DETAILED ACTION
Remarks
Applicant presents a communication filed 26 June 2026 responsive to the 25 March 2026 non-final Office action (the “Previous Action”).
Claims 1, 7, 10-11 and 16 are amended, as well as Figure 9 of the drawings.
Claims 1-20 are pending. Claims 1, 11 and 16 are the independent claims.
Any unpersuasive arguments are addressed in the “Response to Arguments” section below.
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 .
Examiner Notes
Examiner cites particular columns, paragraphs, figures and line numbers in the references as applied to the claims below for the convenience of the applicant. Although the specified citations are representative of the teachings in the art and are applied to the specific limitations within the individual claim, other passages and figures may apply as well. It is respectfully requested that, in preparing responses, the applicant fully consider the references in their entirety as potentially teaching all or part of the claimed invention, as well as the context of the passage as taught by the prior art or disclosed by the examiner.
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
Response to Arguments
Applicant’s arguments are moot in view of the new ground(s) of rejection herein, necessitated by Applicant’s amendments.
Drawings
The Previous Action’s objection to the drawings is withdrawn in view of Applicant’s amendments to Figure 9.
Claim Objections
The Previous Action’s claim objections are withdrawn in view of Applicant’s claim amendments.
Claims 11-20 are objected to for the following informalities:
Claim 11 refers to “structure” test case data in the second to last line of the claim, which appears to be a typographical error that should perhaps read -structured- instead.
Claim 16 includes the same typographical error as claim 11.
Claims 12-15 depend on the objected claim 11 and inherit the same issue.
Claims 17-20 depend on the objected claim 16 and inherit the same issue.
Claim Rejections - 35 USC § 112
The Previous Action’s claim § 112 rejections are withdrawn in view of Applicant’s claim amendments.
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, 11-13, 16 and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Nabeel et al. “Test Code Generation for Telecom Software using a Two-Stage Generative Model” (art of record – hereinafter Nabeel) in view of Maturana et al. (US 2025/0147736) (art of record – hereinafter Maturana) in view of Wen et al. (CN 112966095) (art made of record – hereinafter Wen).
NOTE: Wen is originally in Chinese. Citations herein refer to the machine translation of Wen in the file record.
As to claim 1, Nabeel discloses a system, comprising:
receiving, from database equipment, structured test case data of a collection of structured test case data; (e.g., Nabeel, p. 3 Sec. IV par. 2: given test description written in natural language [structured test case data], the pipeline should generate a test script with test cases [the test description must be stored in some physical storage, i.e., database equipment, and retrieved from that storage to be used]; p. 3 left col. Sec. B last par.: we use test descriptions written in natural language in each test as input for our method)
transforming each structured test case data comprising the collection of structured test case data into a generic executable programming language code representation of the structured test case data, wherein the transforming comprises using a generative artificial intelligence model developed based on a large language model (e.g., Nabeel, p. 4 Figs. 1, 2 and associated text, p. 3 right col. last par.: generating a test script using a prompt “(human language that describes the intention behind the test)”. Then, the LLM is used to generate a test script and test case [see figures, a prompt comprising the test description is input into a Generative LLM and code is produced]; p. 5 Fig. 5 and associated text [see figure, the code and tests cases are generated as Python code, i.e., generic executable programming language code]) that has been trained (e.g., Nabeel, p. 5 left col. par. 1: for test code generation, we evaluate the following LLMs: 1) Mistral-7B trained on the general corpus), wherein the regenerative artificial model uses a process to transform each structured test case data to the generic executable programming language code representation (see above)
executing the generic executable programming language code using a machine language operations pipeline (e.g., Nabeel, p. 4 left col. Sec. C Content Score: Python successfully executed [executing Python requires interpreting it with an interpreter (machine language operations pipeline) that converts it to machine code]) to ensure that network equipment complies with a specified networking protocol standard (e.g., Nabeel, p. 1 right col. par. 2: testing and integration for multivendor interoperability in O-RAN [open radio access network, O-RAN being a networking protocol standard]; p. 1 abstract par. 1: we propose a framework for Automated Test Generation for large-scale Telecom Software system [the device(s) running the software being equipment]).
Nabeel does not explicitly disclose: at least one processor; and at least one memory that stores executable instructions that, when executed by the at least one processor, facilitate performance of the operations, or a large language model that has been trained using a representative sample collection of the structured test case data, and wherein the representative sample collection of the structured test case data is based on feedback data associated with a previous iterative execution of the generative artificial intelligence model; us[ing] a word to vector process, and wherein the word to vector process identifies a cosine similarity representative of a translation confidence value between a first vector value representing each structured test case data and a second vector representing the generic executable programming language code.
However, in an analogous art, Maturana discloses:
at least one processor; (e.g., Maturana par. [0109]]) and
at least one memory that stores executable instructions that, when executed by the at least one processor, facilitate performance of [the] operations (e.g., Maturana, par. [0109])
a large language model that has been trained using a representative sample collection of the structured data, and wherein the representative sample collection of the structured data is based on feedback data associated with a previous iterative execution of the generative artificial intelligence model; (e.g., Maturana, par. [0069]: a code generation user interface that allows a user to enter, as a natural language input, functional descriptions 802 [structured data] of the code to be generated; par. [0071]: the AI component 210 can train the generative AI model 226 using the code in repository 306 to respond to users’ natural language requests with code that meets the requirements of the users’ functional description 802; par. [0062]: the model 226 can be a large language model; par. [0081]: the generated code 804 itself [feedback data] can be stored in the code repository 306 and the system 202 can create a link [feedback data] between the code 804 stored in the code repository 306 and its corresponding functional description 1002 [structured data] stored in the document repository 304. This data can be used to refine the training of the generative AI model; par. [0073]: over time, the system can generate new content to the repositories 304 and 306, functional descriptions 802 [structured data] and the resulting documented control code [feedback data] that was generated based on those descriptions. These types of data can be stored in the repositories so that, when users submit descriptions 802 similar to previous submitted descriptions, the model 226 will recognize these similarities [i.e., it has been trained based on them] and generate code 804 having a high confidence of accuracy based on the code 804 that was generated for previous similar descriptions 802).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the operations of Nabeel such that at least one memory that stores executable instructions that, when executed by the at least one processor, facilitates performance of the operation, as taught by Maturana, as Maturana would provide the advantage of a means of performing the operations automatically on a general-purpose computer. (See Maturana, par. [0109]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the generation of programming language code from structured test case data using a trained large language model taught by Nabeel such that the large language model is trained using a sample collection of structure data based on feedback data associated with a previous generation of the model, as taught by Maturana, as Maturana would provide the advantage of a means of generating code having a high confidence of accuracy. (See Maturana, par. [0073]).
Further, in an analogous art, Wen discloses:
us[ing] a word to vector process, and wherein the word to vector process identifies a cosine similarity representative of a translation confidence value between a first vector value representing each structured data and a second vector representing the executable programming language code and wherein the word to vector process identifies a cosine similarity representative of a translation confidence value between a first vector value representing each structured data and a second vector representing the executable code (e.g., Wen, par. [0125]: a natural language query consists of a sequence of n words; par. [0058]: after the developer inputs a query [structured data], the description embedding representation module in the JEAN model embeds the query into a query vector; par. [0008]: a dataset of JAVA code snippets; par. [0009]: use the code embedding representation module in the JEAN model to embed all code segments [executable programming language code] in the codebase into code vectors; par. [0063]: after obtaining the code embedding vector P and the query vector Q, cosine similarity is used to measure the similarity between the two vectors; par. [0065]: the sim(P,Q) value [the cosine similarity, see the formula in par. [0064]: the higher the relevance [translation confidence] between the code and the natural language query. Then, the code segment of the vector that is the most relevant to the query vector is returned).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the generative artificial model performing a process of translating structured test case data to a generic executable programming language code representation of Nabeel, such that it comprises a word to vector process by identifying a cosine similarity representative of a translation confidence value between a first vector value representing each structured data being translated and a second vector representing the code, as taught by Wen, as Wen would provide the advantage of a means of providing more relevant translations. (See Wen, par. [0065]).
As to claim 2, Nabeel/Maturana/Wen discloses the system of claim 1 (see rejection of claim 1 above), Nabeel further discloses:
wherein the specified network protocol standard is an open radio access network protocol (e.g., Nabeel, p. 1 right col. par. 2: testing and integration for multivendor interoperability in O-RAN [open radio access network, O-RAN being a networking protocol standard and comprises a protocol]).
As to claim 5, Nabeel/Maturana/Wen discloses the system of claim 1 (see rejection of claim 1 above), Nabeel further discloses:
wherein the executable programming language code is a conversion of each of the structured test case data represented in text format to a high-level general-purpose programming language representation of each of the structured test case data (e.g., Nabeel, p. 4 Figs. 1, 2 and associated text, p. 3 right col. last par.” generating a test script using a prompt “(human language that describes the intention behind the test)”. Then, the LLM is used to generate a test script and test case [see figures, a prompt comprising the test description is input into a Generative LLM and code is produced]; p. 5 Fig. 5 and associated text [see figure, the code and tests cases are generated as Python code, i.e., generic executable programming language code]).
As to claim 11, Nabeel discloses a method, comprising:
transforming, by a device, each structured test case data of a collection of structured test case data into a generic executable programming language code representation of the structured test case data, (e.g., Nabeel, p. 4 Figs. 1, 2 and associated text, p. 3 right col. last par.: generating a test script using a prompt “(human language that describes the intention behind the test)”. Then, the LLM is used to generate a test script and test case [see figures, a prompt comprising the test description is input into a Generative LLM and code is produced]; p. 5 Fig. 5 and associated text [see figure, the code and tests cases are generated as Python code, i.e., generic executable programming language code]; p. 3 left col. Sec. B last par.: we use test descriptions written in natural language in each test as input for our method) wherein each structured test case data of the collection of structured test case data is received from a database ([the test descriptions must be stored in some physical storage, i.e., database equipment, and retrieved from that storage to be used]) and
executing, by the device, the generic executable programming language code using a machine language operations pipeline (e.g., Nabeel, p. 4 left col. Sec. C Content Score: Python successfully executed [executing Python requires interpreting it with an interpreter (machine language operations pipeline) that converts it to machine code]) to ensure that network equipment representing a testing framework of networking equipment complies with a specified networking protocol standard (e.g., Nabeel, p. 1 right col. par. 2: testing and integration for multivendor interoperability in O-RAN [open radio access network, O-RAN being a networking protocol standard]; p. 1 abstract par. 1: we propose a framework for Automated Test Generation for large-scale Telecom Software system [the device(s) running the software being equipment, it represents a testing framework because a testing framework is merely a testing environment per par. [0040] of the specification]).
Nabeel does not explicitly disclose a device comprising one or more processor or a database from a group of databases; or wherein the generic executable programming language code is determined to be representative of the structured test case data based on an application of a trained word to vector model that translates each structured test case data to the generic executable programming language code based on a cosine similarity between a first vector value representative of the generic executable programming language code and a second vector value representative of each structure test case data.
However, in an analogous art, Maturana discloses:
a device comprising one or more processor (e.g., Maturana par. [0109]])
a database from a group of databases (e.g., Maturana, par. [0095]: repositories; par. [0133]: data store(s)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the device of Nabeel such that it comprises one or more processors, as taught by Maturana, as Maturana would provide the advantage of a means of performing the operations automatically on a general-purpose computer. (See Maturana, par. [0109]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the database of Nabeel such that it comprises one or multiple databases, as taught by Maturana, as Maturana would provide the advantage of a means of storing data in different locations. (See Maturana, pars. [0133]). It also would have been obvious to try and store the data in a system comprising multiple databases, as Maturana shows that multiple databases can be used in place of a single database and predictably achieve the same result. (See M.P.E.P. § 2143(I)(B)).
Further, in an analogous art, Wen discloses:
wherein the executable programming language code is determined to be representative of the structured data based on an application of a trained word to vector model that translates each structured data to the executable programming language code based on a cosine similarity between a first vector value representative of the executable programming language code and a second vector value representative of each structure data (e.g., Wen, par. [0125]: a natural language query consists of a sequence of n words; par. [0058]: after the developer inputs a query [structured data], the description embedding representation module in the JEAN model embeds the query into a query vector; par. [0008]: a dataset of JAVA code snippets; par. [0008]: training the JEAN model; par. [0009]: use the code embedding representation module in the JEAN model to embed all code segments [executable programming language code] in the codebase into code vectors; par. [0063]: after obtaining the code embedding vector P and the query vector Q, cosine similarity is used to measure the similarity between the two vectors; par. [0065]: the sim(P,Q) value [the cosine similarity, see the formula in par. [0064]: the higher the relevance [translation confidence] between the code and the natural language query. Then, the code segment of the vector that is the most relevant to the query vector is returned [note that Wu is construed as translating in the sense that a natural language form is input and an executable programming language form is output]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the translation of structured test case data to a generic executable programming language code representation taught by Nabeel, such that the executable programming language code is determined to be representative of the structured data based on an application of a trained word to vector model that translates each structured data to the executable programming language code based on a cosine similarity between a first vector value representative of the executable programming language code and a second vector value representative of each structure data, as taught by Wen, as Wen would provide the advantage of a means of providing more relevant translations. (See Wen, par. [0065]).
As to claim 12, Nabeel/Maturana/Wen discloses the method of claim 11 (see rejection of claim 11 above), Nabeel further discloses structured test case data (see rejection of claim 11 above) but Nabeel does not explicitly disclose wherein the transforming comprises using, by the device, a generative artificial intelligence model developed based on a large language model that has been trained using a representative sample collection of the structured test case data.
However, in analogous art, Maturana discloses:
wherein the transforming comprises using, by the device, a generative artificial intelligence model developed based on a large language model that has been trained using a representative sample collection of the structured data (e.g., Maturana, par. [0071]: the AI component 210 can train the generative AI model 226 using the code in repository 306 to respond to users’ natural language requests with code that meets the requirements of the users’ functional description 802 [transform those requests into code]; par. [0057]: descriptions of functional descriptions submitted as natural language prompts; par. [0086]: prompts stored in the system’s prompt repository 308 [representative sample collection of the structure data], which are used to train the model 226; par. [0085]: the model 226 can include a large language model).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the transforming of structured test case data by a trained model of Nabeel such that the transforming comprises using, a generative artificial intelligence model developed based on a large language model that has been trained using a representative sample collection of the structured data, as taught by Maturana, as Maturana would provide the advantage of a means for the model to respond to user’s request with suitable code that meets the requirements of the structured data. (See Maturana, par. [0071]).
As to claim 13, Nabeel/Maturana/Wen discloses the method of claim 12 (see rejection of claim 12 above), Nabeel further discloses structured test case data (see rejection of claim 11 above) but does not explicitly disclose wherein the representative sample collection of the structured test case data is based on feedback data associated with a previous iterative execution of the generative artificial intelligence model.
However, in an analogous art, Maturana discloses:
wherein the representative sample collection of the structured data is based on feedback data associated with a previous iterative execution of the generative artificial intelligence model (e.g., Maturana, par. [0071]: the AI component 210 can train the generative AI model 226 using the code in repository 306 to respond to users’ natural language requests with code that meets the requirements of the users’ functional description 802; par. [0081]: the generated code can be stored in the code repository 306 in association with the prompts and user responses “(e.g., the functional description 802)” [representative sample collection of structured data] used to generate the code 804, and the system can create a link between the code 804 in the repository 306 and its corresponding functional description. This data can be used to refine the training of the generative AI model).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the transforming of structured test case data by a trained model of Nabeel such that the transforming comprises using, a generative artificial intelligence model developed based on a large language model that has been trained using a representative sample collection of the structured data based on feedback data associated with a previous iterative execution of the generative artificial intelligence model, as taught by Maturana, as Maturana would provide the advantage of a means of generating code with a high confidence of accuracy. (See Maturana, par. [0073]).
As to claim 16, Nabeel discloses operations, comprising:
in response to receiving, from database equipment , structured test case data of a group of structured test case data, transforming the structured test case data of a collection of structured test case data into a generic executable programming language code representative of the structured test case data; (e.g., Nabeel, p. 3 Sec. IV par. 2: given test description written in natural language [structured test case data], the pipeline should generate a test script with test cases [the test description must be stored in some physical storage, i.e., database equipment, and retrieved from that storage to be used]; p. 3 left col. Sec. B last par.: we use test descriptions written in natural language in each test as input for our method) and
executing the generic executable programming language code using a machine language operations pipeline (e.g., Nabeel, p. 4 left col. Sec. C Content Score: Python successfully executed [executing Python requires interpreting it with an interpreter (machine language operations pipeline) that converts it to machine code]) to ensure that network equipment represented in a testing framework of networking equipment complies with a specified networking protocol standard (e.g., Nabeel, p. 1 right col. par. 2: testing and integration for multivendor interoperability in O-RAN [open radio access network, O-RAN being a networking protocol standard]; p. 1 abstract par. 1: we propose a framework for Automated Test Generation for large-scale Telecom Software system [the device(s) running the software being equipment, it represents a testing framework because a testing framework is merely a testing environment per par. [0040] of the specification]).
Nabeel does not explicitly disclose: a non-transitory machine-readable medium comprising instructions that, in response to execution, cause a system comprising at least one processor to perform the operations or database equipment of a group of database equipment, or wherein the transforming further comprises using a developed word to vector model that translates the structured test case data to the generic executable programming language code based on a cosine similarity between a first vector value representative of the generic executable programming language code and a second vector value representative of the structure test case data.
However, in an analogous art, Maturana discloses:
a non-transitory machine-readable medium comprising instructions that, in response to execution, cause a system comprising at least one processor to perform operations; (e.g., Maturana par. [0109]]) and
database equipment of a group of database equipment (e.g., Maturana, par. [0095]: repositories; par. [0133]: data store(s)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the operations of Nabeel such that a non-transitory machine-readable medium comprising instructions that, in response to execution, cause a system comprising at least one processor to perform them, as taught by Maturana, as Maturana would provide the advantage of a means of performing the operations automatically on a general-purpose computer. (See Maturana, par. [0109]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the database equipment of Nabeel such that it comprises a group of database equipment, as taught by Maturana, as Maturana would provide the advantage of a means of storing data in different locations. (See Maturana, pars. [0133]). It also would have been obvious to try and store the data in a system comprising a group of database equipment, as Maturana shows that a group of database equipment can be used in place of a single piece of database equipment and predictably achieve the same result. (See M.P.E.P. § 2143(I)(B)).
Further, in an analogous art, Wen discloses:
wherein the transforming further comprises using a developed word to vector model that translates the structured data to the executable programming language code based on a cosine similarity between a first vector value representative of the executable programming language code and a second vector value representative of the structure data (e.g., Wen, par. [0125]: a natural language query consists of a sequence of n words; par. [0058]: after the developer inputs a query [structured data], the description embedding representation module in the JEAN model embeds the query into a query vector; par. [0008]: a dataset of JAVA code snippets; par. [0009]: use the code embedding representation module in the JEAN model to embed all code segments [executable programming language code] in the codebase into code vectors; par. [0063]: after obtaining the code embedding vector P and the query vector Q, cosine similarity is used to measure the similarity between the two vectors; par. [0065]: the sim(P,Q) value [the cosine similarity, see the formula in par. [0064]: the higher the relevance [translation confidence] between the code and the natural language query. Then, the code segment of the vector that is the most relevant to the query vector is returned [note that Wu is construed as translating in the sense that a natural language form is input and an executable programming language form is output]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the translation of structured test case data to a generic executable programming language code representation taught by Nabeel, such that the transforming comprises using a developed word to vector model that translates the structured data to the executable programming language code based on a cosine similarity between a first vector value representative of the executable programming language code and a second vector value representative of the structure data, as taught by Wen, as Wen would provide the advantage of a means of providing more relevant translations. (See Wen, par. [0065]).
As to claim 19, it is a medium claim having limitations substantially the same as those of claim 5. Accordingly, it is rejected for substantially the same reasons.
Claims 3 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Nabeel (“Test Code Generation for Telecom Software using a Two-Stage Generative Model”) in view of Maturana (US 2025/0147736) in view of Wen (CN 112966095) in further view of the O-RAN E2 Interface Test Specification 2.0, (art of record – here O-RAN).
As to claim 3, Nabeel/Maturana/Wen discloses the system of claim 1 (see rejection of claim 1 above), but does not explicitly disclose wherein the structured test case data is text data that is formatted according to a formatting standard comprising a text title of a function associated with a text description of the function, and wherein the text description outlines a group of functional acts necessary to ensure compliance of the network equipment to operate within an open radio access network protocol communication system.
However, in an analogous art, O-RAN discloses:
wherein the structured test case data is text data that is formatted according to a formatting standard comprising a text title of a function (e.g., O-Ran, p. 19: “RIC Service Update Procedure with RAN Function modified (positive case)”) associated with a text description of the function, and wherein the text description outlines a group of functional acts (e.g., O-RAN, p. 19: “5.2.1.4.1.3.2 Procedure Step 1. Initiate the RIC Service Update Procedure…Step 2. At the test Simulator (E2 Node), the following received and transmitted E2 messages are recorded) necessary to ensure compliance of the network equipment to operate within an open radio access network protocol communication system (e.g., O-RAN, p. 11 Sec. 4.3.2.1: Device Under Test (E2 Node) The DUT could be any E2 Node; p. 19 Sec. 5.2.1.4.2.1: the purpose of this test case is to test the RIC Service Update Procedure of the DUT as specified in “O-RAN WG3: E3 Application Protocol”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the structured test data of Nabeel to such that the structured test case data is text data that is formatted according to a formatting standard comprising a text title of a function associated with a text description of the function, and wherein the text description outlines a group of functional acts necessary to ensure compliance of the network equipment to operate within an open radio access network protocol communication system, as taught by O-RAN, as O-Ran would provide the advantage of a means of ensuring the device complies with the O-RAN standard, as suggested by Nabeel. (See O-RAN, p. 19 Sec. 5.2.1.4.2.1, Nabeel at p. 3 left col. Sec. B item 2).
As to claim 17, it is a medium claim having limitations substantially the same as those of claim 3. Accordingly, it is rejected for substantially the same reasons.
Claims 4, 9-10 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Nabeel (“Test Code Generation for Telecom Software using a Two-Stage Generative Model”) in view of Maturana (US 2025/0147736) in view of Wen (CN 112966095) in further view of Kim (US 2018/0181456) (art of record – hereinafter Kim).
As to claim 4, Nabeel/Maturana/Wen discloses the system of claim 1 (see rejection of claim 1) but does not explicitly disclose wherein the network equipment comprises at least one of base station equipment, internet of things equipment, a user equipment, or satellite based equipment.
However, in an analogous art, Kim discloses:
wherein the network equipment comprises at least one of base station equipment, internet of things equipment, a user equipment, or satellite based equipment (e.g., Kim, par. [0002]: an IoT framework that performs a test on IoT software and/or hardware; par. [0021]: the IoT hardware may include an IoT device).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the testing of Nabeel such that the network equipment under test comprises internet of things equipment, as taught by Kim, as Kim would provide the advantage of a means of ensuring such equipment functions as expected. (See Kim, par. [0003]).
As to claim 9, Nabeel/Maturana/Wen discloses the system of claim 1 (see rejection of claim 1 above), and further discloses compliance with the specified networking protocol standard (see rejection of claim 1 above) but does not explicitly disclose wherein the network equipment is first network equipment, and wherein compliance with the specified networking protocol standard comprises determining that the first network equipment is interoperable with second network equipment.
However, in an analogous art, Kim discloses:
wherein the network equipment is first network equipment, and wherein compliance with the specified standard comprises determining that the first network equipment is interoperable with second network equipment (e.g., Kim, par. [0023]: the test case may include an interoperability test. The interoperability test may define test cases to test and end-to-end function between two different IoT devices developed based on the same standard “(e.g. the IoT devices manufactured by different manufacturers)”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the determination of compliance with a specified networking protocol standard of Nabeel such that determining compliance with the standard includes determining that the first network equipment is interoperable with second network equipment, as taught by Kim, as Kim would provide the advantage of a means of ensuring multivendor interoperability, as suggested by Nabeel. (See Nabeel, p. 1 right col. second to last par.).
As to claim 10, Nabeel/Maturana/Wen/Kim discloses the system of claim 9 (see rejection of claim 9 above) but does not explicitly disclose wherein the first network equipment is manufactured by a first manufacturing entity and the second network equipment is manufactured by a second manufacturing entity.
However, in an analogous art, Kim discloses:
wherein the first network equipment is manufactured by a first manufacturing entity and the second network equipment is manufactured by a second manufacturing entity (e.g., Kim, par. [0023]: the test case may include an interoperability test. The interoperability test may define test cases to test and end-to-end function between two different IoT devices developed based on the same standard “(e.g. the IoT devices manufactured by different manufacturers)”).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the determination of compliance with a specified networking protocol standard of Nabeel to include determining that equipment manufactured by different manufacturers is interoperable, as taught by Kim, as Kim would provide the advantage of a means of ensuring multivendor interoperability, as suggested by Nabeel. (See Nabeel, p. 1 right col. second to last par.).
As to claim 18, it is a medium claim having limitations substantially the same as those of claim 9. Accordingly, it is rejected for substantially the same reasons.
Claims 6-8 are rejected under 35 U.S.C. 103 as being unpatentable over Nabeel (“Test Code Generation for Telecom Software using a Two-Stage Generative Model”) in view of Maturana (US 2025/0147736) in view of Wen (CN 112966095) in further view of Libbey et al. (US 2025/0165228) (art of record – hereinafter Libbey).
As to claim 6, Nabeel/Maturana/Wen discloses the system of claim 1 (see rejection of claim 1 above), and further discloses structured test case data of the collection of structured test case data (see rejection of claim 1 above) but does not explicitly disclose wherein the generative artificial intelligence model translates keywords included in each of the structured test case data of the collection of structured test case data based on a first library of keywords associated with a generic executable programming language specification and a second library of statistical occurrence associations of the keywords in the generic executable programming language specification.
However, in an analogous art, Libbey discloses:
wherein the generative artificial intelligence model translates keywords included in each of the structured data (e.g., Libbey, par. [0063]: a generative language model, such as an LLM, may be used to generate programming language code. For example, a user may input a prompt into the LLM such as “change the format of my webpage to look different” and in response the LLM may generate a computer program to perform the operation requested) based on a first library of keywords associated with a generic executable programming language specification and a second library of statistical occurrence associations of the keywords in the generic executable programming language specification (e.g., Libbey, par. [0050]: typically, a token [keyword] may be an integer that corresponds to the index of a text segment “(e.g., a word)” in a vocabulary dataset; par. [0014]: each token of a plurality of tokens [first library]; par. [0078]: the generative language model generates a plurality of values [second library], each of the values indicative of a probability of a respective token being a next token; par. [0089]: the probability that a token is the next token given one or more previously generated tokens [i.e., that probability is a statistical occurrences association of the tokens]; par. [0004]: only a token compliant with the grammar of the programming language [programming language specification] is selected as the next token).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the generative artificial intelligence model generation of programming language code based on structured test case data in a collection of structured test case data taught by Nabeel/Maturana/Wen such that the generative artificial intelligence model translates keywords included in each of the structured data based on a first library of keywords associated with a generic executable programming language specification and a second library of statistical occurrence associations of the keywords in the generic executable programming language specification, as taught by Libbey, as Libbey would provide the advantage of a means of generating code using a language model, as suggested by Nabeel. (See Libbey, par. [0046]). Libbey would also provide the advantage of a means of generating a sequence of code that is most likely to be appropriate. (See Libbey, par. [0077]).
As to claim 7, Nabeel/Maturana/Wen/Libbey discloses the system of claim 6 (see rejection of claim 1 above), but Nabeel/Maturana/Wen does not explicitly disclose wherein the generative artificial intelligence model uses a word to vector process to obtain the vector representation of each keyword included in the first library of keywords associated with the generic executable programming language specification, and wherein the vector representation comprises a group of number values corresponding to each keyword.
However, in an analogous art, Libbey discloses:
wherein the generative artificial intelligence model uses a word to vector process to obtain a vector representation of each keyword included in the first library of keywords associated with the generic executable programming language specification, and wherein the vector representation comprises a group of number values corresponding to each keyword (e.g., Libbey, par. [0004]: am unnormalized probability that the token [keyword, in the library associated with the generic executable programming language specification as set forth above with respect to claim 6] corresponding to that value is the next token; Fig. 4 and associated text, par. [0079]: the values in the tensor 520 are the unnormalized probabilities referred to above in FIG. 3. The tensor may be a 1-D tensor 520 or a vector, as in the illustrated example).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify generative artificial intelligence model of Nabeel/Maturana/Wen such that the generative artificial intelligence model uses a word to vector process to obtain a vector representation of each keyword included in the first library of keywords associated with the generic executable programming language specification, and wherein the vector representation comprises a group of number values corresponding to each keyword, as taught by Libbey, as Libbey would provide the advantage of a means of generating code using a language model, as suggested by Nabeel. (See Libbey, par. [0046]). Libbey would also provide the advantage of a means of generating a sequence of code that is most likely to be appropriate. (See Libbey, par. [0077]).
we
As to claim 8, Nabeel/Maturana/Wen/Libbey discloses the system of claim 7 (see rejection of claim 7 above), but Nabeel/Maturana does not explicitly disclose wherein the group of number values corresponding to each keyword captures a relationship between a first keyword included in the first library of keywords and a second keyword included in the first library of keywords.
However, in an analogous art, Libbey discloses:
wherein the group of number values corresponding to each keyword captures a relationship between a first keyword included in the first library of keywords and a second keyword included in the first library of keywords (e.g., Libbey, par. [0014]: each token [keyword] of a plurality of tokens [first library]; par. [0078]: the generative language model generates a plurality of values, each of the values indicative of a probability of a respective token being a next token; par. [0089]: the probability that a token is the next token given one or more previously generated tokens [i.e., that probability is a statistical occurrences association of the tokens]).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the generative artificial intelligence model generating programming language code taught by Nabeel/Maturana/Wen to include generating a group of number values corresponding to each keyword in a library that captures a relationship between a first keyword included in the library and a second keyword included in the library, as taught by Libbey, as Libbey would provide the advantage of a means of generating code using a language model, as suggested by Nabeel. (See Libbey, par. [0046]). Libbey would also provide the advantage of a means of generating a sequence of code that is most likely to be appropriate. (See Libbey, par. [0077]).
Claims 14 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Nabeel (“Test Code Generation for Telecom Software using a Two-Stage Generative Model”) in view of Maturana (US 2025/0147736) in view of Wen (CN 112966095) in further view of Strenski et al. (US 2023/0236813) (art of record – hereinafter Strenski).
As to claim 14, Nabeel/Maturana/Wen discloses the method of claim 11 (see rejection of claim 11 above), and further discloses a natural language processing machine learning model and the testing framework (see rejection of claim 11 above) but does not explicitly disclose wherein the transforming comprises using, by the device, a natural language processing machine learning model to convert the generic executable programming language code into the testing framework.
However, in an analogous art, Strenski discloses:
wherein the transforming comprises using, by the device, a machine learning model to convert the generic executable programming language code into the framework (e.g., Strenski, par. [0019]: the ML model-based compiler may convert source code in a first language to source code in a second language, compiled executable code, where the executable code runs on a first or second device; par. [0048]: the ML model based compiler may be configured to convert source code to run on a second device).
It would have been obvious to one or ordinary skill in the art to modify the natural language processing machine learning model taught by Nabeel/Maturana, such that the model is used to transform the generic executable programming language code into the framework, as taught by Strenski, as Strenski would provide the advantage of a means of converting to the code to be able to run in the framework. (See Strenski, par. [0048]).
As to claim 20, it is a medium claim having limitations substantially the same as those of claim 14. Accordingly, it is rejected for substantially the same reasons.
Claim 15 is rejected under 35 U.S.C. 103 as being unpatentable over Nabeel (“Test Code Generation for Telecom Software using a Two-Stage Generative Model”) in view of Maturana (US 2025/0147736) in view of Wen (CN 112966095) in view of Strenski (US 2023/0236813) in further view of Janakiraman (US 2025/0321865) (art of record – hereinafter Janakiraman).
As to claim 15, Nabeel/Maturana/Wen/Strenski discloses the method of claim 14 (see rejection of claim 14 above), wherein the natural language processing machine learning model is trained using feedback received in response to executing the generic executable programming language code in the testing framework.
However, in analogous art, Janakiraman discloses:
wherein the natural language processing machine learning model is trained using feedback received in response to executing the generic executable programming language code in the testing framework (e.g., Janakiraman par. [0061]: system 102 uses the large language model to generate test output from the prompt. In particular, system 102 generates test output 304 that include test code for the prompt “(e.g., ‘schedule a meeting next week in the morning with my manager’)”. The test generation system 102 executes the test output 304 “(e.g., the generated test code)” to perform a function test. In doing so, the system 102 can generate feedback 306 indicating whether the test output adequately performs a test; par. [0062]: the system 102 performs an act 308 of adjusting the parameters of the large language model 302 based on the feedback).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the generative artificial intelligence model generating test code taught by Nabeel/Maturana/Wen to include trained using feedback received in response to executing the generic executable programming language code in the testing framework, as taught by Janakiraman, as Janakiraman would provide the advantage of a means of improving accuracy of the model. (See, Janakiraman, par. [0037]).
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 TODD AGUILERA whose telephone number is (571)270-5186. The examiner can normally be reached M-F 11AM - 7:30PM 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, Hyung S Sough can be reached at (571)272-6799. 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.
/TODD AGUILERA/Primary Examiner, Art Unit 2192
Prosecution Insights