Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Interpretation
2. Upon further consideration, the claim term “module” is not being interpreted under 35 U.S.C. 112(f). In view of the specification and the relevant field of endeavor, one of ordinary skill in the art would understand the claimed “application modules available in an environment of an interpreter to refer to software modules, libraries, packages, APIS or other executable software components available to an interpreter environment, rather than as a nonce substitute for “means.” The claims further recite the modules in the context of an interpreter. The claims further recite the modules in the context of an interpreter, documentation regarding the modules and function calls to functions of the modules, which provides sufficient software context and identifies the modules as known software structures having callable functions. Accordingly the term “module” is understood as software/application modules available for use by the interpreter and the prior interpretation under 35 U.S.C. 112(f) is withdrawn.
Response to Arguments
3. Applicant’s arguments with respect to claims 1 and 14 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.
Specifically, newly added limitations are taught by newly cited Mishchenko (US 2024/0319971).
Claim Rejections - 35 USC § 103
4. 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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
5. Claims 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Lee et al. (US 2023/0315856), herein Lee in view of Mishchenko et al. (US 2024/0319971), herein Mishchenko.
Regarding Claim 1:
Lee discloses a method comprising:
applying a history to an input of a trained natural language model to generate a first textual output, wherein the history contains a first user query, and wherein the first textual output includes, as a command in a language of an interpreter, a request for documentation regarding a first module of a plurality of application modules available in an environment of the interpreter (Lee: ¶61 discloses multiple machine learning models trained on corpora and context, ¶65 discloses the system receives user queries via natural language input and generates a template query through intent inference based on the natural language request. ¶75 and ¶98-100 discloses that the functions carried out by the natural language analysis device are implemented through an API that supports natural language processing, template query generation, methods, filters, filter values and filter ranges. In particular, Lee discloses API calls including getting methods and filters and further discloses a JSON response identifying parameters of interest, including methods associated with the intended function call such as query_alert. The retrieved method, filter and parameter information constitutes documentation or descriptive API information regarding available callable functionality of the management system Additionally, ¶128 explicitly states the computer code may include higher level instructions executed by a computer using an interpreter. Therefore Lee teaches a first module/function group of a plurality of available software commands/functions in an interpreter software execution environment associated with the first user query),
;
in response to receiving the first textual output, applying the first textual output to the interpreter to and receiving, from the interpreter, a first interpreter output as a response to the first textual (Lee: p[0075] p[0085] and p[0095] discloses the finalized query is provided to a system which executes the task and ¶128 discloses higher level instructions may be executed by a computer using an interpreter);
in response to receiving the first interpreter output, updating the history by adding a representation of the first interpreter output to the history (Lee: p[0075] p[0077] p[0086] and p[0094] discusses how results from execution are stored and inform further interactions, user adjustments based on system outputs are captured and used to retrain the models, effectively updating the system’s memory/history);
in response to updating the history, applying the updated history to the input of the trained natural language model to generate a second textual output that comprises, in the language of the interpreter, a first function call to a first function of the first module (Lee: p[0075] p[0077] p[0085] and p[0094) discloses that the revised or re-submitted final query is executed by invoking a command to the system); and
in response to receiving the second textual output, applying the first function call of the second textual output to the interpreter (Lee: p[0075] p[0077] p[0086] and p[0094) discloses that the revised or re-submitted final query is executed via a backend system).
Lee does not explicitly disclose:
wherein the history does not include information regarding documentation of the modules of the plurality of application modules, wherein the first module is associated with the first user query;
wherein the first interpreter output includes documentation regarding the first module.
However, Mishchenko discloses
wherein the history does not include information regarding documentation of the modules of the plurality of application modules, wherein the first module is associated with the first user query (Mishchenko: ¶29, ¶34, ¶43-46 disclose that natural language models are limited to information from training data, that the mode receives and uses current external tool/API documentation, that an API may be Python API, that the system accesses a manifest and description of the API and that the manifest/description provides explanations of call functions available with the tool or application Therefore before the API description is accessed, the initial conversation containing the user request does not already include the documentation of the available external API module functions, rather the documentation is obtained from the external manifest API description after the relevant tool or module is identified based on the user query);
wherein the first interpreter output includes documentation regarding the first module (Mishchenko: ¶34, ¶43-46 discloses accessing a manifest and description of the web API, wherein the web API may be a Python API and wherein the description of the web API may include descriptions of available call functions and how to use such call functions).
Lee and Mishchenko are in the same field of endeavor, i.e., both disclose systems or methods for using natural language models or natural language interfaces to generate executable commands, queries, function calls or API calls for interacting with external software systems. It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to disclose modifying Lee to access and Use Mishchenko’s external tool/API documentation, manifest information and API description when generating executable commands or function calls for available software functions. The modification would have allowed Lee’s trained natural language model to use current documentation for available external functions before generating a command function call, therefore improving the model’s ability to correctly invoke available software functionality without relying solely on potentially outdated training data, because Mishchenko explicitly teaches that its solution enables “natural language models to receive and readily use in addition to any training data, information that is most recent and most specific as provided by a publisher of an external tool or application (Mishchenko ¶29).
Regarding Claim 2:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 1, further comprising receiving the first user query by: generating, using the trained natural language model, a third textual output (Lee: p[0064-0069] discloses that trained ML models can generate multiple different outputs (queries or commands) based on different prompts);
applying the third textual output to the interpreter, wherein the third textual output comprises a command to return at least one prior user input (Lee: p[0095] and p[0096] discloses commands that interact with the system context and memory, such as a finalized query can be converted to a complex SQL query or invoke identified functions, obtaining a list of HTTP connections with sources, modify a security setting etc. Since users interactively refine commands based on past queries and result context in p[0075-0077] and the system maintains dialog context/history. A function that queries or surfaces previous input or system state is reasonable encompassed);
and receiving a second interpreter output from the interpreter in response to applying the third textual output thereto, wherein the second interpreter output is representative of the first user query (Lee: p[0075-0077], p[0095] the output of interpreter executed queries may include representations of prior query content or system history, e.g., summaries of prior actions or visualized parameter settings – which could include fata reflective of prior user inputs).
Regarding Claim 3:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 1, wherein the first module is at least one of a knowledgebase query module, a reservation module, a server management module, a database management or access module, a user privileges modification or query module, a user biographical information modification or query module, a telecommunications module, a commercial services query module, or a map query module (Lee: p[0044-0057] discloses a body of information and returning summaries and reports about system state reasonably teaches a knowledgebase query module server management module database access module user privileges module).
Regarding Claim 4:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 1, further comprising: adding a representation of a second interpreter output to the history, wherein the second interpreter output is received from the interpreter in response to applying the second textual output thereto (Lee: p[0075-0077] and p[0085] discloses “the NL analysis device receives the finalized query and generates a system command… such that the tasks may be performed to return results” and “the user can view the results and adjust… to generate an updated finalized query, any number of times…” this means the returned results from the second textual output (a finalized query) inform what gets stored and fed back into future inference);
subsequent to adding the representation of the second interpreter output to the history, applying the history to the input of the trained natural language model to generate third textual output (Lee: p[0077], p[0095] p[0094] in multiple cases the history (which now includes the second interpreter output) is implicitly or explicitly reused to produce new text – the third textual output); and
presenting a representation of the third textual output (Lee: p[0092-0093] discloses that after each ML inference step the generated query is presented to the user interface).
Regarding Claim 5:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 4, wherein presenting the representation of the third textual output comprises:
applying the third textual output to the interpreter, wherein the third textual output comprises a command to provide a representation of the second interpreter output to a user (Lee: p[0095] the finalized queries are converted into commands that return results to the user, satisfying the requirement that the third output retrieves and presents previous output).
Regarding Claim 6:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 1, further comprising: adding a representation of a second interpreter output to the history, wherein the second interpreter output is received from the interpreter in response to applying the second textual output thereto, and wherein the second interpreter output includes an error message (Lee: p[0086] discloses error feedback as part of execution results);
subsequent to adding the representation of the second interpreter output to the history, applying the history to the input of the trained natural language model to generate third textual output, wherein the third textual output represents a request for additional information related to the first function call message (Lee: p[0075] user provides clarifying follow-up, the model generates refined query);
presenting a representation of the third textual output; message (Lee: p[0093] presents results via editable templates or suggestions);
responsive to presenting the representation of the third textual output, receiving a first user response message (Lee: p[0075] discloses system receiving new input from user after viewing output);
in response to receiving the first user response, adding a representation of the first user response to the history message (Lee: p[0075] p[0086] discloses updating history and retraining based on it);
subsequent to adding the representation of the first user response to the history, applying the history to the input of the trained natural language model to generate fourth textual output message (Lee: p[0070-0085] the machine learning model generates further query output based on new-extended history);
and applying the fourth textual output to the interpreter, wherein the fourth textual output comprises a second function call to the first function of the first module message (Lee: p[0095] the system re-executed updated query with modified parameters any number of times).
Regarding Claim 7:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 6, wherein presenting a representation of the third textual output comprises applying the third textual output to the interpreter, wherein the third textual output comprises a command to provide a representation of the second interpreter output to a user, and wherein receiving the first user response comprises:
generating, using the trained natural language model, a fifth textual output (Lee: p[0095] supports this by discloses presentation of previous results to user);
applying the fifth textual output to the interpreter, wherein the fifth textual output comprises a command to return at least one prior user input (Lee: p[0075]iterative query refinement supported);
and receiving a third interpreter output from the interpreter in response to applying the fifth textual output thereto, wherein the third interpreter output is representative of the first user response (Lee: p[0095] returns updated results based on this response).
Regarding Claim 8:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 1, further comprising:
receiving the first user query (Lee: p[0072] the user inputs natural language phrase via the NL interface interpreted as the first query);
adding a representation of the first user query to the history (Lee: p[0075] p[0085] and p[0094] discloses the natural language query and subsequent actions are stored in the system’s) context/history for further inference and learning);
and prior to receiving the first user query:
generating, using the trained natural language model, a third textual output (Lee: p[0075] p[0086]] the model can generate textual output based on initial prompts or historical interaction even before receiving a new user query);
applying the third textual output to the interpreter, wherein the third textual output comprises a request to return information about a set of modules that are usable by the interpreter, wherein the first module is a member of the set of modules (Lee: p[0095-0096] discloses the system can be queried about capabilities of modules and supported operations. This includes requesting lists of actions, query fields, etc., and their relationship to backend modules);
receiving a second interpreter output from the interpreter in response to applying the third textual output thereto, wherein the second interpreter output is representative of capabilities of each module of the set of modules(Lee: p[0095] the management system returns structured data describing the capabilities of each queried module); and adding a representation of the second interpreter output to the history (Lee: p[0094-0095] the results returned are retained in history for further model training and contextual refinement).
Regarding Claim 9:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 1, further comprising, prior to applying the history to the input of the trained natural language model to generate the second textual output and subsequent to adding the representation of the first interpreter output to the history:
applying the history to the input of the trained natural language model to generate third textual output (Lee: p[0094-0095] a new query is generated by the ML model using updated history after storing the first interpreter output);
applying the third textual output to the interpreter, wherein the third textual output comprises a request for at least one of information about the first module or information about the first function (Lee: p[0096] discloses the system requesting function/module details using ML-inferred textual commands, such as supported actions, schema or descriptions);
and adding a representation of a second interpreter output to the history, wherein the second interpreter output is received from the interpreter in response to applying the third textual output thereto (Lee: p[0075], p[0094] discloses the output from this metadata/function query is stored in the systems history).
Regarding Claim 10:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 1, wherein the trained natural language model includes more than a billion parameters and has been trained on a corpus of generic speech, and wherein the history includes, prior to adding a representation of the first user query thereto, representations of at least two examples of goal-oriented dialog using the interpreter (Lee: p[0066] explicitly discloses GPT-3, GPT-Neo 2.7B and similar large models with billions of parameters trained on general language corpora).
Regarding Claim 11:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 1, wherein the trained natural language model has been trained using a plurality of representations of goal-oriented dialog using the interpreter (Lee: p[0072-0074] and p[0077] discloses training the model with feedback data, examples of user-to-user system interactions and generated queries, all tied to performing backend actions-representing goal oriented dialog).
Regarding Claim 12:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 11, wherein the plurality of representations of goal-oriented dialog using the interpreter used to train the trained natural language model comprises a representation of at least one instance of each of: calling a function, receiving an exception in response to calling a function, loading a module, loading documentation about a module or a function, receiving a user query, and generating a user response (Lee: p[0095] and p[0074] discloses various tasks performed via management modules: function invocation, search queries, module capabilities, user responses and error handling).
Regarding Claim 13:
The proposed combination of Lee in view of Mishchenko further discloses the method of claim 1, wherein the history includes, prior to adding a representation of the first user query thereto, a representation of at least one of a past user interaction, information about a user, or a list of modules accessible by the interpreter (Lee: p[0094], p[0074] disclose the system history as storing user profiles prior actions and available system modules for contextual refinement).
Regarding Claim 14:
Claim 14 has been analyzed with regard to claim 1 (see rejection above) and is rejected for the same reasons of obviousness used above.
Regarding Claim 15:
Claim 15 has been analyzed with regard to claim 4 (see rejection above) and is rejected for the same reasons of obviousness used above.
Regarding Claim 16:
Claim 16 has been analyzed with regard to claim 6 (see rejection above) and is rejected for the same reasons of obviousness used above.
Regarding Claim 17:
Claim 17 has been analyzed with regard to claim 7 (see rejection above) and is rejected for the same reasons of obviousness used above.
Regarding Claim 18:
Claim 18 has been analyzed with regard to claim 8 (see rejection above) and is rejected for the same reasons of obviousness used above.
Regarding Claim 19:
Claim 19 has been analyzed with regard to claim 9 (see rejection above) and is rejected for the same reasons of obviousness used above.
Regarding Claim 20:
Claim 20 has been analyzed with regard to claim 13 (see rejection above) and is rejected for the same reasons of obviousness used above.
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 IAN SCOTT MCLEAN whose telephone number is (703)756-4599. The examiner can normally be reached "Monday - Friday 8:00-5:00 EST, off Every 2nd Friday".
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, Hai Phan can be reached at (571) 272-6338. 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.
/IAN SCOTT MCLEAN/Examiner, Art Unit 2654
/HAI PHAN/Supervisory Patent Examiner, Art Unit 2654
Prosecution Insights