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 .
DETAILED ACTION
Claims 1-20 are presented for examination.
Claims 1, 9 and 17 were amended.
This is a Final Action.
Response to Arguments
Applicant’s arguments with respect to claim(s) 1-20 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
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 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 of this title, 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 1-3, 6-7, 9-12, 14-15 and 17-20 are rejected under 35 U.S.C. 103 as being unpatentable over Tan et al. (US 11,270,077) in view of Hung et al. (US 11,423,235) further in view of Xie et al. (US 11,531,821)
1. Tan teaches, A method comprising:
receiving, by an agent, a prompt from a user (Abstract – receives a natural language input from a user, Col 3: lines 16-18 – teaches a NPL based conversation service receives user utterance);
…the LLM trained using a taxonomy having a plurality of root classifications, and a set of child classifications under each root classification (Col 4: lines 15-61 – teaches hierarchical structure 100 illustrates a user utterance 110 received at a node illustrated as root 112, incorporating multiple child ndoes extending down, each nod representing a node of a separate domain and trained for a specific knowledge area);
receiving, from the LLM, a classification corresponding to one of the child classifications based on the prompt (Col 5: lines 56-62 – teaches each IC (in-domain classifier) represents a multi-class identifier that assigns a real in-domain label or identifies a signature of sub-domains);
Tan does not explicitly teach,
inputting the prompt into a large language model (LLM), wherein the LLM is configured to generate natural language text as output;
deploying a hierarchical set of skills, to iteratively identify one or more LLM calls based on the classification, wherein each classification in the hierarchical set of skills is mapped to a respective skill and each skill is an LLM call that is configured to invoke the LLM to perform a task;
responsive to detecting an exception during deployment of the hierarchical set of skills, overriding the deployment by performing an exception activity.
However, Huang teaches,
inputting the prompt into a large language model (LLM) (Col 10: lines 19-44 – teaches input… intent and entity data… to a predictive model – disclosing inputting user-derived data into a deep learning predictive model (LLM equivalent), Huang), wherein the LLM is configured to generate natural language text as output (Col 2: lines 10-11 – teaches output a response to the user input using the selected chatbot – discloses generating natural language response, Huang);
deploying a hierarchical set of skills (Col 3: lines 17-29 – adaptive orchestration… selects a single-task dialogue system… at any point in conversation – disclosing orchestration across multiple functional unites (skills), Huang), to iteratively identify one or more LLM calls based on the classification (Fig 7: 743 – teaches predict dialogue path choice based on intents/entities – disclosing iterative selection of dialogue path based on classification, Huang), wherein each classification in the hierarchical set of skills is mapped to a respective skill and each skill is an LLM call that is configured to invoke the LLM to perform a task (Col 2: lines 7-8 - select a chatbot of the set likely to have a best response to the user input – disclosing mapping classification (intent/entity) to specific chatbot (skill), Huang);
It would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which said subject matter pertains to allow Tan to be combined with Huang as taught by Huang because both Tan and Huang are in the same field of endeavor of dialogue systems, classification and orchestration POSITA would be motivated to integrate Tan’s hierarchical taxonomy classification with Huang’s adaptive orchestration to address exceptions and improve robustness in the system.
However, Xie teaches,
responsive to detecting an exception during deployment of the hierarchical set of skills (Fig 8 - determine… confidence score… if intent is uncertain… - discloses detecting unreliable classification to exception condition, Xie, overriding the deployment by performing an exception activity.(Figs 4-5 – teaches modify… input… re-evaluate intent… ask user for confirmation – discloses overriding initial processing and performing alternative action, Xie).
It would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which said subject matter pertains to combine Xie with Tan and Huang because all three references are directed to natural language processing systems that classify user inputs and dynamically control subsequent processing. A POSITA would have been motivated to incorporate Xie’s confidence based exception handling into the hierarchical and orchestrated framework of Tan and Huang in order to improve robustness and accuracy when initial classifications are uncertain or incorrect. This combination represents a predictable use of kown techniques, classification, orchestration and adaptive reprocessing to improve conversational system performance, yielding predictable results consistent with KSR.
2. The combination of Tan, Huang and Xie teach, The method of claim 1, wherein inputting the prompt into the LLM comprises an auto- regressive input that iteratively progresses through hierarchical classifications until a final child node is encountered (claim 1 – teaches inputting the prompt into a multi-domain context-based hierarchy to a leave node by selecting a parent domain node then selecting a subdomain node until labels the natural language input with a classification label, Tan).
3. The combination of Tan, Huang and Xie teach, The method of claim 1, wherein the hierarchical set of skills comprises a chain of two or more skills performed based on the classification (Col: 5: lines 61-20 – teaches ICs may execute recursively in the IC of the subdomains).
6. The combination of Tan, Huang and Xie teach, The method of claim 1, further comprising:
retrieving, from a database, a plurality of categories that classify conversations based on content of the conversations (Abstract, Col 10: lines 12-18 – new intents and entities are discovered and used to update an existing dialogue path (which requires classification), Huang);
determining whether the prompt aligns with one of the plurality of categories (Col 10: lines 12-18 - teaches predictive model 157 applied to determine the most likely task or intent the user input, Huang);
responsive to determining that the prompt does not align with one of the plurality of categories, retrieving, from the database, a set of previously proposed categories (Abstract & Col 7: lines 49-51– additional data is continually collected and used to retain model 157 and expand the intent/entity datable, Huang);
determining whether the prompt aligns with one of the set of previously proposed categories (Col 10: lines 12-18 - teaches predictive model 157 applied to determine the most likely task or intent the user input, Huang);
responsive to determining that the prompt does not align with one of the set of previously proposed categories, providing the prompt to the LLM to generate a new category for classifying the prompt (Abstract, Col 10: lines 12-18 – new intents and entities are discovered and used to update an existing dialogue path (which requires classification), Huang;
receiving output from the LLM, the output comprising at least one recommended category; and tagging the prompt with the at least one recommended category from the output (Col 10: lines 1-18 - teaches new intents are used to update the dialogue path and stored for future classicization, Huang).
7. The combination of Tan, Huang and Xie teach, The method of claim 6, wherein determining whether the user prompt aligns with one of the plurality of categories comprise:
generating a prompt embedding representing the user prompt (Claim 1 - teaches processing the user unput to determine intent data and entity data which is based on feature vector, Huang);
retrieving a plurality of category embeddings each represents a respective category (Col 10: lines 1-18– teaches processing moves to block 330, where any new intent and entity values discovered are used to update a master intent/entity set, wherein each intent and entity is combined together to generate a feature vector.. different chatbots have different intent and entity sets which are collectively from the feature space (i.e. teaching multiple category representations (intent/entity sets) from chatbots, Huang); and
determining a similarity between the prompt embedding with one or more of the plurality of category embeddings (Col 10: lines 32-39 – teaches from all the workspaces 1 through N the predictive model selects one of the workspaces as providing the best answer for the user’s input, Huang).
Claims 9 and 17 are similar to claim 1 hence rejected similarly.
Claims 10-16 is similar to claims 2-8 respectively hence rejected similarly.
Claim 18-20 is similar to claims 2, 3 and 6 respectively hence rejected similarly.
Claims 4, 8 12 and 16 are rejected under 35 U.S.C. 103 as being unpatentable over Tan et al. (US 11,270,077) in view of Hung et al. (US 11,423,235) and Xie et al. (US 11,531,821) further in view of Lara et al. (WO2024/136871)
All limitations of claim 1 are taught above.
4. The combination of Tan, Huang and Xie do not explicitly teach, wherein detecting the exception comprises detecting a hallucination made by the LLM using at least one additional LLM.
However, Lara teaches, wherein detecting the exception comprises detecting a hallucination made by the LLM using at least one additional LLM (Paragraph 21-22 – teaches a conversation data may be automatically generated by one or more language generative models, which may be employed as evaluation data that is used to evaluate the performance of the conversational agent via one or more benchmarks).
It would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which said subject matter pertains to allow Tan to be combined with Lara as taught by Lara because both Tan and Lara are in the same field of endeavor of conversation AI based on classification and training/evaluation; POSITA would be motivated to integrate Lara’s techniques as a natural way to improve Tan’s exception handling beyond simple off-topic detection.
Claim 12 is similar to claim 4 hence rejected similarly.
All limitations of claim 7 are taught above.
8. The combination of Tan, Huang and Xie does not explicitly disclose, wherein the LLM is trained by:
accessing a plurality of training examples, each training example including a conversation between an agent and a user, wherein the conversation is generated by an additional large language model.
However, Lara teaches,
accessing a plurality of training examples, each training example including a conversation between an agent and a user, wherein the conversation is generated by an additional large language model.(Paragraph 21-22 – teaches generating synthetic conversation data using one LLM then training another model on it)
It would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which said subject matter pertains to allow Tan to be combined with Lara as taught by Lara because both Tan and Lara are in the same field of endeavor of conversation AI based on classification and training/evaluation; POSITA would be motivated to integrate Lara’s techniques as a natural way to improve Tan’s exception handling beyond simple off-topic detection.
Claim 8 is similar to claim 16 hence rejected similarly.
Claims 5 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Tan et al. (US 11,270,077) in view of Hung et al. (US 11,423,235) and Xie et al. (US 11,531,821) further in view of Vibbert et al. (US 2016/0042735)
All limitations of claim 1 are taught above.
5. The combination of Tan, Huang and Xie do not explicitly teach, wherein detecting the exception comprises detecting content in a response generated by the LLM that violates a guardrail.
However, Vibbert teaches, wherein detecting the exception comprises detecting content in a response generated by the LLM that violates a guardrail (Paragraph 11 - the computing device may determine whether execution of the second dialog task should be prevented).
It would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which said subject matter pertains to allow Tan to be combined with Vibbert as taught by Vibbert because both Tan and Vibbert are in the same field of endeavor of NPL processing based on conversation system; POSITA would be motivated to integrate Vibbert’s teaching to improve Tan’s exception handling (off-topics) with Vibbert’s guardrail enforcement.
Claim 13 is similar to claim 5 hence rejected similarly.
Conclusion
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 AMRESH SINGH whose telephone number is (571)270-3560. The examiner can normally be reached Monday-Friday 8am-5pm.
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, Ann J. Lo can be reached at (571) 272-9767. 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.
/AMRESH SINGH/Primary Examiner, Art Unit 2159