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 .
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 .
1. Claims 1-20 are presented for examination.
Response to Amendment/Response to Arguments
2. 2.1 The rejection under 112 has been withdrawn since applicant’s amendments and remarks are persuasive and overcome the rejection.
2.2 The rejection under 103 has been withdrawn since applicant’s amendments and remarks are persuasive and overcome the rejection. Applicant argues the combination of Wang et al. (US 20190332680 A1) in view of Almaer et al. (US 20240362209 A1) fail to disclose the amended limitation of claim 1, 11 and 17. Applicant argument has been persuasive and overcome the rejection.
This amendment was added in response to the non-final rejection made by the Office. As a result, the previous rejection has been withdrawn, and a new rejection has been made in its place
Brown et al. (US 20240281600 A1) in view of (Almaer et al. (US 20240362209 A1).
Brown discloses wherein executing the computer program code performs the at least one of (i) modifying the stored content of the environment (Fig. 4, step 420, updated database to include successful generated query) or (ii) modifying the stored data attribute of the item of the environment, at the location within the environment ([0077],[0089], a user request to retrieve certain resources (e.g., data objects, field values of objects, etc.) from the endpoint 140), and send the computer-readable input to a large language model (LLM) to cause the LLM (Fig. 2, step 206, [0093], the computing system provides, to a large language model (LLM)) to generate, based on the context of the environment, a set of computer program code to perform the task at a location within the environment ([0031], [0077]-[0086], instruct an LLM to automatically generate queries which may be suitable for the endpoint based on a natural language statement of a user request ). In addition, please see the rejection below.
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 (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.
3. 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.
3.1 Claim(s) 1-3 and 5-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Almaer et al. (US 20240362209 A1) in view of Brown et al. (20240281600 A1).
Regarding claims 1, 11 and 17, Almaer discloses a non-transitory, computer-readable storage medium a ([0023], [0024], a non-transitory, processor-readable medium storing processor-executable instructions) comprising instructions recorded thereon, wherein the instructions, when executed by at least one data processor of a system ([0023], [0024], the computing system includes a processor and a memory coupled to the processor. The memory stores computer-executable instructions that, when executed by the processor), cause the system to ([0023], may cause the processor to):
receive a natural language instruction [0036], a user provides a first data request using natural language for an endpoint) to perform a task (Abstract, [0013], [0076],[0077], retrieval of data satisfying one or more criteria, the request including at least one data request parameter; searching a database storing example queries based on the request to identify at least one matching query ) within an environment ([0076], an endpoint 140 refers to a remote computing device that communicates with a network to which it is connected and may be a mobile device, a desktop computer, a virtual machine, an embedded device, a server, or the like) that is communicatively coupled to the system (Fig. 1, [0075], the system 100),
wherein the task includes at least one of (i) modifying stored content of the environment ([Fig. 4, step 420, update databased to included successful generated query or (ii) modifying a stored data attribute of an item of the environment ([0077],[0089], a user request to retrieve certain resources (e.g., data objects, field values of objects, etc.) from the endpoint 140),
generate a computer-readable input based on the received instruction ([0076], the code generating engine 114 performs auto-generation of source code and configured to convert user requests that are expressed in natural language to code for queries corresponding to the user requests).
wherein the computer-readable input (code generating engine 114) includes a context of the environment ([0077], context of e-commerce, the requested data may relate to products, collections, customers, orders, carts, checkouts, and other store resources that can be used to build a custom purchasing experience for an online store) and a computer-readable form of the received instruction ([0077], [0078], Upon processing a data request from a user, the code generation engine 114 may generate a query that corresponds to the data request for the endpoint);
send the computer-readable input to a large language model (LLM) to cause the LLM (Fig. 2, step 206, [0093], the computing system provides, to a large language model (LLM)) to generate, based on the context of the environment, a set of computer program code to perform the task at a location within the environment ([0031], [0077]-[0086], instruct an LLM to automatically generate queries which may be suitable for the endpoint based on a natural language statement of a user request ); and
execute, by the at least one data processor, the computer program code received from the LLM to perform the task in the environment (Fig 2, step 208, Fig. 3, step 314, Fig. 4, steps 410-420, [0020], [0034], [0036]-[0040], [0107], updating the database by including the generated query; process for automatically generating queries for an endpoint that are based on user requests and that comply with the requirements of the endpoint),
wherein executing the computer program code performs the at least one of (i) modifying the stored content of the environment (Fig. 4, step 420, updated database to include successful generated query) or (ii) modifying the stored data attribute of the item of the environment, at the location within the environment ([0077],[0089], a user request to retrieve certain resources (e.g., data objects, field values of objects, etc.) from the endpoint 140).
However, Almaer fails to disclose a set of computer program code to perform the task at a location within the environment; and execute, by the at least one hardware processor, the computer program code received from the LLM to perform the task in the environment.
Brown discloses a set of computer program code to perform the task at a location within the environment (Fig. 4; [0094][0105][0106]; provides, to the LLM, a further input prompt for instructing the LLM to generate a revised query, the input prompt may include the error data associated with the generated query (i.e. providing a recommended modification to the first command. A trained ML model may be fine-tuned, meaning that the values of the learned parameters may be adjusted slightly in order for the ML model to better model a specific task); and
execute, by the at least one hardware processor, the computer program code received from the LLM to perform the task in the environment (Fig. 4, [0094][0105][0106]; the requesting user may then modify the generated query, incorporate the query into source code, or otherwise manipulate the code of the generated query; provides, to the LLM, a further input prompt for instructing the LLM to generate a revised query, the input prompt may include the error data associated with the generated query, that is, a representation of the error data may be inserted in an input prompt to the LLM with instructions to generate a new query corresponding to the first user request).
Brown and Almaer are analogous art. They relate to a language mode program. Therefore, before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to modify a large language model taught by Almaer, incorporated with executing natural language programs, taught by Brown, in order to improved techniques of generating code for interacting with an endpoint, and automatically-generated queries using an LLM.
Regarding claims 2, 12 and 18, Almaer discloses the task is non-deterministic (Fig. 4, [0104]-[0105], when the error response provided to LLM another input to generate revised query), and wherein causing the LLM to generate the set of computer program code ([0002], [0030], that leverage use of large language models (LLMs) for generating source code), comprises:
instructing the LLM to output computer program code ([0031], the programmer may instruct an LLM to automatically generate queries which may be suitable for the endpoint based on a natural language statement of a user request) that includes:
a prompt configured to cause the LLM to produce a task result for the task (Fig. 3,[0033], the LLM may not be capable of generating queries that are suitable for a variety of user request), and
code that when executed by the at least one data processor of the system, causes the system to send the prompt to the LLM (Fig. 3, [0031]-[0033], code generation systems may use hard-coded prompt templates when forming input prompts to an LLM for instructing the LLM to generate a query);
wherein executing the computer program code to perform the task comprises executing the code to send the prompt to the LLM ([0030]-[0033], [0035], generating code for interacting with an endpoint. A system and methods for producing automatically generated queries using an LLM).
Regarding claims 3, 7, 13, 15 and 20, Almaer discloses the environment is a structured digital environment ([0079], store a database which the query is generated), wherein the natural language instruction includes a request to write the task result to the location in the structured digital environment ([0036]-[0040][0107], in response to a user request, the system may return a cached query corresponding to the previous data request (i.e. accessing a stored query); the cached query is the same type of as the current user request (requesting to observe a property); an endpoint executes the query and provides a response (if execution is successful or failure); a success response includes field values (context parameters)), wherein causing the LLM to generate the set of the computer program code ([0031], [0032], an LLM to automatically generate queries which may be suitable for the endpoint based on a natural language statement of a user request) comprise:
prompting the LLM to generate code that is configured to use the context of the environment to identify the location within the structured digital environment ([0031], [0032], [0023], an LLM to automatically generate queries which may be suitable for the endpoint based on a natural language statement of a user request; the user request includes at least one request parameter (context)).
prompting the LLM to generate code that when executed causes the at least one data processor to write task result to the identified location (par. [0013], [0023], [0057], providing, to a large language model (LLM), an input prompt to generate a query purporting to retrieve data satisfying the one or more criteria, the input prompt including the at least one data request parameter and the at least one matching query as an example; and receiving, from the LLM, a result including the generated query).
Regarding claim 5 and 8, Brown discloses the environment comprises a chat thread (Fig. 9A-[9G [0075], the user is requesting the system to determine if a number is divisible by another number), and wherein causing the natural language to generate the set of computer program code (Fig. 4, Fig. 5, Fig. 9B, [0075], writing the computer code based on the user request) comprises:
prompting the natural language (Fig. 2, [0043], [0054], natural language commands, user command by a speaker for voice communication with the use) to generate code (Fig. 4, [0059]-[0060], change the natural language commands to computer code, computer language like javascript) that when executed causes the at least one data processor to write a specified value and the task result to the chat thread (Fig. 3-Fig. 12, At 1202, the software is run in correspondence with an appropriate processing device, such as a mobile device, smart assistant, or personal computer. The software comprises a natural language processor as described earlier in this document. At 1204, a user input is received that includes a request to perform a function or task. The user input comprises any suitable type of input. For example, the user input may correspond to a natural language input that is either spoken or typed into the system. At 1206, a determination is made whether an edge case or exception is identified for the requested functionality).
Regarding claim 6, Brown discloses instructing the LLM to output the prompt comprises instructing the LLM to generate JavaScript instructions that include the prompt ([0052], [0067], [0202], [0213], Code Generation using LLMs. an initial prompt for the LLM. Prompt engineering is performed to take the user's goal and express that goal as a prompt for the LLM to accomplish that goal using computer programming languages such as, for example, Javascript, Python, Java).
Regarding claim 9 and 16, Brown discloses executing the computer program code causes the system to observe a value (Fig. 3, Fig. 4, Fig. 13C) or a state of the environment ([0057], FIG. 3, Fig. 9A-Fig. 9G, proceeds by analyzing the words and symbols in a natural language statement and the processing receives an abstract syntax tree (AST) and run the program to evaluate the result).
Regarding claim 10, Brown discloses generate a transcript including the computer-readable input (Fig. 9F, computer readable program) and the set of computer program code to perform the task (Fig. 9B, Fig. 10, execute new logic to perform desired functionality).
Regarding claims 14 and 19, the combination of Wang and Almaer disclose:
Wang discloses causing the natural language to generate the set of computer program code to perform the task comprises:
causing the natural language to generate a first set of computer program code, comprising a first type of code, to perform a first task (Fig. 25, [0374], The NLU 2514 can generate an intent from the translated text, and pass the intent to a reasoner 2516. The reasoner 2516 analyzes the intent and determines a task to perform in response to the intent. The reasoner 2516 can further initiate execution of the task, and determine an action to perform with the results of the task); and
Almaer discloses causing the LLM to generate a second set of computer program code, comprising a second type of code different from the first type, to perform a second task (Abstract, [0013], [0023], [0060], [0106], Fig. 3, Fig. 4, the generated query; receiving, from the endpoint, a response indicating an error associated with the generated query; and providing, to the LLM, a further input prompt for generating a revised query, the further input prompt including error data associated with the error).
3.2 Claim(s) 4 is/are rejected under 35 U.S.C. 103 as being unpatentable over Almaer et al. (US 20240362209 A1) in view of Brown et al. (20240281600 A1) further in view of Wang et al. (US 20190332680 A1).
Regarding claim 4, the combination of Almaer and Brown disclose the limitations of claim 1 but fails to disclose the limitations of claim 4. However, Wang discloses the code that when executed causes the at least one processor to write the task result to the identified location includes extensible markup language (XML) ([0355], [0393], The input intent 2440 can be expressed as a set of objects, actions, and/or parameters in a data structure. As one example, the input intent 2440 can be formatted using Extensible Markup Language (XML)).
Brown and Almaer are analogous art. They relate to a language mode program. Therefore, before the effective filing date of the claimed invention, it would have been obvious to a person of ordinary skill in the art to modify the computer program product for multi-lingual device, taught by Wang, incorporated with teaching of Brown and Almaer, as stated above, in order to capable of receiving verbal input in multiple languages, and further capable of providing conversational responses in multiple languages.
Citation Pertinent prior art
4. The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Gillman (US 20240028312 A1) discloses the natural language description of user-requested data transformation task for execution with a subset of the first output is received (510) by the machine learning engine. A large language model is directed (512) to identify archetype of user-requested data transformation task. The user-requested data transformation task is applied (514) to a subset of the first output using the archetype to generate second output. The second output is displayed (516) by the machine learning engine.
Cuomo (US 20250086310 A1) discloses a method for preserving user privacy in prompt data i.e. natural language text, by a large language model privacy preservation system for different purposes e.g. personalized content recommendations, quality control, refined model training, optimized system resources, and research in natural language processing and sentiment analysis. Can also be used in contextual data, and user input.
Watson et al. (US 20240319970 A1) discloses the code generation module 302 executes a process for using a large language model to generate executable code in a manner that preserves privacy and confidentiality of proprietary data. An exemplary process for using a using a large language model to generate executable code in a manner that preserves privacy and confidentiality of proprietary data is generally indicated at flowchart 400 in FIG. 4.
Bischo et al. (US 20240256588 A1) discloses a natural language query from a user into a cell of a notebook environment, the natural language query performed with respect to a data warehouse, the data warehouse modeled in a data warehouse graph; in response to receiving the natural language query: determining, using directed edges of a notebook graph structure, a set of precedent cells from which the cell depends, the notebook graph structure being a directed acyclic graph; and determining, using edges of a user graph structure populated based on activities of users relative to the notebook.
Conclusion
5. 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.
6. inquiry concerning this communication or earlier communications from the examiner should be directed to Kidest Worku, whose telephone number is 571-272-3737. Examiner can be reached Mon-Fir (9am-5pm ET).
If attempts to reach the examiner by telephone are unsuccessful, the examiner's supervisor, Ali Mohammad, can be reached on 571-272-4105. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Examiner interviews are available via telephone 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.
Information regarding the status of an application may be obtained from the Patent Application information Retrieval IPAIRI system. Status information for published applications may be obtained from either Private PMR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAG system, contact the Electronic Business Center (EBC) at 866-217 - 9197. If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/KIDEST WORKU/Primary Examiner, Art Unit 2119
Prosecution Insights