DETAILED ACTION
This Office Action is in reply to Applicants response after Non-Final rejection received on July 23, 2026. Claim(s) 1-20 is/are currently pending in the instant application.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Response to Amendment
The Examiner acknowledges the Applicants filing of supplemental amendments on July 23, 2026. Claims 1, 3, 5-14, 17, and 18 are amended. No claims are canceled at this time.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to a judicial exception (i.e., a law of nature, a natural phenomenon, or an abstract idea) without significantly more.
Claims 1-20 are directed to one of the four statutory classes of invention (e.g. process, machine, manufacture, or composition of matter). The claims include a system or “apparatus”, method or “process”, or product or “article of manufacture” and is delivery aware audience segmentation which is a process (Step 1: YES).
The Examiner has identified independent method Claim 8 as the claim that represents the claimed invention for analysis and is similar to method Claim 1 and apparatus Claim 13. Claim 8 recites the limitations of (abstract ideas highlighted in italics and additional elements highlighted in bold)
obtaining training data including activity data, conversion data, and content reach data;
computing a conversion loss by comparing predicted outcome data with actual outcome data;
training, in a conversion training phase, a conversion predictor of a machine learning model using the conversion loss;
computing a reach loss by comparing predicted reach data with actual reach data; and
training, in a reach training phase following the conversion training phase, a reach predictor of the machine learning model using the reach loss.
These limitations, under their broadest reasonable interpretation, cover performance of the limitation as mental processes. Obtaining multiple types of data, calculating conversion loss, training a predictor using conversion loss, computing a reach predictor based on reach loss recites a concept performed in the human mind. But for the “machine learning model”, the claim encompasses collecting data, performing calculations, developing algorithms or equations for conversion and reach, and developing a model by computing reach loss using his/her mind and/or pen and paper. The mere nominal recitation of conversion prediction and reach prediction being performed by generic machine learning model using known techniques does not take the limitations out of the mental processes grouping. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as a concept performed in the human mind, then it falls within the “Mental Processes” grouping of abstract ideas. Accordingly, the claim recites an abstract idea. The processor and memory including instructions executed by the processor in Claim 13 is just applying generic computer components to the recited abstract limitations. The method using a machine learning model in Claim 8 appears to be just software. Claims 1 and 13 are also abstract for similar reasons. (Step 2A-Prong 1: YES. The claims are abstract)
These limitations, under their broadest reasonable interpretation, cover performance of the limitation as certain methods of organizing human activity. Obtaining multiple types of data, calculating conversion loss, training a predictor using conversion loss, computing a reach predictor based on reach loss recites a commercial interaction. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation as a commercial interaction, then it falls within the “Certain Methods of Organizing Human Activity” grouping of abstract ideas. Accordingly, the claim recites an abstract idea. The processor and memory including instructions executed by the processor in Claim 13 is just applying generic computer components to the recited abstract limitations. The method using a machine learning model in Claim 8 appears to be just software. Claims 1 and 13 are also abstract for similar reasons. (Step 2A-Prong 1: YES. The claims are abstract)
This judicial exception is not integrated into a practical application. In particular, the claims only recite a machine learning model and a media channel(Claims 1), a machine learning model (Claim 8) and/or a processor and a memory storing instructions (Claim 13). The computer hardware is recited at a high-level of generality (i.e., as a generic processor performing a generic computer function) such that it amounts no more than mere instructions to apply the exception using a generic computer component. Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Therefore claims 1, 8, and 13 are directed to an abstract idea without a practical application. (Step 2A-Prong 2: NO. The additional claimed elements are not integrated into a practical application)
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because, when considered separately and as an ordered combination, they do not add significantly more (also known as an “inventive concept”) to the exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of using a computer hardware amounts to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept. See Applicant’s specification para. [0031] about implantation using general purpose or special purpose computing devices [a server comprises a general purpose computing device, a personal computer, a laptop computer, a mainframe computer, a super computer, or any other suitable processing apparatus.] and MPEP 2106.05(f) where applying a computer as a tool is not indicative of significantly more. Accordingly, these additional elements, when considered separately and as an ordered combination, do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. Thus claims 1, 8, and 13 are not patent eligible. (Step 2B: NO. The claims do not provide significantly more)
Dependent claims 2-7, 9-12, and 14-20 further define the abstract idea that is present in their respective independent claims 1, 8, and 13 and thus correspond to Mental Processes and/or Certain Methods of Organizing Human Activity and hence are abstract for the reasons presented above. The dependent claims do not include any additional elements that integrate the abstract idea into a practical application or are sufficient to amount to significantly more than the judicial exception when considered both individually and as an ordered combination. The dependent claims include steps or processes which are similar to that disclosed in MPEP 2106.05(d), (f), (g), and/or (h) which include activities and functions the courts have determined to be well-understood, routine, and conventional when claimed in a generic manner, or as insignificant extra solution activity, or as merely indicating a field of use or technological environment in which to apply the judicial exception. Therefore, the claims 2-7, 9-12, and 14-20 are directed to an abstract idea. Thus, the claims 1-20 are not patent-eligible.
Response to Arguments
The Applicants remarks begin on page 9 of the supplemental response on July 23, 2026. The Applicant begins with a status of the claims and a summary of the interview on July 15, 2025.
Moving on, the arguments being on page 9 with the Applicant summarizing the 35 U.S.C § 101 and the Applicant cites the August 4, 2025, memorandum on “reminders on evaluating subject matter eligibility of claims under 35 U.S.C § 101” regarding AI and machine learning arts and Example 39’s training of a neural network as not reciting a judicial exception. Applicants state that based on the content of the interview and amendment to the claims the structured training steps including training conversion predictor and training reach predictor provide optimal content delivery. Further, the Applicant cites a solution of phase dependent training as an improvement to machine learning.
The arguments move to Step 2A Prong 1 (remarks page 10) where the Applicant asserts that the claims do not recite a mental process or certain methods of organizing human activity.
Regarding the grouping under mental processes, the Applicants arguments repeat the position based on the August 4th memorandum and citing example 39 as reasoning that the training limitations do not set forth mathematical relationships, formulas, or equations. The arguments further state that the claim limitations do not recite specific algorithms by name.
Further, the position is that based on the content of the interview the amendments recite the staged training without invoking mathematical formulas, the amended claims recite trained machine learning models. The training is used content reach data of a product segment-specific match rate and a segment-specific exposure rate over a plurality of media channels. A per-segment, per-channel computation across a plurality of channels cannot be performed in the mind (remarks page 11).
The Examiner is not in agreement. To begin, the prior rejection under 35 U.S.C § 101 did not mention or enlist the mathematical concepts grouping. Therefore, the argument that the claims do not mention or require mathematical equations or formulas and not persuasive. The Examiner acknowledges that the Applicants position if calculating conversion is purely mathematical, see [0097], where the selector is also a product of a mathematical equation. Simply because the Applicant didn’t include the mathematical functions into the claims, does not in itself remove the claim form the possibility of being considered in part a mathematical concept.
Second, the Examiner does not agree that the claims are similar to Example 39. In the example the training stages are a first training data set comprising collected digital facial images, modified digital facial images, and non-facial digital images. In the example the training set is modified between the first training and the second training to demonstrate as a result, improvement in the model. By comparison the claim is substantially different. The claims include that the model is trained with conversion and content reach data. Next, conversion loss is calculated with predicted and actual outcome data, however the reach predictor is not producing conversion prediction. Then, the training the conversion phase a conversion predictor using conversion loss data. It’s not clear where the data is coming from for these steps.
The arguments move one with the position that the claims do not recite a method of organizing human activity (remarks page 11). The Applicants state that multi-objective optimization is a well-recognized technical challenge in machine learning. When optimizing for two objectives, improving one may degrade the other. The argument includes that the present claims address a technical problem of optimizing segmentation for conversion prediction accuracy and content reach. The alleged solution is said to be multi-phase training that produces a machine learning model capable of jointly optimizing the two stated objectives. The Applicant concludes that the specification does not transform a technical improvement into a business method.
The Examiner does not agree that the analysis for eligibility under 35 U.S.C § 101 would transform a claim into a business method. As previously discussed in the interview “Business Methods” is simply an umbrella category of classifications which are related to common business practices. Related to certain methods of organizing human activity, the disclosure is about predicting conversion and content reach. Conversion and reach are elements of advertising and are equal to the number of potential customers reached and the conversion of the reaches customers to buy the product or service advertised. This is in the space of commercial interaction / marketing / sales activities.
The disclosure and the argument that the key aspect is multi-variable optimization (also known as Pareto optimization) relates to mathematical functions. These problems and subsequent optimizations are well known in the areas of Economics/ Finance, Engineering Optimal Control or Optimal Design, Process Optimization, Electrical Power Systems, and Radio Resource Management to name a few. Simply applying a computer, and in this direct case a machine learning mode, to ingest the reach and conversion data in order to make accurate predictions is not integrate the claim to a practical application. In this case the computer is used as a tool to perform the calculations necessary and to do so quicker than a person could perform them.
The arguments move on to Step 2A, Prong 2 (remarks page 12) where the Applicant states that the claims are integrated into a practical application of improving machine learning and selecting a media channel with the model. The arguments recite claim 1 and take the position that the specific limitations embody novelty of the invention rather than nominal or generic computer steps. The argument further point to several specific limitations: (i) computing conversion loss by comparing data and training a conversion predictor in a training phase using the conversion loss; (ii) computing reach loss comparing reach data and training a reach predictor using reach loss; (iii) content reach data comprising a product of segment-specific match rate and segment-specific exposure rate over media channels; (iv) generating a reach prediction for the user segment, selecting a media channel, and providing content. Applicant alleges this is a particular order training method for use of a trained model to output a select communication channel and deliver content. Applicant also states these features are not present in conventional segmentation systems.
The remarks on page 13 include specification citations for the described improvements. First, Applicants point to training the reach predictor to jointly optimize user segmentation for two objectives. The argument provides a citation from the specification where “jointly optimizes user segmentation for conversion and reach yields more efficient content delivery as comparted to the existing content audience segmentation systems”.
Second, the argument recites content reach data, what is a measure of users who matched in the media and interacted with the media, and this “enables content providers to target more users within an audience with their rendered content within a media channel”.
Third, generating reach prediction and selecting a media channel based on the prediction is provides the specification described improvement in content delivery. For these reasons the Applicant asserts practical application under Step 2A, Prong 2.
The Examiner is not persuaded. The application of machine learning, which requires a computer or processor while not expressly claiming one, using generic and well known techniques does not constitute integration into practical application, rather simply applying a computer as a tool. Regarding specific limitations (i) the claim fails to include where actual data and predicted data is from and the training step comes after the prediction, this leaving the question of where the prediction happens. (ii) the reach loss comparison is actual and predicted and the training is the most basic training a model with available data. (iii) the product of match rate and exposure rate is drawn to the mathematical formulas and calculations in [0109]. When executed by a computer it’s not more than use as a tool to perform complex calculations. (iv) generating a reach prediction and selecting a channel to provide content is alone or in combination does not demonstrate improvement to the machine learning.
The argument for joint optimizing for two objectives is not more than performing mathematical calculations and providing users with targeted content. The Applicants tie to a generic technological field is not sufficient to show integration to practical application. The machine learning lacks detail regarding the training data and training steps and is only providing the most basic step of a model trained by conversion and reach data or a model trained by conversion loss data.
The arguments conclude with Step 2B arguing the inventive concept is significantly more than the judicial exception. Applicants argument is that neural network architectures including multi-layer and hierarchical attention networks would be recognized by one of skill in the art to involve deep complexity. Training such models for optimization is a technical challenge and can lead to suboptimal results. The arguments state that the training represents an unconventional implementation of a sequence of training where later phases depend on earlier phases.
The Examiner disagrees. The claims do not include any mention of neural networks or their underlying details of multi-layering or hierarchal structure. The claims simply use the most generic and basic term “machine learning” as a placeholder. Further, optimization is not exactly a technical challenge when it’s implemented by a generic computer device.
Regarding the multi-step training, the Examiner does not agree that the defined sequence of phases is clear. The Applicant has model generation, calculation, training steps which (a) do not clearly tie together and expressly stating the phases of training to reach the result. (b) the steps also seem like they could be out of order as the model is trained first and then conversion and reach training steps occur later. The claims do not sufficiently tie together the steps as the argument alleges. There is no functional details of the model training. There is a lack of linking the training phases together and there is a lack of sufficient detail in order for the machine learning to not be considered generic machine learning.
In summary, the claims remain rejected under 35 U.S.C § 101. The claims are not in condition for allowance.
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 DYLAN C WHITE whose telephone number is (571)272-1406 and email dylan.white@uspto.gov. The examiner can normally be reached M-F 7:30-4:00 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.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Beth Boswell can be reached on (571)272-6737. 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.
/DYLAN C WHITE/Primary Examiner, Art Unit 3625 September 14, 2026
Prosecution Insights