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 .
Response to Arguments
Applicant’s arguments with respect to claim(s) 1-3, 5-6, 8-10, 12-13, 15-16, and 18-19 have been considered but are moot in view of the new ground of rejections.
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 1-2, 5-6, 8-9, 12-13, 15-16, and 18-19 are rejected under 35 U.S.C. 103 as being unpatentable over Aftab, Omar (“Copilot in Power Virtual Agents: Next-generation AI assists bot building with natural language”) in view of Hirzel et al. (US Pub 2019/0066694) and in further view of Gelfenbeyn et al (US Pub 2017/0300831).
Regarding claim 1, Aftab discloses a computer-implemented method comprising:
configuring the agent for use by a digital assistant (page 2 – “Introducing Copilot in Power Virtual Agents, a new feature that uses the latest generative AI capabilities to make this possible. An author can simply state: “allow a user to start planning an event, collect user’s email address and phone number, and let them choose the event type from wedding, corporate, and social event,” and a dialog to do so will be instantly created—complete with trigger phrases, entities, variables, and appropriate branching. The author can then select part of the dialog and use additional natural language prompts to refine it—or use the Power Virtual Agents intuitive graphical user interface to tweak it manually” - agent to be with Power Virtual Agents), wherein the configuring comprises:
defining, based on natural language input from a user, specification parameters including an identification of the agent, a purpose of the agent, and identification of one or more assets for implementing the purpose (page 2 – “’allow a user to start planning an event, collect user’s email address and phone number, and let them choose the event type from wedding, corporate, and social event,’ and a dialog to do so will be instantly created—complete with trigger phrases, entities, variables, and appropriate branching.” – discloses an event planning bot (identification), to plan events (purpose), with email-phone and event type (assets for the purpose); also see page 3 – “Here are just a few examples of natural language prompts you can use:
‘Let a user check the status of a flight, accepting the flight number and date. For each question, add two message variations.’
‘Accept a user’s name, age, and date of birth, and then repeat their responses back to them.’
‘Create a support ticket, including a title, severity (high/medium/low), description, and an email address to send update notifications to. Summarize the information in an Adaptive Card.’”),
defining configuration information for the one or more assets (page 3 – “Create entire topics from a simple description, including relevant trigger phrases, questions, entity, messages, variables, and other logic.
Add or update content in an existing topic, such as asking for additional questions to be added or updating existing nodes (example: providing message variations).
Summarize information collected from a user in an interactive, graphical Adaptive Card, with all the JSON for the card generated automatically.
Iterate over just part of a dialog—select one or more nodes to scope your request to that specific part of the dialog.” – for example “trigger phrases, questions, entity, messages, variables” are configuration information for the assts that implement the agent’s purpose), and
defining the one or more actions based on the configuration information, the natural language input from the user, or both (page 3 – “Add or update content in an existing topic, such as asking for additional questions to be added or updating existing nodes (example: providing message variations).” – action based on natural language input; also see page 2);
Aftab does not explicitly disclose accessing a container defining an agent configurable to have one or more actions;
generating a specification document that characterizes the agent, wherein generating the specification document comprises: acquiring, from the agent, metadata associated with: the specification parameters, the one or more assets, the one or more actions, or any combination thereof, and writing the specification document to include the metadata and the identification of the agent; and
storing the specification document in a data store that is communicatively coupled to the digital assistant.
Hirzel discloses accessing a container defining an agent configurable to have one or more actions (para 0016, 0020, 0063 - conversational bot specification);
generating a specification document that characterizes the agent (para 0063), wherein generating the specification document comprises: acquiring, from the agent, metadata associated with: the specification parameters, the one or more assets, the one or more actions, or any combination thereof (para 0020 and table 1 – metadata such as “summary, parameters, and responses”; also see para 0024, 0041), and writing the specification document to include the metadata and the identification of the agent (para 0030, 0032); and
storing the specification document in a data store that is communicatively coupled to the digital assistant (para 0023, 0030).
Therefore, it would have been obvious to a person of ordinary skilled in the art before the effective filing date of the claimed invention to modify Aftab with the teachings of Hirzel in order to use a bot that is “self-documenting, so for example, that users who do not know how to use the bot or the web API can find out how to do that by interacting with the bot” (Hirzel, para 0062).
Aftab in view of Hirzel does not disclose wherein the specification document is indexed in the data store to facilitate identification and selection of the agent, from among a plurality of candidate agents, for responding to a natural language utterance received by the digital assistant.
Gelfenbeyn discloses wherein the specification document is indexed in the data store (para 0050) to facilitate identification and selection of the agent, from among a plurality of candidate agents (para 0067, 0072), for responding to a natural language utterance received by the digital assistant (see abstract; 0003, para 0089).
Therefore, it would have been obvious to a person of ordinary skilled in the art before the effective filing date of the claimed invention to modify Aftab in view of Hirzel with the teachings of Gelfenbeyn in order to “conserve network and/or processor resources that may otherwise be consumed by an initial failed attempt to utilize an agent to perform the intent, which is then followed by invoking an alternative agent in another attempt to perform the intent” (Gelfenbeyn, para 0010).
Regarding claim 2, Aftab discloses further comprising:
receiving, by a generative artificial intelligence model, a natural language utterance from the user; and
testing, by the generative artificial intelligence model using the natural language utterance, functionality of the agent to implement the purpose of the agent,
wherein configuring the agent is performed by the generative artificial intelligence model (see page 2 – “Introducing Copilot in Power Virtual Agents, a new feature that uses the latest generative AI capabilities to make this possible. An author can simply state: “allow a user to start planning an event, collect user’s email address and phone number, and let them choose the event type from wedding, corporate, and social event,” and a dialog to do so will be instantly created”).
Regarding claim 5, Hirzel discloses wherein defining the configuration information for the one or more assets comprises:
accessing a specification for each of the one or more assets (para 0016);
extracting the configuration information from the specification for each of the one or more assets (para 0020 – configuration information for each asset such as summary, paramets, etc; para 0063); and
writing the configuration information into the container (para 0063- “based on parsing the API specification, a conversational bot specification is constructed”; para 0032).
Regarding claim 6, Hirzel discloses further comprising:
receiving a natural language utterance as a user input to the digital assistant (para 0063 – “a natural language expression such as a text expression or an utterance or audible speech is received from a user.”); and
generating a response to the natural language utterance based on the specification document and the agent (para 0064-0065).
Regarding claims 8 and 15, see rejection of claim 1.
Regarding claims 9 and 16, see rejection of claim 2.
Regarding claims 12 and 18, see rejection of claim 5.
Regarding claims 13 and 19, see rejection of claim 6.
Claims 3 and 10 are rejected under 35 U.S.C. 103 as being unpatentable over Aftab, Omar (“Copilot in Power Virtual Agents: Next-generation AI assists bot building with natural language”) in view of Hirzel et al. (US Pub 2019/0066694) and in further view of Gelfenbeyn et al (US Pub 2017/0300831) and in further view of Liang et al. (“TaskMatrix.AI: Completing Tasks by Connecting Foundation Models with Millions of APIs”).
Regarding claim 3, Aftab in view of Hirzel and Gelfenbeyn discloses the method of claim 1.
Aftab in view of Hirzel does not disclose wherein the testing is an interactive process between the user and the generative artificial intelligence model, the interactive process comprising:
(i) evaluating the natural language utterance based on the specification document;
(ii) selecting the agent for use in responding to the natural language utterance based on evaluating the natural language utterance;
(iii) generating an execution plan comprising the one or more actions defined for the agent;
(iv) executing, based on the execution plan and the configuration information for the one or more assets, the one or more actions to obtain output data;
(v) generating a response to the natural language utterance based on the output data; and
(vi) evaluating, based on the response, the functionality of the agent to implement the purpose of the agent.
Liang discloses wherein the testing is an interactive process between the user and the generative artificial intelligence model, the interactive process comprising:
(i) evaluating the natural language utterance based on the specification document (Section 2.1, 2.2, 2.3 – MCFM receives the user’s natural language utterance and evaluates it against the API to plan a solution);
(ii) selecting the agent for use in responding to the natural language utterance based on evaluating the natural language utterance (section 2.4 and Fig. 1 – selecting from the various APIs based on the utterance);
(iii) generating an execution plan comprising the one or more actions defined for the agent (section 2.2 – “takes each user instruction and the corresponding conversational context as input and generates a solution outline. Users often use brief expressions to convey their high-level task intentions, so MCFM generates a more comprehensive textual description of the steps required to complete the task” and section 3.3 – MCFM generates a solution outline (execution plan) and then generates action codes comprising the specific API calls);
(iv) executing, based on the execution plan and the configuration information for the one or more assets, the one or more actions to obtain output data (see section 2.1 – “API Executor, which can execute the generated action codes by calling the relevant APIs and return the intermediate and final execution results.” and 2.5);
(v) generating a response to the natural language utterance based on the output data (see figs. 2, 6, 7, and 9); and
(vi) evaluating, based on the response, the functionality of the agent to implement the purpose of the agent (section 2.5 – “To enhance accuracy and reliability, the action executor also requires a verification mechanism to confirm whether the generated code or outcomes satisfy the tasks specified in human instructions”; also see fig. 14; and section 4.4 – “This process can also be automated by incorporating a separate
module to test each feedback item individually and adding them when the test results demonstrate
improvement”).
Therefore, it would have been obvious to a person of ordinary skilled in the art before the effective filing date of the claimed invention to modify Aftab in view of Hirzel with the teachings of Liang in order to use RLHF (Reinforcement Learning with Human Feedback) to benefit from the knowledge and insight of human feedback to enhance MCFM and API selector so that this can result in faster convergence and better performance of complex tasks (Liang, section 2.6).
Regarding claim 10, see rejection of claim 3.
Allowable Subject Matter
Claims 4, 7, 11, 14, 17, 20 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.
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 NAFIZ E HOQUE whose telephone number is (571)270-1811. The examiner can normally be reached M-F 8-5.
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, Ahmad Matar can be reached at (571)272-7488. 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.
/NAFIZ E HOQUE/Primary Examiner, Art Unit 2693
Prosecution Insights