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 .
Claim(s) 1-20 are pending for examination. Claim(s) 1-3, 7, 9-11, and 19-20 have been amended. This action is Final.
Response to Arguments
The claim objection to claim 1 is withdrawn as claim 1 has been amended.
Applicant’s arguments, filed 3/4/2026, with respect to the 35 U.S.C. 101 rejection have been fully considered and are persuasive. The 35 U.S.C. 101 of claim(s) 1-20 has been withdrawn. The examiner agrees the claims as amended provides a specific technical solution to specifical technical problem, see Applicant’s Remarks filed on 3/4/2026, p. 19 of 24.
Applicant’s arguments, filed 3/4/2026, with respect to the 35 U.S.C. 103 rejection have been considered and are moot in view of new grounds of rejection.
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.
Claim(s) 1, 3-8, 10, 12-17, and 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yuan et al. (US 2020/0005117 A1) in view of Beaver (US 2020/0320134 A1).
Regarding Claim 1;
Yuan discloses an apparatus comprising one or more processors and one or more storage devices storing instructions that are operable, when executed by the one or more processors, to cause the one or more processors to:
generate an information technology (IT) support label dataset based on service event data structures related to respective services messages provided to an application framework configured to manage respective application components for IT service management ([0025] and [0040] - Also in an example, an entity may be a keyword or other tracked value that impacts the flow of the conversation. For example, if an end user intent is, “printer is not working”, a virtual agent may ask for a printer model and operating system to receive example replies such as “S7135” and “Windows”. In this scenario, “printer”, “S7135” and “Windows” are entities. As an example, an intent may represent the categorization of users' questions, issues, or things to do. For example, an intent may be in the form of, “Windows 10 upgrade issue”, “How do I update my credit card?”, or the like. As an example, a solution may include or define a concrete description to answer or solve a users' question or issue. For example, “To upgrade to Windows 10, please follow the following steps: 1) backup your data, . . . 2) Download the installer, . . . , 3) Provide installation information, . . . ”, etc. [0047]-[0049] - In operation 310, operations are performed to obtain and label an initial set of chat transcript content. For example, a sample of conversation data (e.g., a set of thousands of conversations, selected from millions of conversation statements) may be evaluated and labeled, such as by human-initiated (manual) labeling. This labeling may identify statements or portions of statements with labels that indicate respective questions, answers, followup questions, followup answers, issues, or the like. Then, in operation 320, operations are performed to train a machine learning model using the sample of labeled conversation data, which provides structured content for training and classification);
generate a training dataset associated... based on [a] set of candidate questions set ([0025] and [0040] and [0047] and [0048]-[0049] - In operation 310, operations are performed to obtain and label an initial set of chat transcript content. For example, a sample of conversation data (e.g., a set of thousands of conversations, selected from millions of conversation statements) may be evaluated and labeled, such as by human-initiated (manual) labeling. This labeling may identify statements or portions of statements with labels that indicate respective questions, answers, followup questions, followup answers, issues, or the like. Then, in operation 320, operations are performed to train a machine learning model using the sample of labeled conversation data, which provides structured content for training and classification and [0052] and [0066] - In a specific example, the machine learning model is trained from a set of structured learning data, with such data including various conversation content (e.g., utterance) types labeled as: a problem, a clarification question, a clarification answer, or a solution. Also in a specific example, the machine learning model is a CRF classifier, such that the CRF classifier is trained to classify the conversation content type (such as a respective type of utterance).);
... train an artificial intelligence (AI) model based on (i) the IT support label dataset and (ii) the training dataset to generate a trained an intent recognition Al model configured for predicting intent associated with service messages ([0025] - : identifying content for a particular support issue (an “intent”); and [0040] and [0048]-[0049] - In operation 310, operations are performed to obtain and label an initial set of chat transcript content. For example, a sample of conversation data (e.g., a set of thousands of conversations, selected from millions of conversation statements) may be evaluated and labeled, such as by human-initiated (manual) labeling. This labeling may identify statements or portions of statements with labels that indicate respective questions, answers, followup questions, followup answers, issues, or the like. Then, in operation 320, operations are performed to train a machine learning model using the sample of labeled conversation data, which provides structured content for training and classification [0052] and [0066]); and
configure an intent recognition engine for a virtual agent system based on the intent recognition Al model ([0025] - The present authoring techniques include the use of knowledge mining workflows, and the organization of knowledge graph and intent data structures, which are suitable for consumption by a virtual agent in a knowledge information service. For example, in the context of a technical support virtual agent, the present AI-assisted content authoring techniques may involve: identifying content for a particular support issue (an “intent”); developing an intent list to identify solutions for multiple types of intents; and identifying and approving suitable questions and answers to use in an interaction and [0051]).
Yuan fails to explicitly disclose
generate, using a candidate question classification model, a candidate questions set based on one or more patterns associated with the service event data structures;
generate a training dataset associated with a for an artificial intelligence (AI) model based on the set of candidate questions set provided by a classification model;
modify one or more parameters of the Al model train an artificial intelligence (AI) model ... to generate a trained ... Al model ....
However, in an analogous art, Beaver teaches:
generate, using a candidate question classification model, a candidate questions set based on one or more patterns associated with the service event data structures ([0031] and [0034] - and [0035]-[0037] - After the models are trained, the question module 230 may use the question model 223 to generate a set of candidate questions 245 from the data 211. The candidate questions 245 may be meant to simulate the questions 245 that are likely to be asked by customers of the company or entity during conversations with the IVA. Depending on the embodiment, the generated questions 245 may be provided to one or more human reviewers who may review the generated questions 245 based on a variety of factors such as grammar and relevance. The reviewers may edit one or more of the questions 245, delete one or more of the questions 245, or may add one or more additional questions 245. After generating the questions 245, the question module 230 may associate each question 245 with an intent.... After generating the responses 243, the response module 240 may have one or more human reviewers review the proposed responses 243 for each question 245);
generate a training dataset associated with a for an artificial intelligence (AI) model based on the set of candidate questions set provided by a classification model ([0031] and [0034] - The response model 221 may be a model that is trained to generate one or more responses 243 based on an input question 245 and an intent and [0035]-[0037]);
modify one or more parameters of the Al model train an artificial intelligence (AI) model ... to generate a trained an intent recognition Al model configured for predicting intent ([0031] and [0034]-[0037] - The question and response engine 205 may generate a plurality of responses 243 that may be used by the IVA engine 135 to respond to received questions 245 according to determined intents; and
[configure an intent recognition engine for a virtual agent system based on the intent recognition Al model] ([0005] and [0031]).
Therefore, it would have been obvious to one of ordinarily skill in the art before the effective filing date of the claimed invention to combine the teachings of Beaver to the candidate questions of Yuan to include generate, using a candidate question classification model, a candidate questions set based on one or more patterns associated with the service event data structures; generate a training dataset associated with a for an artificial intelligence (AI) model based on the set of candidate questions set provided by a classification model; modify one or more parameters of the Al model train an artificial intelligence (AI) model ... to generate a trained an intent recognition Al model configured for predicting intent; and [configure an intent recognition engine for a virtual agent system based on the intent recognition Al model]
One would have been motivated to combine the teachings of Beaver to Yuan to do so as it provides / allows to ensure the IVA is responding appropriately (Beaver, [0002]-[0003]).
Regarding Claim 3;
Yuan in view of Beaver disclose the apparatus to claim 1.
Yaun further discloses wherein the intent precogitation Al model comprises a deep learning model ([0025] - The present authoring techniques include the use of knowledge mining workflows, and the organization of knowledge graph and intent data structures, which are suitable for consumption by a virtual agent in a knowledge information service. For example, in the context of a technical support virtual agent, the present AI-assisted content authoring techniques may involve: identifying content for a particular support issue (an “intent”); developing an intent list to identify solutions for multiple types of intents; and identifying and approving suitable questions and answers to use in an interaction and [0040] - Also in an example, an entity may be a keyword or other tracked value that impacts the flow of the conversation. For example, if an end user intent is, “printer is not working”, a virtual agent may ask for a printer model and operating system to receive example replies such as “S7135” and “Windows” and [0041] - The various responses received in the conversation of the online processing may also be used as part of a telemetry pipeline 146, which provides a deep learning reinforcement 148 of the responses and response outcomes in the conversation model 176.)
Regarding Claim 4;
Yuan in view of Beaver disclose the apparatus to claim 1.
Yaun further discloses wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: perform an intent discovery process related to the service event data structures to generate an IT support intent classification dataset ([0034] and [0047] - For instance, iterative knowledge mining may be used to perform intent discovery in a workflow after chat transcript data is labeled (with human and machine efforts) into structured data... This process will repeat until the quality of intent discovery is sufficient. Accordingly, the operational deployment 200 may utilize automated and AI techniques to assist human editors to perform tasks and work and to make decisions, within a variety of authoring and content management aspects and [0053]) and generate the IT support label dataset based on the IT support intent classification dataset. ([0034] and [0047] - In an example, the operational deployment 200 may include multiple rounds of iterative knowledge mining, editing, and learning processing. For instance, iterative knowledge mining may be used to perform intent discovery in a workflow after chat transcript data is labeled (with human and machine efforts) into structured data and [0052]).
Regarding Claim 5;
Yuan in view of Beaver disclose the apparatus to claim 1.
Yuan further discloses wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: perform one or more clustering techniques with respect to data included in the service event data structures to generate the IT support intent classification dataset ([0047]- For instance, iterative knowledge mining may be used to perform intent discovery in a workflow after chat transcript data is labeled (with human and machine efforts) into structured data. This workflow may first involve use of a machine to automatically group phrases labeled in a “problem” category, extract candidate phrases, and ultimately recommend intents. Human editors can then review the grouping results, make changes to the phrase/intent relationship, and change intent names or content based on machine recommendation results).
Regarding Claim 8;
Yuan in view of Beaver disclose the apparatus to claim 1.
Yuan further discloses wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: parse service management requests related to one or more application programming interface (API) calls into the respective services messages (FIG. 7 and FIG. 8 and [0030]-[0031] - The bot framework 116 may also enable conversations to occur through information services and user interfaces exposed by search engines, operating systems, software applications, webpages, and the like and [0034] and [0079])
Regarding Claim(s) 10 and 12-17; claim(s) 10 and 12-17 is/are directed to a/an method associated with the apparatus claimed in claim(s) 1 and 3-8. Claim(s) 10 and 12-17 is/are similar in scope to claim(s) 1 and 3-8, and is/are therefore rejected under similar rationale.
Regarding Claim(s) 19; claim(s) 19 is/are directed to a/an medium associated with the apparatus claimed in claim(s) 1. Claim(s) 19 is/are similar in scope to claim(s) 1, and is/are therefore rejected under similar rationale.
Claim(s) 2 and 11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yuan et al. (US 2020/0005117 A1) in view of Beaver (US 2020/0320134 A1) and further in view of Zhang et al. (US 2022/0035527 A1).
Regarding Claim 2;
Yuan in view of Beaver disclose the apparatus to claim 1.
Beaver further teaches the candidate question classification model ([0031] and [0034] - and [0035]-[0037]).
Similar rationale and motivation is noted for the combination of Bever to Yuan in view of Beaver, as per claim 1, above.
Yuan in view of Beaver fail to explicitly disclose wherein the classification model is configured as a term frequency-inverse document
However, in an analogous art, Zhang teaches wherein the classification model is configured as a term frequency-inverse document ([0032] - In some embodiments, computing device 104 collects operation commands of the target type and operation commands of other types to train a classification model, so that the classification model can classify input commands. For example, a Latent Semantic Indexing (LSI) model, a Doc2vec model, a Term Frequency-inverse Document Frequency (TF-IDF) model, and the like may be trained to obtain a classification model).
Therefore, it would have been obvious to one of ordinarily skill in the art before the effective filing date of the claimed invention to combine the teachings of Zhang to the classification model of Yuan in view of Beaver to include wherein the classification model is configured as a term frequency-inverse document
One would have been motivated to combine the teachings of Zhang to Yuan in view of Beaver to do so as it provides / allows [data that] can be quickly distinguished, thereby reducing the time for processing commands for acquiring data, reducing the bandwidth consumed, and improving the processing efficiency of operation commands (Zhang, [0026]).
Regarding Claim(s) 11; claim(s) 11 is/are directed to a/an method associated with the apparatus claimed in claim(s) 2. Claim(s) 11 is/are similar in scope to claim(s) 2, and is/are therefore rejected under similar rationale.
Claim(s) 6 and 7 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yuan et al. (US 2020/0005117 A1) in view of Beaver (US 2020/0320134 A1) and further in view of Manjunatha et al. (US 2022/0109672 A1)
Regarding Claim 6;
Yuan in view of Beaver disclose the apparatus to claim 1.
Yuan further discloses encode... service event data structures .... and generate the IT support intent classification dataset .... ([0025] and [0047]-[0049] and [0053]).
Yuan in view of Beaver fail to explicitly disclose wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: encode text ... into embedding vectors; and generate ... classification dataset based on the embedding vectors
Manjunatha further teaches wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: encode text ... into embedding vectors ([0060] - As depicted in the diagram, the unclassified client device 170A (also described herein as the unknown client device 170) can input the private context information 175 and the private device information into the obfuscation model 185 to generate an output vector 210 and [0062]); and generate ... classification dataset based on the embedding vectors ([0062] - The data processing system 105 (or one or more of the components thereof, such as the device classifier 150, etc.) can receive the request for classification and extract the classification request metadata 305 and the output vector 210. The data processing system 105 can generate an input vector to the classification model 205. The classification model 205 can use the input vector as an input, and can generate an unknown device classification 310. The data processing system 105 can generate a device classification message that includes the unknown device classification 310 and content selected based on the unknown device classification 310, and can provide the classification message to the unclassified client device 170A.).
Therefore, it would have been obvious to one of ordinarily skill in the art before the effective filing date of the claimed invention to combine the teachings of Manjunatha to the training of Yuan and Beaver to include wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: encode text ... into embedding vectors; and generate ... classification dataset based on the embedding vectors
One would have been motivated to combine the teachings of Manjunatha to Yuan and Beaver to do so as it provides / allows improve the network utilization of device classification systems (Manjunatha, [0003]).
Regarding Claim 7;
Yuan in view of Beaver disclose the apparatus to claim 1.
Yuan further discloses intent recognition AI model and a classification provided by the intent recognition engine ([0049] and [0053] - The intent discovery process 420 uses a classification technique, such as with a machine learning model, to produce a set of candidate intents 430.)
Yuan in view of Beaver fail to explicitly disclose wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: retrain the ... AI model based on an active learning...
Manjunatha further teaches wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: retrain the ... AI model based on an active learning... ([0071] - The data processing system can apply the output vector and metadata as input to the classification model to receive a classification output. The data processing system can use the classification output as feedback to train the classification models and the obfuscation models. By training the models stored in memory, the data processing system can facilitate device classification of unknown (e.g., unclassified, etc.) client devices based on the classification of known client devices. Training the obfuscation and classification models based on known classifications can include adjusting the weights, biases, or parameters of the model to facilitate relevant device classification.).
Therefore, it would have been obvious to one of ordinarily skill in the art before the effective filing date of the claimed invention to combine the teachings of Manjunatha to the training of Yuan and Beaver to include wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: retrain the ... AI model based on an active learning...
One would have been motivated to combine the teachings of Manjunatha to Yuan and Beaver to do so as it provides / allows improve the
Claim(s) 9, 18, and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yuan et al. (US 2020/0005117 A1) in view of Beaver (US 2020/0320134 A1) and further in view of Burg et al. (US 2022/0067589 A1)
Regarding Claim 9;
Yuan in view of Beaver disclose the apparatus to claim 1.
Yuan further teaches the intent recognition AI model for the intent recognition engine associated with the virtual agent system ([0025] and [0040] and [0048]-[0049] - In operation 310, operations are performed to obtain and label an initial set of chat transcript content. For example, a sample of conversation data (e.g., a set of thousands of conversations, selected from millions of conversation statements) may be evaluated and labeled, such as by human-initiated (manual) labeling. This labeling may identify statements or portions of statements with labels that indicate respective questions, answers, followup questions, followup answers, issues, or the like. Then, in operation 320, operations are performed to train a machine learning model using the sample of labeled conversation data, which provides structured content for training and classification [0052] and [0066]);
Yuan in view of Beaver fail to explicitly disclose wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: generate a first version of the AI model based on a first metrics indicator related to training of the AI model; generate a second version of the AI model based on a second metrics indicator related to the training of the AI model; and select the first version of the AI model or the second version of the AI model as the trained AI model for the intent recognition engine associated with the virtual agent system.
However, in an analogous art, Burg teaches wherein the one or more storage devices store instructions that are operable, when executed by the one or more processors, to further cause the one or more processors to: generate a first version of the AI model based on a first metrics indicator related to training of the AI model ([0040] - The code portion 31 controls the application including receipt of input data and causing the input data to be processed by the two machine learning models. When the application 26 is run, a set of input data is processed by each of model A 32 and model B 33. Each of model A 32 and model B 33 generates output values in the form of node activations and, in the embodiment shown in FIG. 3, the node activations are compiled into A/B test metrics 34); generate a second version of the AI model based on a second metrics indicator related to the training of the AI model ([0040] - The code portion 31 controls the application including receipt of input data and causing the input data to be processed by the two machine learning models. When the application 26 is run, a set of input data is processed by each of model A 32 and model B 33. Each of model A 32 and model B 33 generates output values in the form of node activations and, in the embodiment shown in FIG. 3, the node activations are compiled into A/B test metrics 34); and select the first version of the AI model or the second version of the AI model as the ... model ([0041] A/B testing as performed by the application 26 is a method of testing two different machine learning models, in this case machine learning model A 32 and machine learning model B 33, using the same input data. By collecting statistics on the performance of the two machines learning models, the performance of the machine learning models can be evaluated against each other. This allows the better performing machine learning model to be selected for use in further inference processing).
Therefore, it would have been obvious to one of ordinarily skill in the art before the effective filing date of the claimed invention to combine the teachings of Burg to trained AI model of Yuan in view of Beaver to include wherein the classification model is configured as a term frequency-inverse document
One would have been motivated to combine the teachings of Burg to Yuan in view of Beaver to do so as it provides / allows the better performing machine learning model to be selected for use in further ... processing (Burg, [0041]).
Regarding Claim(s) 18; claim(s) 18 is/are directed to a/an method associated with the apparatus claimed in claim(s) 9. Claim(s) 18 is/are similar in scope to claim(s) 9, and is/are therefore rejected under similar rationale.
Regarding Claim(s) 20; claim(s) 20 is/are directed to a/an medium associated with the apparatus claimed in claim(s) 9. Claim(s) 20 is/are similar in scope to claim(s) 9, and is/are therefore rejected under similar rationale.
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 ASFAND M SHEIKH whose telephone number is (571)272-1466. The examiner can normally be reached Mon-Fri: 7a-3p (MDT).
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, JESSICA LEMIEUX can be reached at (571)270-3445. 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.
/ASFAND M SHEIKH/Primary Examiner, Art Unit 3626
Prosecution Insights