DETAILED ACTION
This Office Action is in response to claims filed on 09/16/2024.
Claims 1-20 are pending.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim 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 1-6, 8, 10, 14-17, and 19-20 are rejected under 35 U.S.C. 103 as being unpatentable over Mishchenko et al. Patent No. US 11,922,144 B1 (hereinafter Mishchenko) in view of Singh et al. Patent No. US 11,693,637 B1 (hereinafter Singh).
With regard to claim 1, Mishchenko teaches a computer-implemented method (Abstract, Disclosed herein are methods, systems, and computer-readable media for integrating a particular external application programming interface (API) with a natural language model user interface) comprising:
identifying one or more parameters of a code snippet (Col. 1, lines 50-51, Embodiments may also include identifying the particular external API based on the received input; Col. 5, lines 57-60, 1. Create an API with endpoints configured to be called by a natural language model or an associated system (e.g., an existing developer API or a wrapper around an existing API).) in a first programming language;
determining a usage of the code snippet (Col. 2, lines 1-5, Embodiments may also include integrating the particular external API with the natural language model user interface, wherein the integrating may include accessing a manifest stored in a location associated with the third-party software API; Col. 5, lines 61-65, 2. Write a description specification documenting the created API and a manifest file that includes the location of the description specification and additional metadata usable for training the natural language model (e.g. explanations of call functions available with the tool or application).);
generating an input for a chatbot in a second programming language based at least in part on the one or more parameters and the usage (Col. 5-Col. 6, lines 66-67 and lines 1-5, Given the API documentation (e.g., the manifest file and the description specification) an the context of a conversation with a user via a user interface of the natural language model, the natural language model(s) and associated system(s) may be configured to automatically determine the parameters needed to invoke the publisher-created API in order to fulfill a user intent or request); and
receiving, from the chatbot, an output corresponding to the code snippet upon injecting the input into the chatbot (Col. 6, lines 8-11, In some embodiments, in response to a conversation with a user, a natural language model and/or associated system may generate digital information and access a particular external API).
However, Mishchenko does not explicitly teach a code snippet in a first programming language or generating an input for a chatbot in a second programming language
In analogous art, Singh teaches a code snippet in a first programming language (Col. 2, In automatically translating a base source code snippet, programmed in a base programming language (e.g., C++), the base source code snippet can be processed using a base-to-natural language (base-to-NL) neural network model, to generate a sequence of outputs)
…
generating an input for a chatbot in a second programming language (Col. 4, The generated target source code snippet can then be rendered as output of a software development application for optional review and/or editing by developer(s). For example, an entire program, or module., can be automatically translated from a base programming language (e.g., C++) to a target language (e.g., Python) by automatically translating snippets as describe above.) based at least in part on the one or more parameters and the usage (Col. 3, In some of those implementations, a natural language descriptor of training instance output of a training instance can be generated based on a document string (also referred to herein as a docstring) that corresponds to the base source code snippet of the training instance input of the training instance. For example, the docstring can be one crafted by a developer to describe the general functionality of the base source code snippet. In some versions of those implementations, a base source code snippet and a docstring for the base source code snippet are identified from a repository … and can be utilized to generate multiple disparate training instances for training the base-to-NL neural network.)
It would have been obvious to one of ordinary skill in the art at the time the invention was filed to apply the teachings of Singh with the teachings of Mishchenko in order to provide a method that teaches, at a high level, translation of a code snippet in a first programming language into a second programming language and consumed as chatbot input. The motivation for applying Singh teaching with Mishchenko teaching is to provide a method that allows for a developer unfamiliar with a base programming language to view and edit the source code through translation in another programming language familiar to the developer (Singh, Col. 2). Mishchenko and Singh are analogous art directed towards natural language processing and software engineering arrangement. Therefore, it would have been obvious for one of ordinary skill in the art to combine Singh with Mishchenko to teach the claimed invention in order to provide code snippet translation as input for natural language processing.
With regard to claim 2, Mishchenko and Singh teach the computer-implemented method of claim 1
Mishchenko further teaches wherein the chatbot comprises at least one of:
an artificial intelligence (AI) bot (Col. 4, lines 44-48, The embodiments discussed herein involve or relate to artificial intelligence (AI). AI may involve perceiving, synthesizing, inferring, predicting and/or generating information using computerized tools and techniques (e.g., machine learning); Col. 7, lines 18-20, In some embodiments, the natural language model may be associated with a chatbot (e.g., the natural language model user interface may include a chatbot).); or
a large language model (LLM).
With regard to claim 3, Mishchenko and Singh teach the computer-implemented method of claim 1
Mishchenko further teaches the code snippet comprises at least one of:
one or more application programming interfaces (APIs) (Col. 1, lines 41-46, For example, embodiments of the present disclosure may include a computer-implemented method for integrating a particular external application programming interface (API) with a natural language model user interface, including receiving a first input at the natural language model user interface.); or
one or more services; and
injecting the input into the chatbot comprises exposing the one or more APIs or services to the chatbot (Col. 2, lines 18-21, In some embodiments, the natural language model user interface may be communicably connected to a natural language model that may be trained to call one or more functions based on the manifest and in response to the received input.).
With regard to claim 4, Mishchenko and Singh teach the computer-implemented method of claim 3
Mishchenko further teaches generating the input for the chatbot in the second programming language comprises inserting, within the input for the chatbot, one or more instructions on the usage of the one or more APIs or services (Col. 6, lines 47-54, In some embodiments, at 110, the computer-implemented method may include receiving a first input at the natural language model user interface. For example, a first input may be a user input (e.g., input at a input/output device 418). A user input may include, e.g., at least one of a natural language input, an indication of a desired external API for integration, or a selection of a desired external API for integration.); and
injecting the input into the chatbot comprises providing the one or more instructions to the chatbot (Col. 6, lines 19-22, A natural language model and/or associated system may, based on instructions received from the user, access an external application containing a database of curated data and computation capabilities; Col. 7, lines 8-12, In some embodiments, the natural language model user interface may be communicably connected to a natural language model configured to recognize and interpret user input (e.g., natural language messages) and external API payloads.).
With regard to claim 5, Mishchenko and Singh teach the computer-implemented method of claim 4
Mishchenko further teaches wherein injecting the input into the chatbot comprises enabling the chatbot to use the one or more APIs or services based at least in part on the instructions (Col. 6, lines 8-11 and lines 17-24, In some embodiments, in response to a conversation with a user, a natural language model and/or associated system may generate digital information and access a particular external API … As another example, in some embodiments, in response to a conversation with a user (e.g., including through interpreting digital natural language input), a natural language model and/or associated system may, based on instructions received from the user, access an external application containing a database of curated data and computational capabilities, deliver accurate and up-to-the minute answers for a wide range of questions, and thereby enhance conversational workflows.).
With regard to claim 6, Mishchenko and Singh teach the computer-implemented method of claim 5
Singh further teaches generating a confidence score (Col. 23, lines 40-53, Col. 23, the system selects one or more source code snippets based on evaluating the source code snippets, For example, the system can select a source code snippet, generated at block 510 or block 512, for presentation if one or more evaluations of the source code snippet satisfy evaluation conditions. For instance, the source code snippet can be selected based on … having an embedding that is within a threshold distance of an embedding of the base source code snippet) that represents a likelihood that the output compiles with the instructions based at least in part on the output, the usage, and the instructions (Col. 24, lines 23-30, At block 606, the system determines whether output, of the target source code snippet, if compiled and executed matches output of the base source code snippet if compiled and executed; Col. 24, lines 45-48, At block 608, the system determines how distant, in embedding space, (Examiner notes: confidence score) a generated target embedding for the target source code snippet is to a generated base embedding to the base source code snippet);
determining that the confidence score fails to meet a certain threshold (Col. 24, lines 52-56, In some implementations, the system determines to exclude the target source code snippet based on the distance between the target embedding and the base embedding being greater than a distance threshold); and
in response to determining that the confidence score fails to meet the certain threshold, prompting the chatbot to produce an additional output that is more likely to comply with the instructions (Col. 24, In some additional or alternative implementations, the system determines whether to select the target source code snippet based on comparing the distance between the target embedding and the base embedding, to an additional distance that is between the based embedding and an additional embedding for an additional target source code snippet.).
It would have been obvious to one of ordinary skill in the art at the time the invention was filed to apply the teachings of Singh with the teachings of Mishchenko in order to provide a method that teaches natural language processing output confidence scoring and remediation. The motivation for applying Singh teaching with Mishchenko teaching is to provide a method that allows for evaluation and input remedying of a natural language processing model’s output such that increases output compliance and reliability (Singh, Col. 12). Mishchenko and Singh are analogous art directed towards natural language processing and software engineering arrangement. Therefore, it would have been obvious for one of ordinary skill in the art to combine Singh with Mishchenko to teach the claimed invention in order to provide output confidence scoring and comparison to select and remedy natural language processing outputs.
With regard to claim 8, Mishchenko and Singh teach the computer-implemented method of claim 3
Mishchenko further teaches generating the input for the chatbot in the second programming language comprises:
generating a description of the one or more APIs or services based at least in part on the usage (Col. 5, lines 61-64, 2. Write a description specification documenting the created API and a manifest file that includes the location of the description specification and additional metadata usable for training the natural language model ); and
inserting, within the input for the chatbot, the description of the one or more APIs or services (Col. 5-Col. 6, lines 66-67 and lines 1-5, 3. Given the API documentation (e.g., the manifest file and the description specification) and the context of a conversation with a user via a user interface of the natural language model, the natural language model(s) and associated system(s) may be configured to invoke the publisher-created API in order to fulfill a user intent or request.); and
injecting the input into the chatbot comprises providing the description of the one or more APIs or services to the chatbot (Col. 10, lines 3-8, From the descriptions of functions, the natural language model may select one or more of the functions which it may determine based on the desired external API, a natural language input from a user of the natural language model user interface, and information scraped from the manifest and/or the description of the particular web API).
With regard to claim 10, Mishchenko and Singh teach the computer-implemented method of claim 3
Mishchenko further teaches obtaining a query from a user (Col. 6, lines 54-56, For example, a user may enter text (e.g., into a chat window also configured to display natural language interface output or into a search bar)); and
wherein injecting the input into the chatbot comprises enabling the chatbot to answer the query via the one or more APIs or services (Col. 6, lines 19-24, a natural language model and/or associated system may, based on instructions received from the user, access an external application containing a database of curated data and computational capabilities, deliver accurate and up-to-the minute answers for a wide range of questions, and thereby enhance conversational workflows.).
With regard to claim 14, Mishchenko and Singh teach the computer-implemented method of claim 3
Mishchenko further teaches providing a user interface that enables a user to create a chatbot interface that facilitates communication between the chatbot and the one or more APIs or services (Col. 7, lines 8-12, In some embodiments, the natural language model user interface may be communicably connected to a natural language model configured to recognize and interpret user input (e.g., natural language messages) and external API payloads.);
detecting, via the user interface, an entry by the user of a query to be answered by the chatbot using the one or more APIs or services (Col. 7, lines 12-18, For example, the natural language model may be trained (e.g., using input data reflecting user input and /or external API payloads, using feedback or validation data associated with correct results) to understand a request implicit or explicit in natural language input and an associated function or other tool associated with responding to the request.); and
detecting, via the user interface, a selection by the user of the one or more APIs or services (Col. 6, lines 54-60, For example, a user may enter text (e.g., into a chat window also configured to display natural language model output or into a search bar), a selection of a user interface element (e.g., from within a drop down menu), perform a click-and-drag operation, or perform an other action with an input device that provide a natural language input and/or indication of a particular external API.).
With regard to claim 15, Mishchenko and Singh teach the computer-implemented method of claim 14
Mishchenko further teaches detecting, via the user interface, a modification by the user of at least one feature of the code snippet (Col. 8, lines 5-7, In some embodiments, the manifest may be customizable by a publisher (E.g., management device or system) of the third-party software API.).
With regard to claim 16, Mishchenko and Singh teach the computer-implemented method of claim 14
Mishchenko further teaches detecting, via the user interface, an entry by the user of at least one instruction for the chatbot interface (Col. 6, lines 27-30, a natural language model and/or associated system may, based on instructions received from the user, automate tasks for the user across one or more external applications); and
in response to the entry, directing the chatbot interface to answer the query according to the instruction (Col. 6, lines 20-24, based on instructions received from the user, access an external application containing a database of curated data and computation capabilities, deliver accurate and up-to-the-minute answers for a wide range of questions; Col. 6, lines 30-31, automate tasks for across one or more external applications (e.g., replying to emails, locating data in a file or a database, or editing documents).
With regard to claim 17, Mishchenko and Singh teach the computer-implemented method of claim 14
Mishchenko further teaches presenting the output to the user via the user interface (Col. 2, lines 22-30, In some embodiments, the one or more functions may be translated into one or more requests to executed by the at least one processor associated with the natural language model user interface. Embodiments may also include transmitting the response message to the user, wherein the transmitting may include summarizing payload contents received from the particular external API and displaying natural language text to the user via the natural language model user interface).
With regard to claim 19, Mishchenko teaches a system (Col. 2, lines 49-50, Embodiments of the present disclosure may include a system) comprising:
at least one storage device configured to store a code snippet in a first programming language (Col. 2, including at least one memory storing instructions.); and
circuitry communicatively coupled to the storage device (Col. 2, lines 51-54, Embodiments may also include at least one processor configured to execute the instructions to perform operations for integrating a particular external application programming interface (API) into a natural language model user interface), wherein the circuitry is configured to:
Claim 19 is a system having similar limitations to claim 1. Thus, claim 19 is rejected for the same rationale as applied to claim 1.
With regard to claim 20, Mishchenko teaches a non-transitory computer-readable medium comprising one or more computer-executable instructions that, when executed by circuitry of a computing device (Col. 3, lines 19-22, Embodiments of the present disclosure may also include a non-transitory computer-readable medium including instructions that may be executable by one or more processors to perform operations), cause the computing device to:
Claim 20 is a non-transitory computer-readable medium having similar limitations to claim 1. Thus, claim 19 is rejected for the same rationale as applied to claim 1.
Claims 7 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Mishchenko in view of Singh as applied to claim 1 and 3 above, and further in view of Keating et al. Pub. No. US 2026/0044679 A1 (hereinafter Keating).
With regard to claim 7, Mishchenko and Singh teach the computer-implemented method of claim 3
Mishcenko teaches handling a communications channel between the chatbot and the one or more APIs or services (Col. 6, lines 60-67, In some embodiments, a plugin causing the natural language model to be communicably coupled to a particular external API may be enabled (e.g., in response to a user input at the natural language model user interface). Enabling the plugin may initiate a plugin service that allows the natural language model user interface (or an associated program) to electronically communicate with a particular external API.)
However, Mishchenko and Singh do not explicitly teach a communications channel intermediary to reduce the size of a function all to comply with a size constraint of a context window of the chatbot.
Keating teaches hooking, in connection with the communication channel, a function call between the chatbot and the one or more APIs or services to reduce a size of the functional call ([0025], The context interface system 120 may operate to reduce the number of functions and corresponding function information that is passed to an LLM in order to increase the total number of functions usable by the context interface system 120.) by excluding one or more features of the function call to comply with a size constraint of a context window of the chatbot ([0027], Through storage 130, the context interface system 120 can retain metadata/information regarding particular function group(s) that are ‘in-context’ or, stated differently, are the current focus of a given user conversation, shown in FIG. 1, as ‘function groups in-context’ 132. This retained knowledge of the context interface system 100 allows a minimum set of function data/information to be passed to the LLM 150 based on the current context of a user conversation; [0031], To minimize context for the LLM 150 for interactions between the context interface system 120 and the LLM 150, context interface system 120 can pass a core set of prompt segments to the LLM 150 for each interaction (prompt interaction) with the LLM 150.); and
after having reduced the size of the function call, injecting the function call in to the context window of the chatbot ([0039], In this manner, the context interface 120 is to determine the prompt segments (and corresponding information) that are to be sent to the LLM 150 for each interaction with the LLM 150, while reducing the number of tokens being used in cases where many functions can be called by the LLM 150, and conversations may be tracked with users over long time periods.).
It would have been obvious to one of ordinary skill in the art at the time the invention was filed to apply the teachings of Mishchenko and Singh with the teachings of Keating in order to provide a method that teaches feature reduction of an associated function call in compliance with natural language processing context window. The motivation for applying Mishchenko and Singh teaching with Keating teaching is to provide a method that allows for reducing token usage such that enables improvements to accuracy and stability of output generated by a large language model without sacrificing the expansive set of capabilities (Keating, [0014]). Keating and Mishchenko and Singh are analogous art directed towards natural language processing and software engineering arrangement. Therefore, it would have been obvious for one of ordinary skill in the art to combine Mishchenko and Singh with Keating to teach the claimed invention in order to provide feature reduction of functions in association with context window of a natural language processing model.
With regard to claim 18, Mishchenko and Singh teach the computer-implemented method of claim 1
Keating further teaches modifying at least one feature of the code snippet to reduce a size of the input to comply with a size constraint of a context window of the chatbot ([0015], embodiments herein provide an architecture that facilitates reducing/erasing issues with consuming the entire token budget of an LLM and/or causing the LLM to hallucinate by replacing the simple, static function list that an 'out-of-the-box' LLM typically supports with a context interface system that, among other features as discussed for embodiments herein, provides for reducing the number of functions that are passed to an LLM for each interaction (prompt interaction) with the LLM.).
Rationale to claim 7 applied here.
It would be obvious to one of ordinary skill in the art to modify the at least one feature of the code snippet input into a chatbot with compliance to a context window to achieve the benefits of improvements to accuracy and stability of output generated by the chatbot, with reasonable expectation of success.
Claim 9 is rejected under 35 U.S.C. 103 as being unpatentable over Mishchenko in view of Singh as applied to claim 3 above, and further in view of Mullins et al. Patent No. US 11,741,140 B2 (hereinafter Mullins).
With regard to claim 9, Mishchenko and Singh teach the computer-implemented method of claim 3
Mullins teaches generating the input for the chatbot in the second programming language comprises inserting, within the input for the chatbot, scores that represent popularities of the one or more APIs or service relative to one another (Col. 8, lines 41-51, The processing may further comprise identifying one or more conversational bots that have one or more manifests or performing a skill corresponding to the identified specific skill and/or skill type. In examples where more than one conversational bot is identified as being capable of performing an identified specific skill from natural language input 106B, those conversational bots may be ranked. The ranking may comprise one or more metrics (e.g., popularity of website corresponding to conversational bot, popularity of the conversational bot by other users, history of using conversational bot by user 104).); and
injecting the input into the chatbot comprises influencing the chatbot to use a certain API or service included in the one or more APIs or services to answer a query based at least in part on the scores (Col. 15, lines 45-49, Natural language skill match engine 616 may perform operations associated with identifying relevant words and phrases from a natural language input provided to a conversational bot and matching those words and/or phrases to one or more highest ranked skills; Col. 18, lines 44-45, the skill is caused to be executed by the highest-ranked conversational bot.).
It would have been obvious to one of ordinary skill in the art at the time the invention was filed to apply the teachings of Mullin with the teachings of Mishchenko and Singh in order to provide a method that teaches popularity ranking and query of particular APIs to service a natural language processing request. The motivation for applying Mullin teaching with Mishchenko and Singh teaching is to provide a method that allows for expanded capabilities of a natural language processing model (Mullin, Col. 4). Mishchenko and Singh and Mullin are analogous art directed towards natural language processing and software engineering arrangement. Therefore, it would have been obvious for one of ordinary skill in the art to combine Mullin with Mishchenko and Singh to teach the claimed invention in order to provide API and service popularity rankings to service a natural language processing request.
Claim 11 is rejected under 35 U.S.C. 103 as being unpatentable over Mishchenko in view of Singh as applied to claim 3 above, and further in view of Microsoft Corporation “TypeChat Doc Pages Basic TypeScript Usage” (hereinafter TypeChat Usage).
With regard to claim 11, Mishchenko and Singh teach the computer-implemented method of claim 3
Mishchenko further teaches wherein the code snippet comprises a function call (Col. 9, lines 45-46, In some embodiments, the natural language model may be configured to generate a function call); and further comprising:
handling a communication channel between the chatbot and the one or more APIs or services (Col. 6, lines 60-67, In some embodiments, a plugin causing the natural language model to be communicably coupled to a particular external API may be enabled (e.g., in response to a user input at the natural language model user interface). Enabling the plugin may initiate a plugin service that allows the natural language model user interface (or an associated program) to electronically communicate with a particular external API.);
receiving the function call in connection with the communication channel (Col. 9, lines 45-53, In some embodiments, the natural language model may be configured to generate receive digital information, which may be used by the model to generate an output, consistent with disclosed embodiments. In some embodiments, the function call generated by the natural language model may be converted (e.g., by the natural language model or a program, application, script, or module) into an API call, which may be configured to be interpretable by the particular external API);
However, the combination does not explicitly teach function call failure detection or function call repair instruction.
Typechat Usage teaches detecting a failure in the function call by comparing the function call against a set of validators (Pg. 3, TypeChat describes types to language models to help guide their responses. In this case, we are using TypeScriptJsonValidator which uses the TypeScript compiler to validate data against a set of types); and
instructing the chatbot to modify the function call to fix the failure (Pg. 7, A translator takes both a model and a validator, and provides a way to translate some user input into objects following our schema. To do so, it crafts a prompt based on the schema, reaches out to the model, parses out JSON data, and attempts validation. Optionally, it will craft repair prompts and try if validation fails).
It would have been obvious to one of ordinary skill in the art at the time the invention was filed to apply the teachings of TypeChat Usage with the teachings of Mishchenko and Singh in order to provide a method that teaches natural language processing function call failure and repair process. The motivation for applying TypeChat Usage teaching with Mishchenko and Singh teaching is to provide a method that allows for a model to validate output data in association with a particular data schema, such that increase reliability of generated data (TypeChat Usage, Pg. 3). Mishchenko and Singh and TypeChat Usage are analogous art directed towards natural language processing. Therefore, it would have been obvious for one of ordinary skill in the art to combine TypeChat Usage with Mishchenko and Singh to teach the claimed invention in order to provide validator functionality to detect and repair faulty instructions to a natural language processing model to improve model output.
Claims 12-13 are rejected under 35 U.S.C. 103 as being unpatentable over Mishchenko in view of Singh in view of TypeChat FAQ as applied to claim 11 above, and further in view of Microsoft Corporation ““TypeChat Doc Pages Frequently Asked Questions (FAQ)” (hereinafter TypeChat FAQ).
With regard to claim 12, Mishchenko, Singh, and TypeChat FAQ teach the computer-implemented method of claim 11
Singh further teaches wherein generating the input for the chatbot in the second programming language comprises:
compiling the code snippet to an intermediary scripting language (Col. 2, lines 44-52, The generated sequence of outputs can be used to generate a natural language output that is descriptive of the base source code snippet. For example, the natural language output can be generated as a sequence of tokens, where each token in the sequence is selected based at least in part of the probability distribution for corresponding one of the outputs. For instance, the natural language output can be generated, token-by-token, by selecting the highest probability token for each output of the sequence.) with a declaration library;
…; and
converting the code snippet from the intermediary scripting language into the input for the chatbot (Col. 3, lines 21-24, The natural language output, generated by processing the base source code snippet using the base-to-NL neural network model, can then be processed using an NL-to-target neural network model to generate a sequence of outputs).
However, the combination does not explicitly teach an intermediary scripting language with a declaration library or generation of a set of validators from the declaration library.
TypeChat Usage teaches intermediary scripting language with a declaration library (Pg. 2, TypeChat uses TypeScript types as the “specification language” for response from language models. The approach for sending a request is minimal that includes the user’s inputs, your types, and text requesting the model to translate the user input into a JSON object in alignment with TypeScript types)
generating the set of the validators from the declaration library (Pg. 2, Once receiving an AI response, TypeChat uses the TypeScript compiler API under the hood to validate the data based on the types you provided. If validation fails, TypeChat sends a repair prompt back to the model that includes diagnostics from the TypeScript compiler.)
It would have been obvious to one of ordinary skill in the art at the time the invention was filed to apply the teachings of TypeChat FAQ with the teachings of Mishchenko, Singh, and TypeChat Usage in order to provide a method that teaches translation of a particular code snippet into an intermediary language associated with validators to be used input for a natural language processing model. The motivation for applying TypeChat FAQ teaching with Mishchenko, Singh, and TypeChat Usage teaching is to provide a method that allows for integration of a natural language interface into an application, such that the interface enables accurate, safe, and reliable natural language processing (TypeChat FAQ, Pg. 2-Pg. 3). Mishchenko, Singh, and TypeChat Usage and TypeChat FAQ are analogous art directed towards natural language processing. Therefore, it would have been obvious for one of ordinary skill in the art to combine TypeChat FAQ with Mishchenko, Singh, and TypeChat Usage to teach the claimed invention in order to provide generation of an intermediary scripting language and associated validators to enables an interface between application and natural language processing techniques.
With regard to claim 13, Mishchenko, Singh, TypeChat FAQ and TypeChat Usage teach the computer-implemented method of claim 12
Mishchenko further teaches wherein the code snippet is defined by a user (Col. 14, lines 27-32, System administrators, network administrators, software developers, engineers, and end-users are each a particular type of user. Automated agents, scripts, playback software and the like acting on behalf of one or more people may also constitute a user).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to IVAN A CASTANEDA whose telephone number is (571)272-0465. The examiner can normally be reached Monday-Friday 9:30AM-5:30PM EST.
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, Aimee Li can be reached at (571) 272-4169. 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.
/I.A.C./Examiner, Art Unit 2195
/Aimee Li/Supervisory Patent Examiner, Art Unit 2195
Prosecution Insights