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 .
Allowable Subject Matter
Claims 9 and 12 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Regarding claim 9, the claim is indicated as allowable because the cited references of Chopra and Mourya fail to disclose the language of performing the non-linear mapping processing using probability values of a convolution feature, conditional to a length of the input text meeting a first and second preset length thresholds. The language of claim 9 would not be obvious in view of the cited references. Therefore, the claim language is indicated as allowable and would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Regarding claim 12, the claim is indicated as allowable because the cited references of Chopra and Mourya fail to disclose the language of modifying the training dataset used to train the classifier based on the received category label not matching a prediction category obtained by generating a classification result. The language of claim 12 would not be obvious in view of the cited references. Therefore, the claim language is indicated as allowable and would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation is: “processing circuitry configured to…” in claims 13-20.
Because this claim limitation is being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it is being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have this limitation interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation recites sufficient structure to perform the claimed function so as to avoid it being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-8, 10, 11, and 13-20 are rejected under 35 U.S.C. 103 as being unpatentable over Chopra (US PG Pub 20230112369) in view of Mourya (US PG Pub 20230351257).
As per claims 1 and 13, Chopra discloses: A method and an apparatus, comprising: processing circuitry (Chopra; p. 0035 - The server computing device 106 is a device including specialized hardware and/or software modules that execute on a processor and interact with memory modules of the server computing device 106, to receive data from other components of the system 100, transmit data to other components of the system 100, and perform functions for automated analysis of customer interaction text to generate customer intent information and a hierarchy of customer issues as described herein…; see also p. 0085) configured to: receive an input text and scenario information corresponding to the input text (Chopra; Fig. 2, item 202; p. 0006-0007- The server computing device captures a plurality of computer text segments, each segment including (i) a first portion comprising a transcript of an interaction between a customer and an agent (input text) and (ii) a second portion comprising notes about the interaction between the customer and the agent (scenario information)…; see also p. 0038); perform semantic enhancement processing on the input text based on the scenario information to obtain a semantic enhancement result (Chopra; Fig. 2, item 206; p. 0006-0007 - The server computing device executes the trained neural network using the interaction embeddings to generate an interaction summary for each computer text segment (semantic enhancement), the interaction summary comprising a text string identifying a primary topic of the first portion of the computer text segment; see also p. 0037 & p. 0040); encode the semantic enhancement result to obtain a text encoding (Chopra; Fig. 2, item 208; p. 0006-0007- The server computing device converts each interaction summary into a multidimensional vector representing the interaction summary (encoding)); and apply non-linear mapping processing on the text encoding to obtain a text classification result of the input text (Chopra; Fig. 2, items 210 & 212; p. 0006-0007- The server computing device aggregates the multidimensional vectors into one or more clusters based upon a similarity measure between the respective multidimensional vectors. The server computing device aligns the one or more clusters of vectors with attributes of the interaction summaries to generate a hierarchical mapping of customer issues (text classification result); see also Fig. 6, item 616 & p. 0081). Although Chopra, in p. 0006-0007 & p. 0081, discloses hierarchical mapping of multidimensional vectors. Chopra fails to explicitly disclose that the applied mapping processing on the text encoding to obtain a text classification result of the input text is non-linear. Mourya does teach applying non-linear mapping processing on the text encoding to obtain a text classification result of the input text (Mourya; p. 0114-0119 – The disclosure presents a plurality of machine learning classification models, some of which may be non-linear, such as Support Vector Machine (SVM): SVM is a supervised machine learning algorithm that may be used for classification tasks. In the fallback classification process, an SVM model may be trained using the existing intent training phrases as feature vectors and their corresponding intent labels as target variables. SVM aims to find a hyperplane that best separates the training data into different intent classes. By utilizing a kernel function, the SVM model may handle non-linear decision boundaries, allowing it to classify fallback utterances into the closest matching existing intent). Therefore, it would have been obvious to one of ordinary skill in the art to modify the method and apparatus of Chopra to include applying non-linear mapping processing on the text encoding to obtain a text classification result of the input text, as taught by Mourya, because by employing classification models such as logistic regression, decision tree, random forest, SVM, or AutoML, the system may perform the classification of fallback utterances into existing intent classes. It is to be noted that each model has its own capability to learns the patterns and relationships between training phrases and intent labels, enabling accurate prediction and identification of the most appropriate existing intent for a given fallback utterance (Mourya; p. 0120).
As per claims 2 and 14, Chopra in view of Mourya disclose:
The method and apparatus according to claims 1 and 13, wherein the scenario information indicates a purpose of a conversation between a plurality of speakers from which the input text is generated (Chopra; p. 0003 - some agents enter a brief summary or notes of the specific interaction, noting the reasons behind the customer interaction, such as issue(s) faced by customers or product interests of customers).
As per claims 3 and 15, Chopra in view of Mourya disclose: The method and apparatus according to claims 1 and 13, wherein the processing circuitry is configured to: perform vectorization processing on the input text to obtain a vectorization result of the input text (Chopra; p. 0008 - generating an interaction embedding comprises tokenizing (vectorization), by the server computing device, the computer text segment into a plurality of tokens); and perform the semantic enhancement processing on the vectorization result based on which of a plurality of semantic enhancements is determined to correspond to the scenario information to obtain the semantic enhancement result, each of the plurality of semantic enhancements being associated with a different type of scenario information (Chopra; p. 0054-0058 - the interaction embedding module 110 can generate specific word embeddings for each of the intent model 108a and short summary model 108b after evaluating the number of tokens that are contained in the CSR notes…Once the model 108b is trained, the server computing device 106 can subsequently execute (312b) the trained model 108b using as input interaction embeddings generated from new interactions to automatically generate a short summary for the interaction transcript and CSR notes to which new interactions correspond… The intent and generated summary which correspond to the “semantic enhancement” are trained using the CSR notes, which correspond to the “scenario information”, such that the semantic enhancements are determined to correspond to the scenario information).
As per claims 4 and 16, Chopra in view of Mourya disclose: The method and apparatus according to claims 3 and 15, wherein the input text includes speech text of a plurality of speakers (Chopra; p. 0006-0007- …a first portion comprising a transcript of an interaction between a customer and an agent (plurality of speakers)…; see also p. 0038), and the processing circuitry is configured to: when pre-configured information of the scenario information indicates the input text includes key text, set a sentence identifier corresponding to the key text in the vectorization result with a first type of identifier to obtain a first semantic enhancement result; and when pre-configuration information of the scenario information indicates the input text does not include the key text, set the sentence identifier corresponding to the speech text in the input text with a second type of identifier to obtain a second semantic enhancement result, wherein the plurality of speakers are associated with different values of the second type of identifier (Chopra; p. 0052-0054 – this part of the disclosure teaches the tokenization of text using one-hot encoders (identifiers), such that it teaches the identification of key text (e.g. the disclosed “activate, phone, access, device, setup”), and assigns an identifier, such as the one-hot vector for each key text to generate the tokens. Additionally, p. 0051 describes the identification of notes-specific words and phrases which correspond to “pre-configuration information” used to set the identifiers and generate the tokens; see also p. 0072 - The NER model can be configured to recognize specific named entities—in the context of a financial services organization, some named entities may include ‘401(k),’ ‘IRA,’ ‘Insurance,’ ‘Brokerage,’ and other types of products or services. The NER model further can be configured to identify small variations in entity names, alternate entity names, etc. and tag the intents or short summaries using one common name. For example, the clustering module 112 can execute the NER model against the following intents: “opening 403(b) plan” and “opening tax sheltered annuities plan.” The NER model identifies the entity ‘403(b)’ in the first intent, and ‘tax sheltered annuities’ in the second intent. The NER model can tag each of these intents with the tag ‘403(b)’ as a common tag covering both entities (set a sentence identifier). FIG. 11 is a diagram showing exemplary entities 1104 detected by the system 100 from the cleaned intents, using an NER model. As shown in FIG. 11, the module 112 detects the entity ‘electronic fund transfer’ from the first four intents in 1102, the module detects 112 the entities ‘ekit for electronic fund transfer,’ and ‘electronic fund transfer’ from the next five intents, the module 112 detects the entity ‘ESPP’ in the next four intents, and the module 112 detects the entity ‘wire transfer’ in the last three intents; see also p. 0066).
As per claims 5 and 17, Chopra in view of Mourya disclose: The method and apparatus according to claims 4 and 16, wherein the processing circuitry is configured to: add identity prompt words of the plurality of speakers at endpoint positions of the speech text in the input text; add separator symbols between adjacent speech texts of any two speaking objects; and based on the input text with the added identity prompt words and the added separator symbols, perform the semantic enhancement processing on the input text based on the scenario information (Chopra; p. 0048 - Sort utterances in chronological order and combine all messages from an interaction: the module 110 can analyze the text data to confirm that the utterances made by both participants during the interaction are sorted according to chronological order by, e.g., using a timestamp allocated to specific utterances/messages in the interaction data, and further that all messages from a given interaction are included in the transcript. For example, during a text chat session, the system can assign a session identifier to all of the messages exchanged between the participants. In addition, in some embodiments the system can capture a customer identifier and/or an agent identifier in connection with the interaction. Then, the module 110 can use the session identifier, customer identifier, agent identifier and/or timestamp to select the messages allocated to the user interaction and ensure that all of the messages are contained in the transcript. This step can be performed on both voice call and text chat data).
As per claims 6 and 18, Chopra in view of Mourya disclose: The method and apparatus according to claims 1 and 13, wherein the processing circuitry is configured to: perform embedded encoding on the semantic enhancement result to obtain an embedded encoding result (Chopra; p. 0066 - the clustering module 112 performs the embedding creation step (606) to generate word/phrase embeddings (embedded encoding) based upon the cleaned intents and/or short summaries (semantic enhancement)); perform attention encoding processing on the embedded encoding result to obtain an attention encoding result (Chopra; p. 0066 – discloses the use of pretrained language models for clustering such as BERT and RoBERTa, all of which employ self-attention encoding mechanisms in the generation of embeddings); and perform the non-linear mapping processing on the attention encoding result to obtain the text encoding (Chopra; Fig. 2, items 210 & 212; p. 0006-0007- The server computing device aggregates the multidimensional vectors into one or more clusters based upon a similarity measure between the respective multidimensional vectors. The server computing device aligns the one or more clusters of vectors with attributes of the interaction summaries to generate a hierarchical mapping of customer issues (text classification result); see also Fig. 6, item 616 & p. 0081). Although Chopra, in p. 0006-0007 & p. 0081, discloses hierarchical mapping of multidimensional vectors. Chopra fails to explicitly disclose that the applied mapping processing on the text encoding to obtain a text classification result of the input text is non-linear. Mourya does teach performing the non-linear mapping processing on the attention encoding result to obtain the text encoding (Mourya; p. 0114-0119 – The disclosure presents a plurality of machine learning classification models, some of which may be non-linear, such as Support Vector Machine (SVM): SVM is a supervised machine learning algorithm that may be used for classification tasks. In the fallback classification process, an SVM model may be trained using the existing intent training phrases as feature vectors and their corresponding intent labels as target variables. SVM aims to find a hyperplane that best separates the training data into different intent classes. By utilizing a kernel function, the SVM model may handle non-linear decision boundaries, allowing it to classify fallback utterances into the closest matching existing intent). Therefore, it would have been obvious to one of ordinary skill in the art to modify the method and apparatus of Chopra to include performing the non-linear mapping processing on the attention encoding result to obtain the text encoding, as taught by Mourya, because by employing classification models such as logistic regression, decision tree, random forest, SVM, or AutoML, the system may perform the classification of fallback utterances into existing intent classes. It is to be noted that each model has its own capability to learns the patterns and relationships between training phrases and intent labels, enabling accurate prediction and identification of the most appropriate existing intent for a given fallback utterance (Mourya; p. 0120).
As per claims 7 and 19, Chopra in view of Mourya discloses: The method and apparatus according to claims 1 and 13, wherein the processing circuitry is configured to: add identity prompt words of a plurality of speakers at endpoint positions of speech text in the input text; when pre-configuration information of the scenario information indicates the input text includes key text, and the key text includes the speech text of multiple speakers, add marker symbols between adjacent speech texts of any two speaker, and use the input text with the added identity prompt words and the added marker symbols as a third semantic enhancement result; and when the pre-configuration information of the scenario information indicates the input text does not include the key text, add the marker symbols between the adjacent speech texts of any two speakers, and use the input text with the added identity prompt words and the added marker symbols as a fourth semantic enhancement result (Chopra; p. 0048 - Sort utterances in chronological order and combine all messages from an interaction: the module 110 can analyze the text data to confirm that the utterances made by both participants during the interaction are sorted according to chronological order by, e.g., using a timestamp allocated to specific utterances/messages in the interaction data, and further that all messages from a given interaction are included in the transcript. For example, during a text chat session, the system can assign a session identifier to all of the messages exchanged between the participants. In addition, in some embodiments the system can capture a customer identifier and/or an agent identifier in connection with the interaction. Then, the module 110 can use the session identifier, customer identifier, agent identifier and/or timestamp to select the messages allocated to the user interaction and ensure that all of the messages are contained in the transcript. This step can be performed on both voice call and text chat data).
As per claims 8 and 20, Chopra in view of Mourya discloses: The method and apparatus according to claims 1 and 13, wherein the processing circuitry is configured to: detect a length of the input text; when the length of the input text is greater than a first preset length threshold, perform feature extraction processing on the input text to obtain statistical features, perform logistic regression processing on the statistical features to obtain the text classification result; and when the length of the input text is less than or equal to the first preset first length threshold, perform the semantic enhancement processing on the input text (Chopra; p. 0055 - In some embodiments, the interaction embedding module 110 can generate specific word embeddings for each of the intent model 108a and short summary model 108b after evaluating the number of tokens that are contained in the CSR notes. For example, for a computer text segment where the number of tokens in the CSR notes is greater than a predetermined threshold (e.g., seven tokens), the module 110 can generate word embeddings from the computer text segment that are only used for training and execution of the short summary model 108b. Likewise, for a computer text segment where the number of tokens in the CSR notes is at or below the predetermined threshold, the module 110 can generate word embeddings from the computer text segment that are only used for training and execution of the intent model 108a (performing the semantic enhancement using intent model upon determining the number of tokens is at or below the threshold)… In other words, the disclosure of Chopra teaches the determination of a length of the input text and the comparison of the length to a threshold for generating words embeddings).
Chopra, however, fails to disclose perform feature extraction processing on the input text to obtain statistical features, perform logistic regression processing on the statistical features to obtain the text classification result. Mourya does teach perform feature extraction processing on the input text to obtain statistical features, perform logistic regression processing on the statistical features to obtain the text classification result (Mourya; p. 0114-0115 - The classification may be performed by at least one of the following machine learning classification models: logistic regression… Logistic Regression: Logistic regression is a statistical model used to estimate the probability of a binary outcome. In the context of fallback classification, a logistic regression model may be trained using existing intent training phrases as features and their corresponding intent labels as target variables. The model learns the relationship between the training phrases and intents, enabling it to predict the intent class for a given fallback utterance. By analyzing the probabilities assigned to each intent class, the model determines the closest matching existing intent for the fallback).
Therefore, it would have been obvious to one of ordinary skill in the art to modify the method and apparatus of Chopra to include performing the non-linear mapping processing on the attention encoding result to obtain the text encoding, as taught by Mourya, because by employing classification models such as logistic regression, decision tree, random forest, SVM, or AutoML, the system may perform the classification of fallback utterances into existing intent classes. It is to be noted that each model has its own capability to learns the patterns and relationships between training phrases and intent labels, enabling accurate prediction and identification of the most appropriate existing intent for a given fallback utterance (Mourya; p. 0120).
As per claim 10, Chopra in view of Mourya disclose: The method according to claim 1, wherein the input text includes speech text from a plurality of speakers (Chopra; p. 0006-0007- …a first portion comprising a transcript of an interaction between a customer and an agent (plurality of speakers)…; see also p. 0038), and the method further comprises: obtaining speech from the plurality of speakers; performing text conversion processing on the speech to obtain converted text; and generating the input text based on the converted text (Chopra; p. 0030 - the database 102a can store a digital audio recording of the voice call and/or a transcript of the voice call (e.g. as generated by a speech-to-text module that converts the digital audio recording into unstructured text); see also p. 0039 - when only the digital audio recording is stored (e.g. as a .mp4 file), the system 100 can convert the audio into unstructured text using, e.g., a speech-to-text conversion module that analyzes the waveforms in the audio file and converts them into natural language text; see also p. 0041).
As per claim 11, Chopra discloses:
A method for training a text classification model, the method comprising: obtaining a first training dataset including a sample text, a first category label corresponding to the sample text, and scenario information corresponding to the sample text (Chopra; p. 0038 - Each computer text segment includes (i) a first portion comprising a transcript of an interaction (sample text) and (ii) a second portion comprising notes about the interaction (scenario information). In some embodiments, the computer text segment comprises unstructured text containing the transcript and the notes (e.g. CSR summary) stored in a database record or table for a given interaction…; see also p. 0072-0073 – Tagging intents with entity tags using NER model (category label)); performing, via an initialized text classification model, semantic enhancement processing on the sample text based on the corresponding scenario information to obtain a semantic enhancement result (Chopra; Fig. 2, item 206; p. 0006-0007 - The server computing device executes the trained neural network using the interaction embeddings to generate an interaction summary for each computer text segment (semantic enhancement), the interaction summary comprising a text string identifying a primary topic of the first portion of the computer text segment; see also p. 0037 & p. 0040); encoding, via the initialized text classification model, the semantic enhancement result through the initialized text classification model to obtain a text encoding corresponding to the sample text (Chopra; Fig. 2, item 208; p. 0006-0007- The server computing device converts each interaction summary into a multidimensional vector representing the interaction summary (encoding)); and performing, via the initialized text classification model, non-linear mapping processing on the text encoding to obtain a text classification result corresponding to the sample text (Chopra; Fig. 2, items 210 & 212; p. 0006-0007- The server computing device aggregates the multidimensional vectors into one or more clusters based upon a similarity measure between the respective multidimensional vectors. The server computing device aligns the one or more clusters of vectors with attributes of the interaction summaries to generate a hierarchical mapping of customer issues (text classification result); see also Fig. 6, item 616 & p. 0081); calculating a loss value based on the text classification result and the first category label; and updating parameters of the initialized text classification model based on the loss value to obtain a trained text classification model.
Although Chopra, in p. 0006-0007 & p. 0081, discloses hierarchical mapping of multidimensional vectors. Chopra fails to explicitly disclose that the applied mapping processing on the text encoding to obtain a text classification result of the input text is non-linear, calculating a loss value based on the text classification result and the first category label, and updating parameters of the initialized text classification model based on the loss value to obtain a trained text classification model. Mourya does teach performing, via the initialized text classification model, non-linear mapping processing on the text encoding to obtain a text classification result corresponding to the sample text (Mourya; p. 0114-0119 – The disclosure presents a plurality of machine learning classification models, some of which may be non-linear, such as Support Vector Machine (SVM): SVM is a supervised machine learning algorithm that may be used for classification tasks. In the fallback classification process, an SVM model may be trained using the existing intent training phrases as feature vectors and their corresponding intent labels as target variables. SVM aims to find a hyperplane that best separates the training data into different intent classes. By utilizing a kernel function, the SVM model may handle non-linear decision boundaries, allowing it to classify fallback utterances into the closest matching existing intent); calculating a loss value based on the text classification result and the first category label (Mourya; p. 0223 - Optimization and Iteration: During training, the model undergoes an optimization process to adjust its internal parameters and optimize its performance. This is typically done by minimizing a loss function that quantifies the discrepancy between the predicted intent labels (classification result) and the true intent labels of the training data (first category label). The optimization process aims to find the best parameters that minimize this discrepancy); and updating parameters of the initialized text classification model based on the loss value to obtain a trained text classification model (Mourya; p. 0223 - Optimization and Iteration: During training, the model undergoes an optimization process to adjust its internal parameters and optimize its performance. This is typically done by minimizing a loss function that quantifies the discrepancy between the predicted intent labels and the true intent labels of the training data. The optimization process aims to find the best parameters that minimize this discrepancy). Therefore, it would have been obvious to one of ordinary skill in the art to modify the method and apparatus of Chopra to include performing, via the initialized text classification model, non-linear mapping processing on the text encoding to obtain a text classification result corresponding to the sample text, calculating a loss value based on the text classification result and the first category label, and updating parameters of the initialized text classification model based on the loss value to obtain a trained text classification model, as taught by Mourya, because by employing classification models such as logistic regression, decision tree, random forest, SVM, or AutoML, the system may perform the classification of fallback utterances into existing intent classes. It is to be noted that each model has its own capability to learns the patterns and relationships between training phrases and intent labels, enabling accurate prediction and identification of the most appropriate existing intent for a given fallback utterance (Mourya; p. 0120).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure. The prior art made of record and not relied upon includes: Lin (US PG Pub 20230320642) discloses methods, systems, and other implementations for processing, analyzing, and modelling psychotherapy data. The implementations include a method for analyzing psychotherapy data that includes obtaining transcript data representative of spoken dialog in one or more psychotherapy sessions conducted between a patient and a therapist, extracting speech segments from the transcript data related to one or more of the patient or the therapist, applying a trained machine learning topic model process to the extracted speech segments to determine weighted topic labels representative of semantic psychiatric content of the extracted speech segments, and processing the weighted topic labels to derive a psychiatric assessment for the patient (Lin; Abstract).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Rodrigo A Chavez whose telephone number is (571)270-0139. The examiner can normally be reached Monday - Friday 9-6 ET.
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, Richemond Dorvil can be reached at 5712727602. 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.
/RODRIGO A CHAVEZ/Examiner, Art Unit 2658
/RICHEMOND DORVIL/Supervisory Patent Examiner, Art Unit 2658
Prosecution Insights