed, beyond simply stating what data is input or output from the operations. Where BASCOM claimed inventive distribution of functionality within a network to filter Internet content, the instant claims appear to recite only generic machine learning operations such as training, input, output, and normalization of data in relation to a generic machine learning model.
As such, the arguments are not persuasive, and the outstanding 35 USC 101 rejections are maintained.
Claim Rejections Under 35 U.S.C. & 103
Applicant’s arguments with respect to the claims have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
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 Stephen Alvesteffer whose telephone number is (571)272-8680. The examiner can normally be reached M-F 8:00-6:00.
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, Peter Vasat can be reached at 571-270-7625. 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.
/SA/Examiner, Art Unit 3715
/PETER S VASAT/Supervisory Patent Examiner, Art Unit 3715
Prosecution Insights