DETAILED ACTION
This communication is in response to the Amendments/Arguments filed on 05/18/2026. Claims 1 and 15-33 are pending and have been examined. Hence, this action has been made FINAL.
Any previous objection/rejection not mentioned in this Office Action has been withdrawn by the Examiner.
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 .
Examiner Note
The Examiner of record has changed from Satwant Singh to SPE Paras Shah.
Compact Prosecution
The Examiner recommends reviewing the allowed claims and incorporating subject matter from the allowed set 15-20 and claims 23, 26 which have been already noted.
Response to Arguments and Amendments
With respect to the 35 USC 101 rejections,
the Applicant asserts on page 11 under Step 2A, Prong One:
At Step 2A, Prong One, the claim is not "directed to" an abstract idea, because it does not recite any mental process, mathematical concept, or method of organizing human activity. Rather, it recites a concrete, computer-implemented process that operates on multimedia data streams captured by meeting input/output devices to generate and iteratively update a strategic message model. The claimed operations-"obtaining input data stream comprising multimedia data generated by meeting data input/output devices," "formulating a strategic message model by machine learning," "monitoring a presentation by detecting delivery issues," and "updating the strategic message model based on a reaction by the audience"-each require specific computer and sensor functionality that cannot be performed mentally. Humans cannot, by observation alone, process synchronized multimedia input data streams from microphones, cameras, and connected presentation systems; cannot execute machine- learning algorithms to formulate predictive models; and cannot automatically update such models based on digitally captured delivery and reaction data. These are inherently technological operations requiring computer processing of data in forms imperceptible to the human mind.
The Examiner's assertion that a human "can determine by observing the discussion/event what the topic/message is and what information needs to bc presented" improperly oversimplifies the claim by ignoring its explicit requirement that the strategic message model be formulated by machine learning. A human mind cannot execute a supervised or unsupervised learning algorithm, perform vectorization of multimedia content, or update learned parameters from audience-reaction data in real time. As the Federal Circuit has emphasized in McRO V. Bandai Namco Games, 837 F.3d 1299 (Fed. Cir. 2016), a claim directed to "automatic" generation or analysis of expressive content using specific computational rules is patent-eligible because it recites a technological improvement over human mental processes. The present claim is analogous: it automates and improves the process of generating and refining strategic messages by machine learning using data unavailable to, and unprocessable by, human cognition.
The Examiner respectfully disagrees with these assertions. First, the Examiner notes that the claim limitation makes no mention of “synchronized multimedia input data streams”. Thus, rendering the Applicant’s assertions moot. Further, even if such limitations were added, these limitations would relate to pre-solution activity of gathering the multimedia data stream for further processing. The Applicant further asserts that humans “Cannot execute machine learning algorithms”. The Examiner agrees. However, “machine learning” models have been interpreted as additional elements. The Examiner also refers the Applicant to Example 47, claim 2 of the SME examples provided by the USPTO which is in stark contrast to the Applicant’s assertions which notes “neural networks” to be considered as additional elements. The Applicant further notes that humans cannot update models. This is not true. Developers have the ability to analyze output data generated by general models in order to provide further adjustments to the models. The Applicant has not provided any limitation in the claims that notes how the models are updated and what these models are apart from generally noting that these models are updated based on delivery content capturing the presentation and audience reaction. Thus, contrary to the Applicant’s assertion they are not “inherently” technological operations that require computer processing. This is not the standard that is applied under prong one of Step 2A as it appears the Applicant is coflating Prong 2 of Step 2A. The analysis under Step 2A, prong one involved determining whether the claim limitations are “directed to” an abstract idea minus the additional elements.
With respect to the 2nd paragraph of the Applicant’s assertions above, the Applicant appears to note that the claim has been oversimplified by ignoring that that “strategic message model” is formulated from “machine learning”. This analysis is incorrect. The Examiner has not oversimplified the claim language since the “strategic message model” that has been formulated using “machine learning” has been interpreted as an additional element and that is the reason it has not been addressed in Step 2A, prong One. The Applicant further notes “McRO V. Bandai Namco Games” however the Applicant appears to be confusing Prong Two of Step 2A from Prong One.
The Applicant asserts under Step 2A, Prong Two:
At Step 2A, Prong Two, even if one were to characterize some aspect of the claim as involving an abstract idea such as "analyzing and improving communication," the claim integrates that idea into a practical application. The recited system does not merely analyze information; it uses machine learning to formulate and update a strategic message model in real time based on computer-captured delivery content and audience reactions. This feedback loop improves the functioning of the computing system itself by enabling the machine-learning model to self-adapt using incoming performance data. The claim therefore applies any underlying idea in a particular technological context-multimedia data capture and adaptive machine learning-to achieve a tangible technological result: improved delivery of strategic messages through dynamic model retraining. Such integration into a practical technological application satisfies Step 2A, Prong Two of the USPTO's 2019 Guidance.
The Examiner respectfully disagrees with this assertion. As noted in the rejection below the “machine learning model” is only being used as a “tool” to provide a strategic message instance using “strategic message model”. Further, there is nothing provided in the claim that sets this “model” apart from any other model such as how this “model” is trained and what it comprises. The Examiner refers the Applicant to SME example 47 which recites “neural network” claim 2. Similar to claim 2 of example 47, where the “ANN” limitation “The limitations in (d) and (e) reciting “using the trained ANN” provide nothing more than mere instructions to implement an abstract idea on a generic computer. See MPEP 2106.05(f). MPEP 2106.05(f) provides the following considerations for determining whether a claim simply recites a judicial exception with the words “apply it” (or an equivalent), such as mere instructions to implement an abstract idea on a computer: (1) whether the claim recites only the idea of a solution or outcome i.e., the claim fails to recite details of how a solution to a problem is accomplished; (2) whether the claim invokes computers or other machinery merely as a tool to 8 perform an existing process; and (3) the particularity or generality of the application of the judicial exception. The judicial exception of “detecting one or more anomalies in a data set using the trained ANN” and “analyzing the one or more detected anomalies using the trained ANN to generate anomaly data” is performed “using the trained ANN.” The trained ANN is used to generally apply the abstract idea without placing any limits on how the trained ANN functions. Rather, these limitations only recite the outcome of “detecting one or more anomalies” and “analyzing the one or more detected anomalies” and do not include any details about how the “detecting” and “analyzing” are accomplished. See MPEP 2106.05(f). The recitation of “using a trained ANN” in limitations (d) and (e) also merely indicates a field of use or technological environment in which the judicial exception is performed. Although the additional element “using a trained ANN” limits the identified judicial exceptions “detecting one or more anomalies in a data set using the trained ANN” and “analyzing the one or more detected anomalies using the trained ANN to generate anomaly data,” this type of limitation merely confines the use of the abstract idea to a particular technological environment (neural networks) and thus fails to add an inventive concept to the claims. See MPEP 2106.05(h).” The present claims similarly incorporate similar attributes by reciting “the strategic message model” using “machine learning” as merely being generally used to apply the abstract idea without placing any limits on how the trained “strategic message model” functions and only recites the outcomes without ANY details on how the “generating” and “updating” occur. Also, the use of “strategic message model” merely “confines the use of the abstract idea to a particular technological environment (neural networks) and thus fails to add an inventive concept to the claims.”
The Applicant asserts under Step 2B:
At Step 2B, the claim also recites "significantly more" than any abstract idea. The ordered combination of elements-acquiring live multimedia data from meeting I/O devices, applying machine- learning algorithms to formulate and update a message model, detecting delivery issues through digital monitoring of presentations, and retraining the model based on real-time audience reactions-represents a specific implementation of a computing architecture that improves both the technological field of automated content generation and the computer's own ability to adaptively process streaming, multimodal data. This improvement is comparable to those held eligible in DDR Holdings V. Hotels.com, 773 F.3d 1245 (Fed. Cir. 2014) (improving website behavior in a networked environment), and McRO (automatic rule-based content generation). It is not a generic application of routine computer functions but rather a coordinated, machine-implemented workflow that could not be performed mentally or manually.
The Examiner's suggestion that "a human brain can also be a learning model that can be trained" conflates metaphor with statutory analysis. The comparison of a neural network to a human brain does not convert a computer-implemented machine-learning model into a mental process any more than referring to "electronic memory" renders a computer indistinguishable from human recollection. The claim explicitly limits the learning to machine execution and specifies that the input comprises multimedia data captured by devices, not human observation. Thus, it is rooted in computer technology, not human thought.
Furthermore, the claimed process yields a concrete, machine-readable result-the updated strategic message model-which is stored, retrained, and used to guide subsequent message generation. This is not a result of human judgment but of computational processing of sensor-based inputs. The technological improvement is therefore both practical and measurable, not abstract.
For at least these reasons, claim 1 is directed to a patent-eligible technological process that uses machine learning and multimedia data capture to improve the generation and adaptive delivery of strategic messages. The claim does not recite or preempt any mental process, law of nature, or mathematical formula, and it integrates any abstract concept into a specific and practical technological application. Accordingly, the rejection under 35 U.S.C. § 101 is respectfully traversed and withdrawal is requested.
The Examiner respectfully disagrees with these assertions. As noted above, with respect to Step 2A, prong Two, the usage of the additional element of “strategic message model” relate to mere instructions to apply the abstract idea and such cannot provide an inventive concept (see MPEP 2106.05(f)). Further, this “model” is recited at a high level of generality as noted and mentioned in Applicant’s Specification para [0047] as relating to generic machine learning model. The additional elements of “g live multimedia data from meeting I/O devices” relate to insignificant extra solution activity which are related to data gathering and output. Therefore, the Examiner notes that the above comments regarding the present claims presenting “a specific implementation of a computing architecture that improves both the technological field of automated content generation and the computer's own ability to adaptively process streaming, multimodal data”. Although, the claims provide an implementation of “Computer architecture”, this has not been deemed as an improvement due to the reasons provided above as these limitations are being used as “mere instructions to apply the abstract idea and at a “high level of generality”. The remaining assertions made by the Applicant are not persuasive for the reasons mentioned in this paragraph.
Applicant on bottom of pages 13-page 16 reiterates the same points already addressed above. Therefore, these arguments are not being re-addressed and the Applicant is directed to each point noted and addressed above with respect to each step of the analysis for Subject Matter Eligibility. Therefore, the Applicant’s arguments are not persuasive.
With respect to the 35 USC 103 rejections,
The Applicant asserts on page 19 with respect to claim 1:
The cited portions of Seleskerov do not disclose this required workflow. The Office Action relies on Scleskerov paragraph [0064] for the claimed "input data stream capturing a live discussion by one or more strategists on topics of strategic messages to deliver to an audience." However, the cited passage merely describes presentation media streams transmitted by a presentation and communications platform to client devices, including audio media streams, video media streams, and other media streams. That disclosure concerns media streams of an online presentation or communication session. It does not disclose an input data stream capturing a live discussion by one or more strategists on topics of strategic messages to deliver to an audience. The rejection therefore treats generic presentation media streams as if they were the claimed strategist-discussion input stream, without identifying where Seleskerov discloses the claimed strategists, the claimed live discussion by those strategists, or the claimed topics of strategic messages to be delivered to an audience.
The Examiner respectfully disagrees. The claims require the capturing of discussion of “one or more strategists” regarding “topics of strategic messages”. Per the specification, it is unclear what a “strategist” is and what “topics of strategic messages” is. The claim has been mapped properly under BRI. Seleskerov in para [0064]. This paragraph describes the transmission of media which comprises audio and/or video which is related to discussion of presentation content with participants. Paragraph [0031], further notes feedback being presented live during this presentation. Therefore, Seleskerov clearly shows capturing of live discussion of presentation from a presenter discussing with participants content. In general, presentations inherently comprise a series of topics that are being presented. Neither the claims nor the specification provide any description of how this “strategic message” or “strategists” different from a “presenter” and their “content” or “presentation material” which is being presented among participants differs from what is currently claimed. Para [26] of the published spec notes “strategic message” to be “ the term “strategic message” indicates a message to audiences as an ordered collection of topics on a strategy that is developed and generated by the strategy management system 110 and delivered to the audiences via the delivery agent system 190”. In other words, the message is just a collection of topics on a strategy that has been generated. A presentation generally comprises a collection of topics on a main topic that is to be presented. The Applicant appears to be taking a narrow interpretation for these elements and such interpretation is not persuasive as the specification/claims make no further distinction. Therefore, Seleskerov properly maps per the BRI standard.
The Applicant asserts on page 20:
The Office Action next relies on Seleskerov paragraph [0038] for "formulating a strategic message model by machine learning." But the cited passage describes a presentation coaching unit using a delivery attributes model to analyze audio, video, and presentation content with machine learning models trained to identify aspects of a presenter's presentation skills and presentation content that are good or may benefit from improvement. That disclosure is materially different from formulating a strategic message model by machine learning from a captured strategist discussion. A delivery-attributes model for critiquing presentation skills is not the claimed strategic message model. The cited model concerns presenter performance attributes, such as presentation skills and content-improvement feedback. It is not shown to be formulated from a live discussion by strategists, and it is not shown to model strategic- message topics for later generation of a strategic message instance.
The Examiner respectfully disagrees. The Applicant appears to be making a general allegation of patentability without providing specific details as to how the claims limitation differs from that of Seleskerov. Instead the Applicant appears to summarize what Seleskerov is directed towards and then summarizes the claim limitations as a whole. How is the claim limitation being argued different from that of the prior art? The claim requires formulation of “strategic message model” by machine learning. That is all the claim requires. Under BRI, Seleskerov in Paragraph [0038] describes the presentation coaching unit used to analyze presenter information (e.g. live discussion by strategist) to provide improvements to “presentation content” (e.g. strategic message instance). This clearly reads on the current claim language.
The Applicant asserts on page 20:
The rejection also relies on Seleskerov paragraph [0064] for "generating a strategic message instance by use of the strategic message model," stating that Seleskerov discloses presentation content such as slides, documents, or other content discussed during a presentation. But the mere existence of slides, documents, or other presentation content is not generation of a strategic message instance by use of the strategic message model. The cited passage does not identify any strategic message model used to generate the content. Nor does it disclose that the presentation content is a "strategic message instance" generated from a model formulated from strategist-discussion multimedia data. The rejection therefore maps ordinary presentation content to the claimed model-generated strategic message instance, but the mapping omits the claimed generative relationship between the strategic message model and the strategic message instance.
The Examiner respectfully disagrees. Again the Applicant appears to be affording a special definition to “strategic message instance”. The claims/specification make not mention of how the “strategic message instance” differs from “presentation content” that has been modified based on the current presentation of the presenter. Para [0075], further notes how feedback to the presentation is provided in the summary reports mentioned in para [0064]. This under BRI reads on the “strategic message instance” since it provides a summary report of how the presenter is presenting this a “message instance”. The claims has been mapped properly. The Applicant also notes “the cited passage does not identify any strategic message model”. The claims do not require any identification of a “strategic message model”. In fact, the earlier mapping already identified the “presentation coaching unit 235” as the “strategic message model”.
Applicant asserts on page 20:
Similarly, although Seleskerov on current review may disclose monitoring a presentation and identifying presentation-style issues, that disclosure does not cure the deficiencies in the earlier limitations. The Office Action cites Seleskerov paragraph [0049] for identifying issues with presenter style, including language usage, monotone delivery, reading slide content, emotional state, eye contact, gaze direction, and body posc. Even assuming this disclosure corresponds generally to detecting delivery issues, anticipation requires the complete claimed combination, not isolated similarity to one part of the claim. Seleskerov's detection of presentation-style issues occurs within a presentation-coaching framework, not within the claimed strategist-discussion-to-strategic-message-model-to-strategic-mesage- instance workflow.
The Examiner respectfully disagrees. Seleskerov teaches ALL of the earlier limitations as noted and such reasoning has been provided above. As noted above, the Applicant appears to note that such teachings of Seleskerov are not within “strategist-discussion-to-strategic-message-model-to-strategic-mesage- instance workflow”. As noted above and due to the lack of explanation on these terms, the Applicant’s assertions appear to be related to general allegations of patentability without providing clear reasoning/evidence of how such terms differ and how such terms per the specification differ from the BRI of the claims. Hence, this limitation has been properly mapped to [0049] of Seleskerov which continuously provides feedback to a speaker’s skills in order to improve the presentation as the claims further indicate the “detection of delivery issues: that are monitored. It is unclear why the Applicant believes this limitation is not taught.
Applicant on page 21 asserts:
Finally, the cited disclosure of updating in Seleskerov does not, on current review, disclose "updating the strategic message model based on a delivery content capturing the presentation and a reaction by the audience," as claimed. The Office Action relies on Seleskerov paragraph [0052], which describes updating a slide attribute model and/or delivery attributes model based on participant reaction information to improve recommendations provided by those models. That is not an update of the claimed strategic message model. Seleskerov's models are directed to slide attributes and delivery attributes for presentation feedback. The rejection does not identify where Seleskerov updates a strategic message model that was previously formulated from a live strategist discussion and used to generate a strategic message instance. Improving presentation recommendations based on participant reaction is not the same as updating the claimed strategic message model in the claimed workflow.
The Examiner disagrees with these assertions. The claim requires “updating the strategic message model based on a delivery content capturing the presentation and a reaction by the audience”. Seleskerov does teach this in para [0052]-[0056]. These sections notes that the model updates based on BOTH the participant reaction information and the online presentation content presented by the presenter. This is clearly taught by Seleskerov. Again, the Applicant’s arguments relate to general allegation of patentability without showing how Seleskerov differs from the claim. Even if this clarification was made. The claims simply requires an update of the “Strategic message model” based a “delivery content capturing the presentation” and “reaction by the audience” both of which Seleskerov teaches. The fact that slide attributes are updated based on participant reaction does not make the claims different from what is claimed as these paragraphs further note that these models are updated.
The Applicant’s arguments with respect to claim 27 are moot in view of new grounds for rejections.
Applicant asserts on page 23 with respect to claim 28:
Claim 28 recites: "The computer implemented method of claim 1, further comprising: processing the input data stream obtained from the meeting data devices to distinguish among different types of content contributed during the live discussion, and preparing structured representations of such content suitable for use by the machine learning that formulates the strategic message model."
The Office Action rejects claim 28 as anticipated by Seleskerov. The Office Action relies on Seleskerov's disclosure that a presentation and communications platform implements an architecture for analyzing audio, video, and/or multimodal media streams and/or presentation content, and that media streams and/or presentation content may be analyzed to extract feature information for processing by various models.
Applicant respectfully traverses the rejection of claim 28. Claim 28 does not merely recite analyzing media streams or extracting features. Claim 28 requires processing "the input data stream obtained from the meeting data devices" to distinguish among different types of content "contributed during the live discussion," and preparing structured representations of "such content" for use by the machine learning that formulates the strategic message model. Thus, claim 28 remains tied to the upstream live strategist discussion and to the formulation of the claimed strategic message model.
The cited portions of Seleskerov do not disclose these limitations. Seleskerov's generalized analysis of audio, video, multimodal media streams, and presentation content occurs in the context of online presentation analysis and presentation coaching. The cited disclosure does not identify a live discussion by strategists, does not identify content contributed during that strategist discussion, and does not disclose structured representations of that strategist-discussion content being used by machine learning to formulate a strategic message model.
At most, the cited portions of Seleskerov disclose feature extraction for models used in a presentation coaching or presentation hosting platform. That is different from claim 28's requirement that content contributed during a live strategist discussion be distinguished and structurally represented for use by the machinc learning that formulates the strategic message model. Because the rejection docs not identify where Seleskerov discloses the claimed strategist-discussion content processing and strategic- message-model formulation relationship, Seleskerov does not anticipate claim 28. Withdrawal of the rejection is respectfully requested.
The Examiner respectfully disagrees with this assertion. The Applicant asserts that Seleskerov does not specifically teach “identify a live discussion by strategists, does not identify content contributed during that strategist discussion”. The claims are not worded in this manner”. Further, “for use by the machine learning model” is intended use of the structured representation. Seleskerov teaches ALL of the limitations as required by claim 28. The claim 28 only requires processing of the “input data stream obtained from the meeting data devices” as disclosed. Seleskerov teaches in [0035] that audio, video and multimodal streams are analyzed. The extraction of features by processing the various streams would distinguish the different types of content. The claim limitation notes “to distinguish among different types…” is an intended result of the “processing”. Further, the claim limitation of “preparing structure representations” is broad. The extraction of “features” from the processing of the various streams reasonably relates to “structured representations” as claimed. The Applicant has not commented on how such is different. Seleskerov then notes that these features extracted are used by the “models”. Thus, showing how the features extracted (e.g. structured representation) is used by the models.
Applicant asserts on page 24, with respect to claim 29:
Claim 29 recites: "The computer implemented method of claim 1, wherein the method includes processing the multimedia data, wherein the processing the multimedia data includes identifying speakers and linguistic content within audio data, converting the identified speech into text, and providing the resulting text for use by the machine learning that formulates the strategic message model."
The Office Action rejects claim 29 as anticipated by Seleskerov. The Office Action relies on Seleskerov's speaker-skills feedback unit identifying issues with a presenter's presentation style, Seleskerov's generation of a transcript of the audio portion of an online presentation or communication session, and Seleskerov's models receiving feature data extracted from presentation media streams, participant media streams, and/or reaction data.
Applicant respectfully traverses the rejection of claim 29. Claim 29 requires more than identifying presentation-style issues or generating a transcript. Claim 29 requires processing the claimed multimedia data by identifying speakers and linguistic content within audio data, converting the identified speech into text, and providing the resulting text for use by the machine learning that formulates the strategic message model.
The cited disclosure of Seleskerov does not disclose providing resulting text from the claimed live-discussion multimedia data for use by machine learning that formulates a strategic message model. Seleskerov's transcript generation concerns an online presentation or communication session. The cited disclosure of feature data being provided to models relates to models used for analyzing presentation media, participant media, or reaction data. The rejection does not identify a strategic message model, much less a strategic message model formulated using text derived from a captured live strategist discussion.
Further, identifying issues with a presenter's language usage or presentation style is not the same as identifying speakers and linguistic content within audio data from a strategist discussion and converting that speech into text for formulation of a strategic message model. The cited speaker-skills feedback unit is directed to evaluating presenter performance. Claim 29 is directed to preparing linguistic information from multimedia data for model formulation. These are different operations performed for different purposes in different workflows.
Accordingly, Seleskerov does not disclose all limitations of claim 29 arranged as claimed, and withdrawal of the rejection of claim 29 is respectfully requested.
The Examiner respectfully disagrees. The Examiner notes that the claim requires “wherein the processing the multimedia data includes identifying speakers and linguistic content within audio data, converting the identified speech into text, and providing the resulting text for use by the machine learning that formulates the strategic message model”. More specifically, the claims require from the multimedia data identifying speakers and linguistic content and converting identified speech into text. Then, providing the text to a “model”. The main argument relates to the last limitation. However, the citation from para [0084] provides as an output the transcription of the audio stream. The language of “for use by the machine learning that formulates the strategic message model” recites intended use language. The examiner understands that claim 29 is directed towards model formulation but this is not positively claimed which is why the mapping provided teaches the claimed limitation of converting into text, identifying aspects of the presenter and updating of a model as noted in the citations in the Office Action. The Examiner refers the Applicant to the reasoning provided above with respect to “strategic message model” and “strategist discussion”.
Applicant asserts on page 25, with respect to claim 30:
Claim 30 recites: "The computer implemented method of claim 1, further comprising: processing the multimedia data to detect spoken content and corresponding textual input, and generating text based data derived from audio content for use in training the strategic message model."
The Office Action rejects claim 30 as anticipated by Seleskerov. The Office Action relies on Seleskerov's disclosure that presentation media streams may include an audio component of a presentation where a presenter discusses presentation content, that presentation content may include slides, documents, or other content, that a transcript of the audio portion of an online presentation or communication session may be generated, and that models may receive feature data extracted from presentation media streams, participant media streams, and/or reaction data.
Applicant respectfully traverses the rejection of claim 30. Claim 30 requires processing the multimedia data to detect spoken content and corresponding textual input, and generating text-based data derived from audio content "for use in training the strategic message model." The cited portions of Seleskerov do not disclose this training relationship. Seleskerov's disclosure of an audio component of a presentation and generation of a transcript may show that audio can be converted to text in a presentation environment. However, claim 30 is not satisfied by transcript generation alone. Claim 30 specifically requires text-based data derived from audio content for use in training the strategic message model.
The Office Action does not identify any disclosure in Seleskerov in which text-based data derived from audio content is used to train a strategic message model. Nor does the cited presentation-coaching disclosure identify the claimed strategic message model formulated from a live strategist discussion.
The Office Action's reliance on generic model feature input docs not remedy this deficiency. Feature data extracted from presentation media streams, participant media streams, or reaction data for use by presentation-analysis models is not the same as text-based data derived from audio content for training a strategic message model. Seleskerov's models are directed to presentation coaching, slide attributes, delivery attributes, and reaction analysis, not strategic-message-model training as claimed.
Therefore, Seleskerov does not anticipate claim 30, and withdrawal of the rejection is respectfully requested.
The Examiner respectfully disagrees for the same reasons as mentioned in claim 29. The claim notes “for use in training the strategic message model”. This is intended use of the text based data. Therefore, the claim does not per se require this which is why the claims were mapped in the manner there was.
Applicant asserts on page 26, with respect to claim 31:
The Office Action rejects claim 31 as anticipated by Seleskerov. The Office Action relies on Seleskerov's disclosure that a presentation coaching unit may provide feedback critiques on aspects of presentation skills, including pacing, vocal pattern, volume, monotone speech, and language usage.
Applicant respectfully traverses the rejection of claim 31. Claim 31 requires that multimedia data include audio and text data, that the audio and text data be analyzed to extract linguistic, tonal, and contextual information, and that the extracted information be used by the strategic message model in generating the strategic message instance. This is a specific use of extracted multimodal information in the generation of the strategic message instance.
The cited disclosure of Seleskerov is materially different. Seleskerov's presentation coaching unit may critique presentation skills such as pacing, vocal pattern, volume, monotone speech, and language usage. Those critiques relate to evaluating presenter delivery. They do not disclose extracting linguistic, tonal, and contextual information from audio and text data for use by a strategic message model in generating a strategic message instance.
The rejection does not identify where Seleskerov discloses a strategic message model, where that model uses extracted linguistic, tonal, and contextual information, or where such information is used in generating a strategic message instance. Critiquing a presenter's vocal pattern or language usage after or during a presentation is not the same as using extracted audio/text information in strategic-message- instance generation. The Office Action therefore does not establish anticipation of claim 31. Withdrawal of the rejection is respectfully requested.
The Examiner respectfully disagrees. The Applicant’s arguments relate to general allegations of patentability. For example, the Applicant makes general statements as to why the disclosure of Seleskerov is different. But does not indicate how they are different. The use of “pacing, vocal pattern, volume, language usage” ALL clearly relate to “extracting linguistic, tonal and contextual information” from audio and text data so its unclear what exactly is not present in Seleskerov. The limitation of “used by the strategic message model” is intended use of the analyzing. However, Seleskerov still teaches this asof Seleskerov. This is clearly a “message instance” as claimed contrary to Applicant’s argument as the recommendations.
Applicant asserts on page 26, with respect to claim 32:
Claim 32 recites: "The computer implemented method of claim 1, further comprising: processing the multimedia data including audio and text data to identify patterns, relationships, or context within the captured content, and providing the processed information for application by the machine learning that formulates or updates the strategic message model."
The Office Action rejects claim 32 as anticipated by Seleskerov. The Office Action relies on Seleskerov's disclosure of presentation media streams including audio and presentation content, Seleskerov's presentation coaching unit providing feedback critiques regarding pacing, vocal pattern, volume, monotone speech, and language usage, and Scleskerov's delivery-attributes model analyzing audio, video, and presentation content with machine learning models trained to identify aspects of presentation skills and presentation content that are good or may benefit from improvement.
Applicant respectfully traverses the rejection of claim 32. Claim 32 requires processing multimedia data including audio and text data to identify patterns, relationships, or context within the captured content, and providing that processed information for application by the machine learning that formulates or updates the strategic message model. The claim is therefore directed to deriving content- characterizing information from the claimed multimedia data and applying that information in the formulation or updating of the strategic message model.
The cited portions of Seleskerov do not disclose this limitation. Providing feedback critiques on presentation skills is not identifying patterns, relationships, or context within captured strategist- discussion content for use in formulating or updating a strategic message model. Similarly, a delivery- attributes model that evaluates presentation skills and presentation content does not disclose the claimed use of processed information by machine learning that formulates or updates a strategic message model.
The Office Action's mapping again rests on treating Seleskerov's presentation-coaching models as if they were the claimed strategic message model. But Seleskerov's delivery-attribute and presentation- coaching models are used to provide feedback for improving an online presentation and presenter skills. The rejection does not identify a model formulated from a live strategist discussion, does not identify updating of that strategic message model, and docs not identify processed audio/text information being provided for application by such machine learning.
Accordingly, Seleskerov does not disclose all limitations of claim 32 arranged as claimed. Withdrawal of the rejection of claim 32 is respectfully requested.
The Examiner respectfully disagrees for the same reasons as mentioned above for claim 31. Applicant continues to indicate there is a distinction between “coaching models: and “strategic message models”. The examiner as noted above fails to see the difference as asserted by the applicant. Also, Seleskerov teaches the processing of multimedia data to identify patterns, relationships or context. This is provided in para [0038]. The last portion of the claim and emphasized by the Applicant of “claimed use of processed information by machine learning that formulates or updates a strategic message model” is also taught by Seleskerov whereby using this information in order to provide recommendations/feedback on the presentation as provided in [0039].
The Applicant asserts on page 27, with respect to claim 33:
Claim 33 recites: "The computer implemented method of claim 1, further comprising: processing audio and text portions of the input data stream to derive representative information characterizing the discussion, and using the derived information as part of the training data applied to the strategic message model." The Office Action rejects claim 33 as anticipated by Seleskerov.
The Office Action relies on Seleskerov's disclosure that presentation media streams may include an audio component and presentation content, that a presentation coaching unit may provide feedback critiques on pacing, vocal pattern, volume, monotone specch, and language usage, and that a delivery-attributes model may analyze audio, video, and presentation content with machine learning models trained to identify aspects of presentation skills and presentation content that are good or may benefit from improvement.
Applicant respectfully traverses the rejection of claim 33. Claim 33 requires processing audio and text portions of the input data stream to derive representative information characterizing the discussion, and using that derived information as part of the training data applied to the strategic message model. This limitation ties the processed audio/text information to characterization of the claimed discussion and to training of the claimed strategic message model.
The cited portions of Seleskerov do not disclose these requirements. Seleskerov's presentation coaching unit evaluates or critiques presentation skills. Seleskerov's delivery-attributes model analyzes presentation-related content to identify presentation skills or content that may benefit from improvement. These disclosures do not teach deriving representative information characterizing a live strategist discussion, and they do not teach using such derived information as training data applied to a strategic message model.
The Office Action does not identify any disclosure in Seleskerov in which audio and text portions of a strategist-discussion input stream are processed to characterize the discussion itself. Nor does the rejection identify where Seleskerov applies such derived discussion-characterizing information as training data to a strategic message model. The cited disclosure of a model trained to identify presenter-skill issues is not a disclosure of training a strategic message model using representative information characterizing a strategist discussion.
Accordingly, Seleskerov does not disclose all limitations of claim 33 arranged as claimed. Withdrawal of the rejection of claim 33 is respectfully requested.
The Examiner respectfully disagrees. The examiner refers the Applicant to similar arguments presented in claims 31-32. Again, the Applicant is making general allegations of patentability without providing explicit evidence on how the cited portions of Seleskerov are different from that of the claims. Clearly, each of the limitations have all been addressed and have been understood by the Applicant as noted above in the arguments presented and for which have been mapped by in the prior Office Action. Since these arguments are repetitive in nature the Examiner will not repeat the position henceforth.
Therefore, the Applicant appears to understand for each claim the specific interpretation being made by the Examiner and yet the Applicant appears to note that the examiner has failed to identify how “audio and text portions of a strategist-discussion input stream are processed to characterize the discussion itself. Nor does the rejection identify where Seleskerov applies such derived discussion-characterizing information as training data to a strategic message model”. Other than the repeated arguments of “strategist discussion” and “strategic message model” being different from that of Seleskerov the Applicant has not provided arguments that actually indicate claimed differences and how they are different. Furthermore, the Specification does not present specific evidence to the contrary of the interpretations made by the examiner since the overall goal in both the reference and the application is to provide adaptive delivery and content generation.
Hence, ALL of the Applicant’s arguments have been addressed and are not persuasive.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claim 27 is rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention. More specifically, the Applicant has amended claim 27 to include several limitations. The limitations of concern being “the strategic message instance comprising presentation-control data…” and “wherein the updating modifies one or more weighting values of the strategic message model based on the detected deviation and reaction by the audience”. More specifically, with respect to “presentation-control data”, there is no mention of this limitation anywhere in the claims to understand the full scope and meaning. Therefore, the Examiner understands that the Applicant can be their own lexicographer however when amendments after the original filing are made they should correlate back to the specification or use language used specifically by the specification as filed. Therefore, it is unclear what the exact scope of this term is. Furthermore, with respect to the last limitation of the claim of the updating of the “strategic message model” this is specifically not found in the specification. The closest support comes from published spec para [0047]. However, this section speaks of training a strategic message model based on newly updated content in the base content archive to update the model. Para [0057]-[0058] and [0024] are the only paragraphs that make mention of topic. However, these paragraphs describe the weighting of topics based on relative significance and which can be adjusted. But this is not the same as “wherein the updating modifies one or more weighting values of the strategic message model based on the detected deviation and reaction by the audience”. This was not found to be supported anywhere in the specification.
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 and 27 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. Independent claims 1 and 8 relate to the statutory category of method/process. Claims 1 and 27 recite ”obtaining input data stream capturing a live discussion by one or more strategists on topics of strategic messages to deliver to an audience, the input data stream comprising multimedia data generated by meeting data (input/output) devices located at the live discussion; formulating a strategic message model by machine learning; generating a strategic message instance by use of the strategic message model; monitoring a presentation of the strategic message instance by a presenter by detecting delivery issues; and updating the strategic message model based on a delivery content capturing the presentation and a reaction by the audience”.
With regards to Claims 1 and 27:
“obtaining input data stream capturing a live discussion by one or more strategists on topics of strategic messages to deliver to an audience, the input data stream comprising multimedia data generated by meeting data devices located at the live discussion” as drafted covers mental activity. A human is present and observing/monitoring a live presentation on a particular topic. The presentation can include power point slides, audio recordings, etc. to the audience attending the presentation who are using a computer/other devices to attend and interact with the presentation.
“formulating a strategic message model by machine learning” (claim 1)
“formulating a strategic message model by machine learning including extracting discussion features from the multimedia data” (claim 27) as drafted covers mental activity. A human can determine by observing the presentation, what the topic/message of the presentation is, and what information needs to be presented to the audience and to be used to learn different patterns of the data. The additional limitation of using machine learning to formulate a strategic message model does not provide an inventive concept. The strategic message model is described in paragraph [0047] of the as filed specification as a generic machine learning model. It can be said that a human brain can also be a learning model that can be trained.
“generating a strategic message instance by use of the strategic message model” (claim 1)
“generating a strategic message instance by use of the strategic message model, the strategic message instance comprising presentation-control data for guiding delivery of the strategic message instance” (claim 27)
as drafted covers mental activity. A human can determine what the topic/message is being presented and how to present the information based on patterns determined from past example/patterns.
“monitoring a presentation of the strategic message instance by a presenter by detecting delivery issues” (claim 1)
“monitoring a presentation of the strategic message instance by a presenter by detecting delivery issues including detecting a deviation between the presentation control data and delivery content captured during the presentation” (claim 27)
as drafted covers mental activity. A human can, by observing the audience, determining if the topic/message of the discussion is getting thru to the audience based on the feedback and reactions by the audience as well as following a script and ensuring points of the presentation are being delivered.
“updating the strategic message model based on a delivery content capturing the presentation and a reaction by the audience” (claim 1)
“updating the strategic message model based on a delivery content capturing the presentation and a reaction by the audience, wherein the updating modifies one or more weighting values of the strategic message model based on the detected deviation and the reaction by the audience” (claim 27)
as drafted covers mental activity. A human can, by observing the audience’s reaction or feedback, update/change what the topic/message is and/or update/change how the topic/message is delivered, where such modifies a specific pattern by biasing certain delivery aspects and content due to the weights.
This judicial exception is not integrated into a practical application. The additional limitation of using machine learning to create the strategic message model does not provide an inventive concept. The strategic message model is described in paragraph [0047] of the as filed specification as a generic machine learning model. Accordingly, the additional element does not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to the integration of the abstract idea into a practical application, the additional element of using machine learning to create the strategic message model is noted. Mere instructions to apply a generic machine learning model cannot provide an inventive concept. The additional limitations in the claims noted above are directed towards insignificant pre-solution activity. The claims are not patent eligible.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1, 21, 22, 24, 25, and 28-33 is/are rejected under 35 U.S.C. 102(a)(2) as being anticipated by Seleskerov et al. (US 2022/0138470).
Regarding Claim 1, Seleskerov et al discloses a computer implemented method comprising: obtaining input data stream capturing a live discussion by one or more strategists on topics of strategic messages to deliver to an audience, the input data stream comprising multimedia data generated by meeting data devices located at the live discussion (FIG. 3 is a diagram showing examples of data exchanged between the presentation and communications platform 110 and the client devices 105a, 105b, 105c, and 105d. As discussed in the preceding examples, the presentation and communications platform 110 may transmit one or more presentation media streams 305 to the each of the client devices 105 over the network 120. The one or more presentation media streams 305 may include one or more audio media streams, one or more video media streams, and/or other media streams) (page 8, paragraph [0064]); formulating a strategic message model by machine learning (The presentation coaching unit 235 may utilize a delivery attributes model 1170 to analyze audio, video, and presentation content with machine learning models trained to identify aspects of the presenter's presentation skills and the presentation content are good and those that may benefit from improvement) (page 4, paragraph [0038]); generating a strategic message instance by use of the strategic message model (The presentation content may include a set of slides, a document, or other content that may be discussed during presentation) (page 8, paragraph [0064]); monitoring a presentation of the strategic message instance by a presenter by detecting delivery issues (The speaker skills feedback unit 1630 may identify issues with the presenter's presentation style, such as the language usage, language patterns, monotone delivery, reading of slide content, emotional state of the presenter, eye contact and/or gaze direction of the presenter, body pose of the presenter, and/or other information about the presenter and/or the participants) (page 6, paragraph [0049]); and updating the strategic message model based on a delivery content capturing the presentation and a reaction by the audience (Returning to FIG. 2, the model updating unit 220 may be configured to update the slide attribute model 1180 and/or the delivery attributes model 1170 based on the participant reaction information determined by the stream processing unit 215. The slide attribute model 1180 and/or the delivery attributes model 1170 may analyze the online presentation, and the presentation designer unit 230 and the presentation coaching unit 235 may use the inferences output by the slide attribute model 1180 and/or the delivery attributes model 1170 to provide feedback to the presenter for improving the online presentation content and/or the presentation skills of the presenter. The model updating unit 220 may utilize the reaction data obtained from the participants of the online presentation to improve the recommendations provided by the slide attribute model 1180 and/or the delivery attributes model 1170) (page 6, paragraph [0052]).
Regarding Claim 21, Seleskerov et al discloses the method, further comprising: processing the input data stream by preparing manageable data units classified according to media types (The frame and filtering preprocessing unit 410 may be configured to determine whether a particular media stream contains audio, video, or both at a particular time and to process the stream using to convert the media stream into an appropriate format to serve as an input to the machine learning models for analyzing that type of content) (page 9, paragraph [0071]), wherein the processing includes: extracting audio streams from video media (The frame and filter preprocessing unit 410 may be configured to perform feature extraction on the media streams and/or reaction data) (page 8, paragraph [0068]); and parsing video frames prior to formulating the strategic message model (Initially, the frame and filtering preprocessing unit 410 may process the stream vi to generate an input or inputs for models that process features from video content) (page 9, paragraph [0071]).
Regarding Claim 22, Seleskerov et al discloses the method, wherein the processing of multimedia data includes: identifying speakers and language within audio media (The language usage detection model 620 may be configured to analyze features extracted from video content of the presenter or a participant to identify language usage of that person and to output high-level features information that represents the language usage) (page 10, paragraph [0080]); transcribing the audio media data ( A transcript 715 of the audio portion of the online presentation and/or communication session may be generated by the stream processing unit 215 by analyzing the spoken content provided by the presenter and the participants) (page 11, paragraph [0084]); and synchronizing text media with corresponding video and audio streams based on timestamps (Furthermore, the reactions information generated by the analyzer unit 415 by analyzing the audio content, video content, and/or multi-modal content captured by the client devices 105 of the participants may also include a timestamp indicating when each reaction occurred) (page 5, paragraph [0047])..
Regarding Claim 24, Seleskerov et al discloses the method, wherein monitoring the presentation comprises: detecting delivery issues based on real-time capture of presentation conditions (For example, a participant may utter the word “what?” or utterance “huh?” during the presentation if they do not understand something that is being presented. The feedback and reporting unit 225 may be configured to maps this reaction to a “confused” reaction that may be sent to the client device 105a of the presenter to help the presenter to gain an understanding that at least some of the participants may be confused by a portion of the presentation) (page 10, paragraph [0080]); and responding in real time by modifying at least one of: a message format, timing, prioritization, or delivery channel (The presentation coaching unit 235 may provide suggestions for alternative language and/or language to be avoided during a presentation) (page 10, paragraph [0080]), wherein the modification is selected based on metadata of the presented content and a delivery profile associated with the session (With respect to the participants, the feedback and reporting unit 225 may be configured to identify certain language usage of a participant as being a reaction that may be sent to the client device 105a of the presenter to help the presenter to gain an understanding of the audience engagement in near real time during the presentation) (page 10, paragraph [0080]).
Regarding Claim 25, Seleskerov et al discloses the method, wherein the processing comprises: preparing manageable data units classified by media type (The frame and filtering preprocessing unit 410 may be configured to determine whether a particular media stream contains audio, video, or both at a particular time and to process the stream using to convert the media stream into an appropriate format to serve as an input to the machine learning models for analyzing that type of content) (page 9, paragraph [0071]); extracting audio streams (The frame and filter preprocessing unit 410 may be configured to perform feature extraction on the media streams and/or reaction data) (page 8, paragraph [0068]); parsing video frames (Initially, the frame and filtering preprocessing unit 410 may process the stream vi to generate an input or inputs for models that process features from video content) (page 9, paragraph [0071]); identifying speakers and language (The language usage detection model 620 may be configured to analyze features extracted from video content of the presenter or a participant to identify language usage of that person and to output high-level features information that represents the language usage) (page 10, paragraph [0080]); transcribing audio ( A transcript 715 of the audio portion of the online presentation and/or communication session may be generated by the stream processing unit 215 by analyzing the spoken content provided by the presenter and the participants) (page 11, paragraph [0084]); synchronizing text media with corresponding streams (Furthermore, the reactions information generated by the analyzer unit 415 by analyzing the audio content, video content, and/or multi-modal content captured by the client devices 105 of the participants may also include a timestamp indicating when each reaction occurred) (page 5, paragraph [0047]); and generating metadata for each data unit (The one or more presentation media streams 305 may include one or more audio media streams, one or more video media streams, and/or other media streams. The one or more presentation media streams may include an audio component of the presentation where the presenter is discussing presentation content being shared with the participants) (page 8, paragraph [0064]); and wherein monitoring includes: detecting delivery issues (For example, a participant may utter the word “what?” or utterance “huh?” during the presentation if they do not understand something that is being presented. The feedback and reporting unit 225 may be configured to maps this reaction to a “confused” reaction that may be sent to the client device 105a of the presenter to help the presenter to gain an understanding that at least some of the participants may be confused by a portion of the presentation) (page 10, paragraph [0080]) and responding in real time by modifying a delivery characteristic selected based on the metadata (The presentation coaching unit 235 may provide suggestions for alternative language and/or language to be avoided during a presentation) (page 10, paragraph [0080]).
Regarding Claim 28, Seleskerov et al discloses the method, further comprising: processing the input data stream obtained from the meeting data devices to distinguish among different types of content contributed during the live discussion ( The presentation and communications platform 110 implements an architecture for efficiently analyzing audio, video, and/or multimodal media streams and/or presentation content) ([age 3, paragraph [0034]), and preparing structured representations of such content suitable for use by the machine learning that formulates the strategic message model ( A technical benefit of this architecture is the media streams and/or presentation content may be analyzed to extract feature information for processing by the various models, and the high-level feature information output by the models may then be utilized by both the presentation coaching unit 235 and the presentation hosting unit 240) (page 3, paragraph [0035]).
Regarding Claim 29, Seleskerov et al discloses the method, wherein the method includes processing the multimedia data, wherein the processing the multimedia data includes identifying speakers and linguistic content within audio data (The speaker skills feedback unit 1630 may identify issues with the presenter's presentation style, such as the language usage, language patterns, monotone delivery, reading of slide content, emotional state of the presenter, eye contact and/or gaze direction of the presenter, body pose of the presenter, and/or other information about the presenter and/or the participants) ([age 6, paragraph [0049]), converting the identified speech into text ( A transcript 715 of the audio portion of the online presentation and/or communication session may be generated by the stream processing unit 215 by analyzing the spoken content provided by the presenter and the participants) (page 11, paragraph [0084]), and providing the resulting text for use by the machine learning that formulates the strategic message model (The models may be configured to receive feature data extracted from the presentation media streams 305, the participant media streams 310, and/or the reactions data 315) (page 9, paragraph [0074]).
Regarding Claim 30, Seleskerov et al discloses the method, further comprising: processing the multimedia data to detect spoken content and corresponding textual input (The one or more presentation media streams may include an audio component of the presentation where the presenter is discussing presentation content being shared with the participants. The presentation content may include a set of slides, a document, or other content that may be discussed during presentation) (page 8, paragraph [0064]), and generating text based data derived from audio content ( A transcript 715 of the audio portion of the online presentation and/or communication session may be generated by the stream processing unit 215 by analyzing the spoken content provided by the presenter and the participants) for use in training the strategic message model (page 11, paragraph [0084]) (The models may be configured to receive feature data extracted from the presentation media streams 305, the participant media streams 310, and/or the reactions data 315) (page 9, paragraph [0074]).
Regarding Claim 31, Seleskerov et al discloses the method, wherein the multimedia data includes audio and text data, and the processing comprises analyzing the audio and text data to extract linguistic, tonal, and contextual information used by the strategic message model in generating the strategic message instance (The presentation coaching unit 235 may provide feedback critiques on aspects of the presentation skills, such as but not limited to pacing, vocal pattern, volume, whether the presenter is speaking in monotone, and/or language usage) (page 4, paragraph [0038]).
Regarding Claim 32, Seleskerov et al discloses the method, further comprising: processing the multimedia data including audio and text data (The one or more presentation media streams may include an audio component of the presentation where the presenter is discussing presentation content being shared with the participants. The presentation content may include a set of slides, a document, or other content that may be discussed during presentation) (page 8, paragraph [0064]) to identify patterns, relationships, or context within the captured content (The presentation coaching unit 235 may provide feedback critiques on aspects of the presentation skills, such as but not limited to pacing, vocal pattern, volume, whether the presenter is speaking in monotone, and/or language usage) (page 4, paragraph [0038]), and providing the processed information for application by the machine learning that formulates or updates the strategic message model (The presentation coaching unit 235 may utilize a delivery attributes model 1170 to analyze audio, video, and presentation content with machine learning models trained to identify aspects of the presenter's presentation skills and the presentation content are good and those that may benefit from improvement) (page 4, paragraph [0038]).
Regarding Claim 33, Seleskerov et al discloses the method, further comprising: processing audio and text portions of the input data stream (The one or more presentation media streams may include an audio component of the presentation where the presenter is discussing presentation content being shared with the participants. The presentation content may include a set of slides, a document, or other content that may be discussed during presentation) (page 8, paragraph [0064]) to derive representative information characterizing the discussion (The presentation coaching unit 235 may provide feedback critiques on aspects of the presentation skills, such as but not limited to pacing, vocal pattern, volume, whether the presenter is speaking in monotone, and/or language usage) (page 4, paragraph [0038]), and using the derived information as part of the training data applied to the strategic message model The presentation coaching unit 235 may utilize a delivery attributes model 1170 to analyze audio, video, and presentation content with machine learning models trained to identify aspects of the presenter's presentation skills and the presentation content are good and those that may benefit from improvement) (page 4, paragraph [0038]).
Allowable Subject Matter
Claims 15-20 are allowed.
The following is a statement of reasons for the indication of allowable subject matter: Claim 15 teaches similar subject matter as the prior art of Seleskeroy et al. (US 2022/0138470), Daredia et al. (US 2020/0403817, and Advani et al. (US 2018/0145840). However, the prior art fails to teach “formulating a strategic message model by machine learning based on the manageable data units produced by the multimedia content processor, the strategic message model comprising topics and corresponding attributes associated with a topic profile, wherein the formulating includes applying cognitive analysis separately to each media type while maintaining synchronization and fidelity normalization across video resolution and audio volume” and “updating the strategic message model based on a delivery content capturing the presentation and a reaction by the audience wherein the updating includes adapting topic profiles based on audience engagement and feedback associated with the metadata and timestamps recorded during the session” as recited in claim 15.
Claims 16-20 are allowed for being dependent on an allowable base claim.
Claims 23 and 26 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Claim 27 would be allowable if rewritten or amended to overcome the rejection(s) under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), 1st paragraph, and 35 USC 101 set forth in this Office action. More specifically, the closest prior art of record Seleskerov teaches all of the limitations except for “updating the strategic message model based on a delivery content capturing the presentation and a reaction by the audience, wherein the updating modifies one or more weighting values of the strategic message model based on the detected deviation and the reaction by the audience.” Seleskerov does update the model as noted in [0056] but its not clear that its “weighting values” based on “detected deviation and the reaction by the audience”. This allowable subject matter is conditioned on the Applicant’s ability to provide clear support for the limitation noted in the rejections above.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Peng et al. (“Say It All: Feedback for Improving Non-Visual Presentation Accessibility”) is cited to disclose analyzing of presentation slides and transcribing of presenter speech and receiving post presentation feedback (see Figure 1) in order to improve presenter presentation.
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.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PARAS D SHAH whose telephone number is (571)270-1650. The examiner can normally be reached Monday-Thursday 7:30AM-2:30PM, 5PM-7PM (EST), Friday 8AM-noon (EST).
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
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.
/Paras D Shah/Supervisory Patent Examiner, Art Unit 2653
07/28/2026
Prosecution Insights