DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
Claims 1, 4, 8, 13 and 16 have been amended and are hereby entered.
Claims 1-20 are pending and have been examined.
This action is made FINAL.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on July 15, 2025 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Response to Arguments
Applicant's arguments filed October 7, 2025 have been fully considered but they are not persuasive.
Regarding the applicant's arguments against the 101 rejection of claims 1-20 on pages 12-16: Applicant’s arguments directed to Step 2A prong 1 and Step 2A prong 2 analysis were considered. However, these arguments are not persuasive and the Examiner respectfully disagrees for the following reasons:
For Step 2A-Prong 1 and Prong 2 starting in p. 13: The Applicant argues that the claims do not recite a judicial exception and further compares claims 1, 8 and 15 amended steps to Example 39 from the 2019 PEG for the limitation of “training of a neural network" because the claim do not recite none of the abstract idea identified by the Examiner (excluding other abstract ideas that the Applicant alleges where identified such as “mathematical concepts”). However, the Examiner find these arguments unpersuasive. Firstly, because the Applicant is applying an outdated version from the latest Subject Matter Eligibility (SME) guidance issued on July 2024. But also, since the Applicant failed to apply the MPEP 2106.04, subsection II, which further states that the claims are “reciting” a judicial exception when the judicial exception is “set forth” or “described” in the claim. Thus, when analyzing these features of a “generated RNN” and “ML model” recited claim language from the claim limitations in claims 1, 8 and 16, are directed to mental processes for “analyzing” the contentious statement with an RNN that is used to generally apply the abstract idea without placing any limits on how the RNN functions. Rather, the limitation only recites the outcome of analyzing contentious statements, without providing details on how that analyzing function is accomplished. Similarly, the step reciting the “machine learning model” is simply describing the provenance from which the time-series data was outputted without actively and positively reciting and discussing how the ML model achieved such output and how it trained and forecasted the respective data. Thus, the steps do not negate and further still reads in the mental nature of the limitation(s), when analyzing such information and outputting time-series data. Moreover, the concept is merely claimed to be performed on a generic computer with the use of RNN and ML models that are merely used as a tool to perform the concept of data analysis and time-series data outputs (see MPEP 2106.04(a)(2)(III)(B & C)).
Finally, the recitation of “…via a generated recurrent neural network (RNN)…” to analyze the contentious statement, merely indicates a field of use or technological environment in which the judicial exception is performed. Because such 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). Therefore, the amendments did not limit the claimed scope and does not improve the computer functioning of the "computing device" utilizing RNN and ML models or the technological field for peer pressure detection and analytics systems. Rather, the limitations merely recite the functions being performed by generic computer components (e.g. clearly invoking "apply it" to a computer) to achieve the intended result in a high level of generality which cannot integrate the abstract idea into a practical application in Step2A-Prong 2 and/or provide an inventive concept at Step 2B (see MPEP 2106.05(f) and (a)(I - II).Thus, for these reasons, the Examiner respectfully disagrees, and maintains 35 USC § 101 rejection for these pending claims.
Regarding the Applicant's arguments of rejection under 35 USC § 103 for the pending claims on pages 16 – 18: Applicant’s arguments with respect to the 103 rejection with Goeldi and Batcha, alone and in combination “failing to teach” the amended limitations in claims 1, 8 and 16, which are maintained herein, are unpersuasive. Firstly, the Examiner disagrees since the combination of Goeldi and Batcha references still satisfies the amended limitations in the claims under the Broadest Reasonable Interpretation (BRI) of the claim language. The combination of Goeldi and Batcha still reasonably teaches the limitation step related in part to “utilizing a Recurrent Neural Network (RNN) to recognize patterns of social media interactions” and the calculation step for the “balance values” with the basis of a “level of support” within the social interactions. Because, Applicant's arguments fail to comply with 37 CFR 1.111(b) as they amount to general allegations that the claims define a patentable invention without specifically pointing out how the language of the claims patentably distinguishes them from these references that are maintained herein. Also, the Examiner notes that the amended steps for the “generated recurrent neural network (RNN)” used with the computer system for “analyzing” contentious statement(s) is broadly claimed and thus, this particular step is being teach by Batcha and the NN model that could be a “more complex NN model” when analyzing different data sources for “decision-based aggregation” (see C11; L20 – 28 and C14; L20 – 25; Batcha). Similarly, the step for “calculating… a plurality of balance values…based on a level of support within the plurality of social interactions”, such level of support basis is considered descriptive matter that does not hold patentable weight since is not functionally tied as to how the calculation for the balance values integrates the level of support basis. But nonetheless and under BRI, Goeldi teaches this limitation step as the example that discloses a “brand sentiment index gauge 907 tells how positively or negatively social media participants are talking about users' brands, products, or services” that is reflected in neutral, positive or negative values (directed to the calculation of balance values) and the “share of voice chart 909 indicates the percentage of social media posts referring to the users' brands in comparison with their competitors” (see ¶0082; Goeldi).
Lastly, regarding to Applicant’s allegation regarding to Goeldi not mentioning a “counter-balance” with respect to the related and non-functional descriptive limitation of “wherein the plurality of time-series is an input…and a forecast of the at least one machine learning model derived from iteratively refining the counter-balance” in the claims, this is a general and unfounded allegation that is further unpersuasive. Firstly, because there is no function being actively or positively recited for “counter-balancing”. Nonetheless, Goeldi reasonably teaches “sentiment trends” that are generated based on “using a combination of natural language processing, statistical processing, positive/negative keyword modifiers and author and site influence” (see ¶0069; Goeldi) wherein the “positive/negative keyword modifiers” are directed to such adjustment of counter-balancing the peer pressure over time, in accordance to ¶0047 in Applicant disclosure. Also, because during the “sentiment rating processing” may also “require balancing opposing keywords (e.g., both positive and negative keywords in the same post)” (directed to counter-balance), to determine “an overall sentiment rating of how positive, negative, or neutral the social media post is in relation to a brand, product or service” which is combined with the “positive/negative keyword modifiers” (see ¶0067 – 69; Goeldi). For all these reasons stated above, the Examiner respectfully disagrees, and maintains 35 USC § 103 rejection for these pending claims.
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 an abstract idea without significantly more. The analysis of this claimed invention recited in the claims begins in view of independent claim 8, the most representative claim of the independent claims set 1, 8 and 16, as follows:
At Step 1: Claim 8 falls under statutory category of an article of manufacture, while claims 1 and 15 are directed to a process and a system, respectively.
At, Step 2A Prong 1: Claim 8 (representative of claims 1 and 16) recites an abstract idea, which is defined by the following underlined elements (e.g. functional steps) while omitting any hardware components (e.g. represented as “…”):
receiving… a contentious statement from a user;
responsive to determining a topic associated with the contentious statement, detecting…a plurality of media content relating to the topic;
determining…a plurality of social interactions comprising at least one aggregation associated with the plurality of media content;
analyzing…the contentious statement…utilizing a plurality of training samples of the plurality of social interactions;
calculating…a plurality of balance values associated with the plurality of social interactions based on a level of support within the plurality of social interactions derived from the RNN recognizing one or more patterns in the plurality of social interactions;
outputting…a plurality of time-series data associated with the one or more patterns of the contentious statement, the plurality of time-series data configured to be time-stamped and utilized…to generate a metric indicating peer pressure comprising a counter-balance to the plurality of balance values based on the plurality of time-series data;
wherein the plurality of time-series data is an output…utilizing the at least one aggregation as training data and a forecast of the at least one machine learning model derived from iteratively refining the counter-balance; and
generating a visual representation of the plurality of time-series data based on the at least one aggregation wherein the at least one aggregation comprises one or more summaries of pertaining to a particular topic using key point analysis APIs derived from the er training data.
Generally, and as disclosed in the specification in ¶0020, this claimed invention “have the capacity to improve analytics of social media interactions via detecting indicators of peer pressure, but also to improve the field of computing overall by providing machine learning models and natural language processing techniques that reduce the amount of computing resources required to operate efficient networks.” However, the abstract idea(s) of a mental process is recited in claim 8 as it can be practically be performed in the human mind (See MPEP 2106.04(a)(2), subsection III). Specifically, claim 8 recites the steps of “responsive to determining a topic associated with the contentious statement, detecting…a plurality of media content relating to the topic”, “determining…a plurality of social interactions…” and “analyzing…the contentious statement…” to “calculate” balance values and “output” time series data that are further used to “generate a metric indicating peer pressure” while taking the aggregation data as “training data”, and “generate” a “visual representation of the plurality of time-series data” (e.g. steps b - h). In order for these steps to be performed, it requires observation for the detection of media content, judgment for determining social interactions and analyzing contentious statements and evaluation to calculate the balance values, output a time series data and further generate a visual representation, and judgement and/or opinion is required to generating peer pressure metrics.
Step 2A Prong 2: For independent claims 1, 8 and 16, The judicial exception(s) or abstract idea previously identified is not integrated into a practical application (see MPEP 2106.04 (d)). The claims recite the additional element(s) of computing device (from claims 1 and 8); one or more non-transitory computer-readable storage media (from claim 8) and computer system; one or more processors, one or more computer-readable memories, and program instructions (from claim 16); at least one machine learning model and a generated recurrent neural network (RNN) (from claims 1, 8 and 16). These additional elements, individually and in combination, and while considering the claims as a whole, are merely are used as a tool to perform the abstract idea (See MPEP 2106.05(f)). Specifically, all the claim limitation steps are recited as being performed by the computer. While the steps of “analyzing…the contentious statement…”, “calculating…a plurality of balance values associated with the plurality of social interactions…”, and “outputting…a plurality of time-series data associated with the one or more patterns of the contentious statement…to generate a metric indicating peer pressure” with “counter-balance to the plurality of balance values” for “generating a visual representation of the plurality of time-series data…” are recited as being performed with the computer in addition with the aid of a machine learning (ML) model and a generated recurrent neural network (RNN). Firstly, the computer is recited at a high level of generality that is being used as a tool to perform the generic computer functions for analyzing the contentious statement, calculating balance values, outputting time series data and generating peer pressure metrics as well as a visual representation of the time series data. Secondly, the step of “analyzing” the contentious statement with an RNN, the RNN is used to generally apply the abstract idea without placing any limits on how the RNN functions. Rather, the limitation only recites the outcome of analyzing contentious statements, without providing details on how that analyzing function is accomplished. Similarly, the step reciting the “machine learning model” is simply describing the provenance from which the time-series data was outputted without actively and positively reciting and discussing how the ML model achieved such output and how it trained and forecasted the respective data. Thus, these steps mentioned above are further describing and applying the abstract idea without placing any limits on how the technological components are being improved, while distinguishing in the claim language, the performing limitations from functions that generic computer components can perform.
Also, the recitation of “…via a generated recurrent neural network (RNN)…” to analyze the contentious statement, merely indicates a field of use or technological environment in which the judicial exception is performed. Because such 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).
As for the “receiving”, “outputting” and “generating a visual representation” steps in the claim, are really nothing more than links to computer implementing the use of ordinary capacity for implementing the use of ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general-purpose computer or computer components (refer to MPEP 2106.05 f (2)). Thus, in these limitation steps, the computer is used to perform an abstract idea, as discussed above in Step 2A, Prong One, such that it amounts to no more than mere instructions to apply the exception using a generic computer.
Therefore, this analysis is indicative of the fact that even when viewed in combination, the claims’ additional elements do not integrate the abstract idea or judicial exception into a practical application.
Step 2B: For independent claims 1, 8 and 16, these claims do not provide an inventive concept. The recited additional elements of the claim(s) are the following: computing device (from claims 1 and 8); one or more non-transitory computer-readable storage media (from claim 8) and computer system; one or more processors, one or more computer-readable memories, and program instructions (from claim 16); at least one machine learning model and a generated recurrent neural network (RNN) (from claims 1, 8 and 16). These additional elements are not sufficient to amount significantly more than the judicial exception or abstract idea (see MPEP 2106.05). Because, as indicated in Step 2A Prong 2, these additional element(s) claimed are merely, instructions to “apply” the abstract ideas, which cannot provide an inventive concept. Also, the recitation of a computer with the aid of a ML model generally recited to perform the claim limitations amounts to no more than mere instructions to apply the exception using a generic computer component. Thus, even when considered in combination, these additional elements represent mere instructions to implement an abstract idea or other exception on a computer, which do not provide an inventive concept at Step 2B.
For dependent claims 2-7, 9-15 and 17 - 20, these claims cover or fall under the same abstract idea of a mental process. They describe additional limitations steps of:
Claims 2-7, 9-15 and 17 - 20: further describes the abstract idea of the detection of peer pressure in electronic content method and its user’s “media content” from “electronic media source or social networking service” and their “social interactions” on a topic while evaluating their “degree of influence” and “level of support” for each individual and their “positive and negative sentiments” to “ascertain the metric indicating peer pressure”. Thus, being directed to the abstract idea group of “managing personal behavior or relationships or interactions between people” and “engaging in commercial or legal interactions” as these requires evaluation and judgement as well as mathematical calculations.
Step 2A Prong 2 and Step 2B: For dependent claims 6, 9, 14 and 17, these claims recite the additional elements: a machine learning algorithm (from claims 6 and 14); a natural language processing (NLP) engine (from claims 9 and 17); which are also recited to be merely used as a tool to perform the abstract idea to generate and update predictions. Thus, amounting to no more than mere instructions to “apply” the exception using a generic computer component (MPEP 2106.05(f) and (f)(2)). Accordingly, for the same reasons stated above, these additional element(s) claimed cannot provide an inventive concept at Step 2B.
Similarly, these elements and their limitations are “merely indicating a field of use or technological environment (e.g. analyzing social interactions in social media networks by using general computing technology such as ML models and NLP techniques, for this case) in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application” (MPEP 2106.05(h)). These additional elements are further describing that based on this algorithm and NPL engine the “social interactions” can be “utilized to ascertain the metric indicating peer pressure” and to “determine the topic” which are expressed at a high-level of generality (i.e., as a generic computing technology 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 do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claim is directed to an abstract idea.
Finally, the additional elements previously mentioned above, are nothing more than descriptive language about the elements that define the abstract idea, and these claims remain rejected under 101 as well.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Goeldi (U.S. Pub No. 20100121849 A1) in view of Batcha (WO Pub No. 2021060967 A1).
Regarding claims 1, 8 and 16:
This independent claim set is represented by claim 8
Bostick teaches:
one or more non-transitory computer-readable storage media and program instructions stored on the one or more non-transitory computer-readable storage media, the program instructions, when executed by the computing device, cause the computing device to perform a method comprising: (In ¶0047; Fig. 35 (3503, 3506 and 3510); Fig. 3: teaches “methods, apparatuses, and computer-readable media for generating a social network graph to model one or more social networks of related authors of online social media and their corresponding posts of online social media conversations relevant to subject matter of interest in a category”. Refer to ¶0110 – 111 for more details regarding the computer systems’ processor, memory and computer-readable medium.)
receiving, by the computing device, a contentious statement from a user; (In ¶0056; Fig. 2 (203, 206, 205 and 207): teaches that the system would take web content that is “retrieved and broken apart into 20 pieces, with each piece stored individually along with the user-profile information of the authors who posted the content” which is directed to a contentious statement.)
responsive to determining a topic associated with the contentious statement, detecting, by the computing device, a plurality of media content relating to the topic; (In ¶0050; Fig. 2 (211): teaches that the system can determine and provide “quantitative online social media content data” which can include “topics and keywords used by online discussion participants for the brand and the competition”. Also, the system can analyze a “social graph data” based on the collected data and “determine information about the author's social network including which authors are communicating about what topics, who is responding to which posts, what the related content is, and so on” (see ¶0058).)
calculating, by the computing device, a plurality of balance values associated with the plurality of social interactions based on a level of support within the plurality of social interactions derived from the RNN recognizing one or more patterns in the plurality of social interactions; (In ¶0067; Fig. 3 (305); Fig. 9 (909); Fig. 26: teaches a “Sentiment rating processing is then performed using sentiment rating processing module 305” on data such as “the actual text of online social media conversations to find keywords, terms or phrases to determine if a particular post refers to the particular brand, product or service of interest”. “The input to sentiment rating processing module 305 includes the actual text of the social media post…Industry-specific keywords are identified and a value or sentiment rating is assigned to each of these keywords” (e.g. directed to balance values). “Once the keywords are identified, they are processed using a number of factors including how many times the keyword appears in the social media post, the closeness and linguistic context of the keyword in relation to the brand, product or service, and whether the keyword reflects a positive, negative, or neutral sentiment about the brand, product or service”. Also, this processing also requires “balancing opposing keywords (e.g., both positive and negative keywords in the same post) to determine an overall sentiment rating of how positive, negative, or neutral the social media post is in relation to a brand, product or service” which is directed to an aggregation basis. As for the descriptive matter of the level of support within the social interactions used as a basis to calculate the balance values, this does not hold patentable weight since is not functionally tied as to how the calculation for the balance values integrates the level of support basis. However, under BRI, this prior art discloses an example wherein the “brand sentiment index gauge 907 tells how positively or negatively social media participants are talking about users' brands, products, or services” that is reflected in neutral, positive or negative values (directed to the calculation of balance values), then the “brand trend line graph 911 shows how social media participant attitudes and opinions for a user's brand, product or service have changed over time” as well as the “share of voice chart 909 indicates the percentage of social media posts referring to the users' brands in comparison with their competitors” (see ¶0082), which is interpreted as the level of support within the social interactions used as a basis to calculate the balance values, which is in accordance to the example given in ¶0047 from Applicant disclosure. Also, refer to ¶0096 for “a share of voice” category that is displayed and shows “how much conversations in the online social media are talking about this set of brands relative to each other for the date range” and “users can quickly see if their volume of mentions in online social media is high or low in comparison to the competition and can view the chart for a different month for comparison”.)
outputting, by the computing device, a plurality of time-series data associated with the contentious statement, the plurality of time-series data configured to be time-stamped and utilized by the computing device to generate a metric indicating peer pressure comprising a counter-balance to the plurality of balance values based on the plurality of time-series data; (In ¶0070; Figs. 12 and 15 – 16: teaches that “the overall volume of opinions about users' brands, products or services is calculated and trends over time can be determined based on volume trend processing” wherein “the content authored in each unit of time is searched for the terms of interest, and the number of occurrences is added up per unit of time and per term” and “when plotted in a time series, these volume data points describe the volume trend for the brand, product or service” which is directed to outputting a plurality of time-series data associated with the contentious statement. Moreover, at “operation 413, text trend processing is performed on the data” wherein the “text trend processing analyzes the text edge information stored in text edge storage 307 in conjunction with time information to determine text trends over time” and this “processing is used to determine how sentiment changes over time” and further determine “aggregation of opinion leader data over time to determine trends in opinion leader data” to “identify and target social media authors with the most influence” which is directed to the generation of a metric indicating peer pressure or users influence over one or more users, given the different or possible metric examples that can be generated in ¶0047 – 48 from Applicant disclosure. As for the time-series data being time-stamped, before being utilized to generate a metric indicating peer pressure, which is considered to be non-functional descriptive matter, this is directed to the data sources of “raw conversations of each social media post” and “raw content metadata” which includes “the length, context, and time of the post” and further include “author's username, demographic information, number of posts to the social media website, those responding to the author's posts, and the author's contacts” (see ¶0057). But also, in Fig. 12 the search data under the “overview” tab includes timestamps with date and time. Finally, for the balance values configured to counter-balance the peer pressure based on the plurality of time-series data, this is already satisfied by the prior art since the “sentiment trends” are generated based on “using a combination of natural language processing, statistical processing, positive/negative keyword modifiers and author and site influence” (see ¶0069) wherein the “positive/negative keyword modifiers” are directed to such adjustment of counter-balancing the peer pressure over time and in accordance to applicant specs in ¶0047. Refer to ¶0085 – 86 and Figs. 12 - 16 for more details regarding “trend lines display” in a GUI which includes a plurality of time-series data.)
generating a visual representation of the plurality of time-series data based on the at least one aggregation wherein the at least one aggregation comprises one or more summaries of pertaining to a particular topic using key point analysis APIs derived from the training data. (In Figs. 12 – 16: teaches a visual representation generation directed to “trend lines display” in a GUI which includes a plurality of time-series data (see ¶0084 – 86 for more details). See ¶0080 wherein the “GUI” is “utilized to present the quantified and analyzed online social media content in a manner relevant to the user” and “not only provides standard spreadsheet-style visualization such as bar and pie charts, but also highly innovative approaches including: radar screen; heatmaps; geographical visualization; 3D clustering, tag clouds, and timelines” which at least the use of “heatmaps” and “tag clouds, and timelines” can be used for “key point analysis” to give “context to each social media post and gain familiarity with the posting website”.)
Goeldi teaches the determination of social interactions via “the social network analysis (SNA) processing” to analyze “information about the author's social network including which authors are communicating about what topics, who is responding to which posts, what the related content is, and so on” which is developed on networks in social graphs (related to clusters; see ¶0058, Goeldi). However, Goeldi does not explicitly teach the abilities of determining these social interactions, but specifically based on aggregation associated to media content, to analyze the contentious statement via a generated recurrent neural network (RNN) utilizing a plurality of training samples of the plurality of social interactions and wherein the time-series data is outputted by ML model(s) utilizing an aggregation or cluster data as training data and a forecast derived from a ML model that iteratively refined the counter-balance. Thus, Batcha teaches:
determining, by the computing device, a plurality of social interactions comprising at least one aggregation associated with the plurality of media content (In C13; L1 – 12; Fig. 2 (202) Fig. 6 (606) and Fig. 7 (704): teaches that the system uses a ML tool for extracting statements from articles via “Statement Extraction Module (202)” (see C9; L1 – 4) and “article comparison on document similarity is performed using known algorithms such cosine similarity or Euclidean distance to compare the documents (606)” which is further determined if the data (e.g. articles data) is a duplicate before “summarizing similarity of articles using e.g Euclidean distance (704)” by querying an “article summary (702)” from a “database of articles” directed to the aggregation and machine learning based aggregating the data determined and is in accordance to applicant specs in ¶0030.)
analyzing, by the computing device, the contentious statement via a generated recurrent neural network (RNN) utilizing a plurality of training samples of the plurality of social interactions; (In C11; L25 – 28; Fig. 1.0c and Fig. 9 (920, 922, 924 and 926): teaches that “information from the Statement, Article update, Article positivity, historical trend, and global trend” is fed to “a neural network model” or NN model to compute and obtain a “decision based aggregation”. The NN model used for the “decision based aggregation” could be a “more complex NN model” which suggests that is could be an RNN, in order to have the model to “be updated as the system accumulates more historical data and the feedbacks received to improve accuracy over time” (see C11; L20 – 23), which is in accordance to the example given in ¶0030 from Applicant disclosure. Refer to C14; L20 – 25 wherein the NN model used, trains its weights (e.g. “trend positivity (920), article positivity weight (918), historical trend weight (922) and global trend weight (924)”) over time from historical data which is directed to utilizing a plurality of training samples.)
wherein the plurality of time-series data is an output of at least one machine learning model utilizing the at least one aggregation as training data and a forecast of the at least one machine learning model derived from iteratively refining the counter-balance; and (In C11; L25 – 29; Fig. 1.0b – 1.0c: teaches that “as illustrated in FIG. 1 .0c, Information from the Statement, Article update, Article positivity, historical trend, and global trend will be fed to a decision based aggregation to compute using a neural network model that could train its weights over time from historical data. The weights in the neural network would be able to update in a smooth manner over time for an improved accuracy” wherein such update is directed to outputting the plurality of time-series data from the ML using data aggregation/grouping as training data. This is because the Examiner notes that the claimed “at least one machine learning model utilizing the at least one aggregation as training data” seems to be suggested as equivalent to or the same as using an RNN when in view of the examples given in ¶0020, ¶0027 and ¶0029 from Applicant disclosure. As for the use and forecast of the ML model, this is directed to the “parameters retrieved from the historical trend, statement weightage, article update detection, sentiment of article” that are used in the “prediction algorithm to suggest to the user the possible next outcome” and the prediction algorithm “adjusts the weightages of the parameters and computes the aggregation” wherein this aggregation or “Decision based Aggregation” can be further used in a “more complex neural network model” that is updated for further accuracy improvement over time (see C11; L17 – 23).)
It would have been obvious to one of ordinary skill in the art before the earliest effective filing date of the claimed invention to modify Goeldi to provide the abilities of abilities of determining these social interactions, but specifically based on aggregation associated to media content, to analyze the contentious statement via a generated recurrent neural network (RNN) utilizing a plurality of training samples of the plurality of social interactions and wherein the time-series data is outputted by ML model(s) utilizing an aggregation or cluster data as training data and a forecast derived from a ML model that iteratively refined the counter-balance, as taught by Batcha in order to have valuable information that would allow “to automatically predict the outcome after processing an article considering the enormous amount of information on a perception of an entity.” (C2; L24 – 26; Batcha).
Regarding claims 2, 10 and 18:
The combination of Goeldi and Batcha, as shown in the rejection above, discloses the limitations of claims 1, 8 and 16, respectively.
Goeldi further teaches:
wherein the plurality of media content is derived from at least one electronic media source or social networking service and the plurality of social interactions includes one or more likes, comments, shares, posts, and re-posts relating to the topic. (In ¶0057: teaches “The raw post content retrieved from the online social media sources is stored in raw content storage 133 (operation 205)” and includes “the actual text of the relevant social media post”, the “raw content metadata includes information such as the URL of the social media website, and the length, context, and time of the post” and more data that includes “the social media post's author profile data such as the author's username, demographic information, number of posts to the social media website, those responding to the author's posts, and the author's contacts” which are directed to both electronic media source or social networking service and the plurality of social interactions. This data is the used to determine “information about the author's social network including which authors are communicating about what topics, who is responding to which posts, what the related content is, and so on” (see ¶0058).)
Regarding claims 3 and 12:
The combination of Goeldi and Batcha, as shown in the rejection above, discloses the limitations of claims 1 and 8, respectively.
Goeldi further teaches:
wherein calculating the plurality of balance values further comprises calculating the plurality of balance values based on a degree of influence of an individual providing a social interaction of the plurality of social interactions. (In ¶0069; Fig. 4A (402 – 404): teaches “the sentiment rating processing of operation 402 takes into consideration the level of influence the author of the social media post has in determining the sentiment rating. A weighting factor is determined based on the influence of the author of the social media post. The resulting data from sentiment rating processing module 305 is then stored in the sentiment rating storage 309 (operation 404).”)
Regarding claims 4 and 13:
The combination of Goeldi and Batcha, as shown in the rejection above, discloses the limitations of claims 1 and 8, respectively.
Goeldi further teaches:
wherein calculating the plurality of balance values further comprises calculating the plurality of balance values based on level of support within the plurality of social interactions. (In ¶0082; Fig. 9 (909); Fig. 26: teaches that “the share of voice chart 909 indicates the percentage of social media posts referring to the users' brands in comparison with their competitors” which is directed to indicating a strong wave or level of support. Also, refer to ¶0096 for “a share of voice” category that is displayed and shows “how much conversations in the online social media are talking about this set of brands relative to each other for the date range… Users can quickly see if their volume of mentions in online social media is high or low in comparison to the competition and can view the chart for a different month for comparison. Clicking on a section of the chart takes users to the newest posts about that brand.”)
Regarding claims 5, 11 and 19:
The combination of Goeldi and Batcha, as shown in the rejection above, discloses the limitations of claims 1, 8 and 16, respectively.
Goeldi further teaches:
wherein the at least one aggregation includes one or more of a first aggregation value representing positive sentiments and a second aggregation value representing negative sentiments pertaining to the contentious statement. (In ¶0068: teaches “Keywords are assigned with a positive and negative probability value each that express the probability that the keyword means something positive or negative in the context of the specific vertical. Since the same word can have different meanings per industry or topic, these probabilities can be specifically set per vertical” in which each “vertical layer” can “be generated for every conceivable category such as industry, topic of interest, type of website, geographic region, and so on” (see ¶0065) and stores “the aggregated and quantified online social media content in a database” (see ¶0054) which is directed to a first aggregation value of positive sentiments and a second aggregation value of negative sentiments related to the contentious statement.)
Regarding claims 6 and 14:
The combination of Goeldi and Batcha, as shown in the rejection above, discloses the limitations of claims 1 and 8, respectively.
Goeldi further teaches:
wherein calculating the plurality of balance values further comprises providing, by the computing device, a machine learning algorithm trained with a plurality of past social interactions, thereby having learned to derive from the plurality of social interactions the at least one aggregation configured to be utilized to ascertain the metric indicating peer pressure. (In ¶0058: teaches “The SNA processing first calculates a so-called centrality value for each author that expresses the author's degree of influence in a given social network” in which “a version of Brandes' Betweenness Centrality algorithm is applied to the raw social graph for each website. The resulting raw centrality value is then modified with the activity level of the author, i.e. the number of posts written by this person, and an importance score for the website where that author is active.” Finally, an “influence score” can be calculated for each author with its corresponding formula (see ¶0059 for more details).)
Regarding claims 7, 15 and 20:
The combination of Goeldi and Batcha, as shown in the rejection above, discloses the limitations of claims 1, 8 and 16, respectively.
Goeldi further teaches:
wherein the counter-balance is based on a time unit derived from a date and time of each social interaction of the plurality of social interactions wherein weighting parameters are applied to adjust the plurality of balance values and account for peer pressure. (In ¶0067; Figs. 9, 12, 15 – 16 and 21 – 22: teaches that the system can process “sentiment ratings” or values in the content data from “social graph information” directed to balance values. This “sentiment rating processing” includes “natural language and sentence structure analysis to determine which parts of the text of a social media post apply to the particular brand, product or service” wherein the processing may also “require balancing opposing keywords (e.g., both positive and negative keywords in the same post)” directed to “counter-balance”, to determine “an overall sentiment rating of how positive, negative, or neutral the social media post is in relation to a brand, product or service.” Thus, this information is combined with the “social graph data” and further weighted for “level of influence of the author of the social media post” (e.g. “positive/negative keyword modifiers and author and site influence”) and is aggregated over time for “sentiment trend processing” (see ¶0068 – 69) which is directed to the counter-balance based on a time unit derived from a date and time with weighting parameters applied for balance values adjustment while accounting peer pressure.)
Regarding claims 9 and 17:
The combination of Goeldi and Batcha, as shown in the rejection above, discloses the limitations of claims 8 and 16, respectively.
Goeldi further teaches:
wherein determining the topic associated with the contentious statement comprises: utilizing, by the computing device, a natural language processing (NLP) engine to determine the topic based on a social interaction of the plurality of social interactions derived from the user. (In ¶0067: teaches that the system can determine topics (e.g. concepts or subjects) via “sentiment rating processing module 305” wherein the sentiment rating processing “includes analyzing the actual text of online social media conversations to find keywords, terms or phrases to determine if a particular post refers to the particular brand, product or service of interest” and “includes natural language and sentence structure analysis to determine which parts of the text of a social media post apply to the particular brand, product or service” which is directed to a topic being determined. Refer to ¶0085 wherein a combination of NLP, “statistical processing, positive/negative keyword modifiers and author and site influences may be used to rate each post to online social media”.)
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Chang (U.S. Pub No. 20120185544 A1) is pertinent because it “provide[s] techniques that quantize community interactions with social media to understand and influence consumer experiences.”
Lehrer (U.S. Pub No. 20160026720 A1) is pertinent because it “relates to systems and methods for providing a semi-automatic research tool including the ability to create, update, and/or manage a research project and/or content related to the research project.”
Bostick (U.S. Pub No. 20170300823 A1) is pertinent because it “relates generally to the field of social media contextual relationships, and more particularly to predicting influence between users having contextual relationships.”
Mao (U.S. Patent No. 12020265 B1) is pertinent because it is about “Systems, apparatuses, and methods obtain and process data that may be used to identify or discover one or more “influencers” in business, finance, fashion, sports, current events, and other areas, and also to generate an indication of each influencer's expertise and ability to influence others with their posts or comments.”
Cantarero (U.S. Pub No. 20150227579 A1) is pertinent because it is a “method and system for deriving relevant and deeper insights from social media data through computer automation”
Cai (U.S. Patent No. 11563699 B2) is pertinent because it “generally relate to the field of natural language processing, and particularly, embodiments of the present disclosure related to systems and methods for a machine natural language processing architecture for providing virtual agents.”
Bennett (U.S. Patent No. 9177060 B1) is pertinent because it “relates to a computer implemented method, system and apparatus for identifying, collecting and parsing content for providing business intelligence. Particularly, the present invention provides a method, system and apparatus for deriving knowledge from information indicative of human communication, emotions, reactions, and experiences to evaluate trends and decisions that impact business.”
Zhao (U.S. Pub No. 20200159829 A1) is pertinent because it is “relate to using machine learning techniques to analyze text comments.”
Arpat (U.S. Pub No. 20200286000 A1) is pertinent because it “relates generally to social networking, and in particular to inferring sentiment polarity for users of a social networking system.”
Yates (U.S. Pub No. 20230306345 A1) is pertinent because it “relates to the field of artificial intelligence (AI) systems and methods of improving the performance of predictive models utilizing such machines. More specifically, the use of AI systems to monitor and analyze the impact of social media feeds on a particular business or brand.”
Kim (U.S. Pub No. 20130054638 A1) is pertinent because it “relate[s] in general to a technique of analyzing a designated topic in a social network environment, and more specifically to an apparatus and a method that can detect and track the topic based on an opinion and a social influencer for each topic.”
Milner (U.S. Pub No. 20170262451 A1) is pertinent because it is “a system and method for computer-implemented automated category-based valuation of social influence.”
Baughman (U.S. Pub No. 20170221168 A1) is pertinent because it “relates to data analytics, and more particularly to determining and ranking influence of individuals in a social network.”
Johnmar (U.S. Pub No. 20140032657 A1) is pertinent because it “relate[s] generally to information management, and more particularly, to providing analytic measurement of digital and social media content.”
Sarvabhotla (U.S. Pub No. 20140337328 A1) is pertinent because it is “relates to a method and system for capturing, extracting, analyzing, categorizing, synthesizing, summarizing and displaying, the substance and sentiment embodied within such data through a concept centric social media portal.”
Ravazzi, Learning hidden influences in large-scale dynamical social networks (July 23, 2020) is pertinent because the paper aims at “qualitatively and quantitatively infer social influence from data using a systems and control viewpoint” and to “provide a general overview of the main concepts, algorithmic tools, results, and open problems in the systematic study of learning interpersonal influence in networked systems.”
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 Ivonnemary Rivera Gonzalez whose telephone number is (571)272-6158. The examiner can normally be reached Mon - Fri 9:00AM - 5:30PM.
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, Nathan Uber can be reached at (571) 270-3923. 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.
/IVONNEMARY RIVERA GONZALEZ/Examiner, Art Unit 3626
/NATHAN C UBER/Supervisory Patent Examiner, Art Unit 3626
Prosecution Insights