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 .
DETAILED ACTION
The action is in response to claims dated 6/20/2025
Claims pending in the case: 1-20
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claim 1-20 rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for pre-AIA the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claim 1, 10 and 19 recites “classifying training data based on a result of analysis of the neural network data of the source framework”. Although the specification recites the same statement, there is information on how analysis data can be used to classify training data. An explanation of how training data and network information are being used in the classification process was not found in the specification. The specification does not seem to explain how the classification is to be done i.e. how the analysis data from the neural network is to be used to form the classes. The applicant is requested to identify the paragraphs and lines in the specification that supports an explanation of this limitation.
Claim 7 and 16 recites “classifying the training data based on a variable list acquired by performing the lexical and syntactic analysis”. Although the specification recites the same statement, there is no explanation on how this variable list can be used to classify training data. The examiner was unable to find an explanation on how the variable list is being used to classify training data. The applicant is requested to identify the paragraphs and lines in the specification that supports an explanation of this limitation.
All claims dependent on this claim are also rejected under 35 U.S.C. 112(a) due to the virtue of their respective direct and indirect dependencies.
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.
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.
Claim(s) 1, 10 and 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ryu (KR 20210157813 – refer to attached English translation for claim mapping) in view of Campos (US 7069256) and Nandanuru (US 9971581).
Nandanuru not used in the prior office action.
Regarding claim 1, Ryu teaches, a method for converting a neural network, comprising:
separating neural network data of a source framework to form a tree structure by analyzing the neural network data and converting the neural network data in the tree structure to a neural network optimized for a target framework … (Ryu: abstract, Pg. 3 [11], Pg. 5 [1-3, 6]: convert neural network into a tree-structure and optimize it);
classifying training data … and converting the classified training data to a training data structure of the target framework (Ryu: Pg. 5 [3]: training by utilizing parameter reuse; select as per needs of user (classify)); and
creating a neural network and training data of the target framework by combining the converted neural network and the converted training data (Ryu: abstract, Pg. 3 [11], Pg. 5 [1-3, 6]: convert neural network into a tree-structure and optimize it with appropriate training data);
However Ryu does not specifically teach,
using a mapping table that that defines correspondence between instructions and parameters of the source framework and the target framework;
classifying training data based on a result of analysis of the neural network data of the source framework;
Campos teaches, classifying training data based on a result of analysis of the neural network data of the source framework (Campos: col 5 lines 40-45, col 6 line 46-col 7 line 16: training data and validation data based on model analysis (subset of data used for the model); classified by score);
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings of Ryu and Campos because the combination would enable using training data based on a neural network. One of ordinary skill in the art would have been motivated to combine the teachings because the combination would provide “ a technique by which neural network analysis may be performed that provides improved performance in model building, … flexible specification and adjustment of the models being built, and flexible model arrangement and export capability” (see Campos col 1 lines 51-56).
Nandanuru teaches, using a mapping table that that defines correspondence between instructions and parameters of the source framework and the target framework (Nandanuru: col 6 lines 23-32 : mapping table may be used to convert code from one framework to another);
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings of Ryu, Campos and Nandanuru because the combination would enable using mapping table to adapt to a different framework. One of ordinary skill in the art would have been motivated to combine the teachings because the combination would reduce time to migrate application codes from one framework to another (Nandanuru col 1 lines 47-51).
Regarding Claim(s) 10 and 19, this/these claim(s) is/are similar in scope as claim(s) 1. Therefore, this/these claim(s) is/are rejected under the same rationale.
Claim(s) 2-6, 11-15, 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ryu (KR 20210157813 – refer to attached English translation for claim mapping) in view of Campos (US 7069256) and Nandanuru (US 9971581) in further view of Fujita (JP H05181684 – refer to attached English translation for claim mapping).
Regarding claim 2, Ryu, Campos and Nandanuru teach the invention as claimed in claim 1 and further, wherein converting the neural network data in the tree structure comprises:
performing …analysis on neural network code of the source framework based on a previously stored neural network data structure of the source framework (Ryu: abstract, Pg. 3 [11], Pg. 5 [1-3, 6]: convert neural network into a tree-structure – require network analysis);
creating a tree structure formed of instructions and parameters from the neural network code based on a result of the analysis (Ryu: Pg. 4 [3], Pg. 7 [2]: tree-structure of neural network – network instructions and parameters);
Ryu Campos and Nandanuru do not specifically teach,
performing lexical and syntactic analysis on code of the source framework (Fujita: Pg. 3 [4-5]: a lexical / syntactic analysis may be done on source code to generate a syntax tree);
converting the instructions and parameters of the created tree structure based on a mapping table in which instructions and parameters of the target framework are listed
Fujita teaches, performing lexical and syntactic analysis on code of the source framework (Fujita: Pg. 3 [4-5]: a lexical / syntactic analysis may be done on source code to generate a syntax tree);
converting the instructions and parameters of the created tree structure based on a mapping table in which instructions and parameters of the target framework are listed (Fujita: Pg. 3 [1]: modify as per mapping table of the source code);
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings of Ryu, Campos, Nandanuru and Fujita because the combination would enable using lexical and syntactic analysis on source code to generate a tree structure that can be modified. One of ordinary skill in the art would have been motivated to combine the teachings because the combination would automate code modification and increase the efficiency of program development (see Fujita [2-5]).
Regarding claim 3, Ryu Campos, Nandanuru and Fujita teaches the invention as claimed in claim 2 and further, further comprising: validating the instruction based on whether the instruction is present, wherein, when the instruction is not validated, an instruction error message is output (Fujita: Pg. 3 [3]: code analysis may prompt error message as output). It is noted here that it would have been obvious to one skilled in the art to insert error message prompts during validation process as deemed useful. This limitation does not add any inventive concept.
Regarding claim 4, Ryu, Campos, Nandanuru and Fujita teaches the invention as claimed in claim 2 and further, further comprising: validating ranges and fields of the parameters, wherein, when the ranges or fields of the parameters are not validated, a parameter range error message is output (Fujita: Pg. 3 [3]: code analysis may prompt error message as output). It is noted here that it would have been obvious to one skilled in the art to insert error message prompts during validation process as deemed useful. This limitation does not add any inventive concept.
Regarding claim 5, Ryu, Campos, Nandanuru and Fujita teaches the invention as claimed in claim 2 and further, wherein: converting the instructions and parameters of the created tree structure (Ryu: abstract, Pg. 3 [11], Pg. 5 [1-3, 6]: convert neural network into a tree-structure and optimize it to convert the neural network) comprises checking whether an error is present in a structure and operation of the neural network that is converted based on the mapping table (Fujita: Pg. 3 [1-3]: code analysis to identify and correct error), and
when there is no error, neural network code, acquired through conversion to instructions and a parameter structure of the neural network of the target framework, is stored (Ryu: abstract, Pg. 3 [11], Pg. 5 [1-3, 6]: convert neural network into a tree-structure and optimize it with appropriate training data; Pg. 4 [6]: store network).
It is noted here that validation process that comprise checking for errors in code and correcting the errors is a standard process in software coding and code conversion. The limitations as claimed do not add any specifics to amount to anything more than a generalized validation process practiced in the art. Therefore this limitation would have been obvious to one skilled in the art.
Regarding claim 6, Ryu, Campos, Nandanuru and Fujita teaches the invention as claimed in claim 2 and further, wherein: performing the lexical and syntactic analysis, creating the tree structure, and converting the instructions and parameters of the created tree structure are repeated for each line of neural network instruction code (Ryu: abstract, Pg. 3 [11], Pg. 5 [1-3, 6]: convert neural network into a tree-structure and optimize it) (Fujita: Pg. 3 [4-5]: a lexical / syntactic analysis may be done on source code to generate a syntax tree). It would have been obvious to one skilled in the art that the process of conversion and optimization needs to address the entire source code in order for the process to be complete,
Regarding Claim(s) 11, this/these claim(s) is/are similar in scope as claim(s) 2. Therefore, this/these claim(s) is/are rejected under the same rationale.
Regarding Claim(s) 12-15, this/these claim(s) is/are similar in scope as claim(s) 3-6 respectively. Therefore, this/these claim(s) is/are rejected under the same rationale.
Regarding Claim(s) 20, this/these claim(s) is/are similar in scope as claim(s) 15. Therefore, this/these claim(s) is/are rejected under the same rationale.
Claim(s) 7-8 and 16-17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ryu (KR 20210157813 – refer to attached English translation for claim mapping), Campos (US 7069256) and Nandanuru (US 9971581) and Fujita (JP H05181684 – refer to attached English translation for claim mapping) in view of Chaski (US 20180101518).
Regarding claim 7, Ryu Campos, Nandanuru and Fujita teach the invention as claimed in claim 2 and further, wherein converting the classified training data comprises: …;
optimizing the classified training data based on user requirements; and converting the optimized training data to the training data structure of the target framework (Ryu: Pg. 5 [3]: training by utilizing parameter reuse; select as per needs of user) (Campos: col 5 lines 40-45, 58-64 : optimize training data);
a variable list acquired by performing the lexical and syntactic analysis (Fujita: Pg. 3 [4-5]: a lexical / syntactic analysis may be done on source code to generate a syntax tree)
But not, classifying the training data based on a variable list;
Chaski teaches, classifying the training data based on a variable list (Chaski: [25-26]: classify based on variables that quantify the data);
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings of Ryu, Campos, Nandanuru, Fujita and Chaski because the combination would enable classifying the training data based on a variable list acquired by performing the lexical and syntactic analysis. One of ordinary skill in the art would have been motivated to combine the teachings because the combination would automate classifying data by using “several statistical procedures for classifying and predicting group membership” (see Chaski [30]).
Regarding claim 8, Ryu, Campos, Nandanuru, Fujita and Chaski teach the invention as claimed in claim 7 and further, wherein converting the classified training data further comprises: before optimizing the classified training data, detecting an error through comparison and analysis of respective variables and array coefficients of the training data classified using the variable list and the parameters (Fujita: Pg. 3 [1-3]: source code analysis to identify and correct error (parameter list of source code)).
It is noted here that validation process that comprise checking for errors in code and correcting the errors is a standard process in software coding and code conversion. The limitations as claimed do not add any specifics to amount to anything more than a generalized validation process practiced in the art. Therefore this limitation would have been obvious to one skilled in the art.
Regarding Claim(s) 16-17, this/these claim(s) is/are similar in scope as claim(s) 7-8 respectively. Therefore, this/these claim(s) is/are rejected under the same rationale.
Claim(s) 9 and 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Ryu (KR 20210157813 – refer to attached English translation for claim mapping), Campos (US 7069256) and Nandanuru (US 9971581), Fujita (JP H05181684 – refer to attached English translation for claim mapping) and Chaski (US 20180101518) in view of Vantrease (US 20190294413)
Regarding claim 9, Ryu, Campos, Nandanuru, Fujita and Chaski teaches the invention as claimed in claim 7 but not, wherein optimizing the training data is configured to perform at least one of optimization methods for quantization calculation and reduction of a size of a real number;
Vantrease teaches, wherein optimizing the training data is configured to perform at least one of optimization methods for quantization calculation and reduction of a size of a real number (Vantrease: [85-86]: quantization process); It is noted here quantization is a known process for data size reduction and may be done on a broad range of data that may be used as training data.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine the teachings of Ryu, Campos, Nandanuru, Fujita, Chaski and Vantrease because the combination would enable using quantization calculation to reduce data size. One of ordinary skill in the art would have been motivated to combine the teachings because the combination would reduce the size of the data and “as such a device with limited memory space and computation power may be able to handle the inference process, in some cases, in real time” (see Vantrease [18]).
Regarding Claim(s) 18, this/these claim(s) is/are similar in scope as claim(s) 9 respectively. Therefore, this/these claim(s) is/are rejected under the same rationale.
Response to Arguments
Applicant’s arguments on the 112(a) rejections have been fully considered and are not persuasive.
Applicant argues “Paragraph [00065] of the specification clearly states that classification is based on “the result of lexical and syntactic analysis of the neural network. Further, paragraphs [00065 ]—[00066] and [00091 ]—[00094] describe in detail the procedure for analyzing the source framework’s neural network code to extract variable information and classify the training data into input, output, and label categories accordingly.”. As mentioned in the rejection, [65] mentions classification based on a variable list acquired by analysis of the neural network. This paragraph does not explain how the classification is to be done i.e. how the variable list from neural network data is to be used to form the classes. [66] recites “the training data is stored in the form of an array through classification and analysis of the variables, arrays, and argument values of the neural network” which does not provide any insight on how the variables, arrays etc. are to be used to form the classes. [91-94] also mentions variable list and parameters but provides no information how such a list can be used in the classification process to form different classes. Paragraph [101-104, 109] indicated by the applicant does not mention training data at all and therefore does not address how neural network information can be used to classify training data. The applicant also indicates a paragraph [113] as support. However the specification has only 111 paragraphs. Therefore the paragraphs cited by the applicant does not provide the information needed by one of ordinary skill to classify training data. Therefore the 112(a) rejections are maintained.
Applicants’ arguments on the 112(b) rejections have been considered. The examiner finds that the limitations may be considered as extremely broad instead of indefinite. Thus the 112(b) rejections are respectfully withdrawn.
Applicants’ prior art arguments have been fully considered but since they pertain to the amended sections of the claim, they are considered moot in view of the new grounds of rejection presented above.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure in attached 892.
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 extension fee 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 date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MANDRITA BRAHMACHARI whose telephone number is (571)272-9735. The examiner can normally be reached Monday to Friday, 11 am to 8 pm 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, Tamara Kyle can be reached on 571 272 4241. 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.
/Mandrita Brahmachari/Primary Examiner, Art Unit 2144