DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
(a)(2) the claimed invention was described in a patent issued under section 151, or in an application for patent published or deemed published under section 122(b), in which the patent or application, as the case may be, names another inventor and was effectively filed before the effective filing date of the claimed invention.
Claims 1-8, 11-15 & 18-22 are rejected under 35 U.S.C. 102(a) (2) as being anticipated by Phan et al. (US 20230316150, hereinafter Phan).
Regarding Claim 1, Phan discloses a computer implemented method for decision optimization in a multi-record environment (FIG. 4), the method comprising: receiving a request to make a recommendation in relation to a data record ([0021], receiving, by the computing system, one or more user-indicated preference inputs from a user, indicative of desired outputs, desired output characteristics, or other desired outcomes [recommendation] to be optimized, such as in terms of solution quality and/or running time speed, for example, and incorporating the received user-indicated preference inputs as part of the constraints among which to optimize the control actions for optimizing the desired outputs, desired output [recommendation] characteristics, or other desired outcomes to be optimized; [0069], FIG. 4, knowledge parser 406 may ingest and parse inputs);
PNG
media_image1.png
422
600
media_image1.png
Greyscale
defining the recommendation in terms of an optimization problem comprising a plurality of decision objectives comprising objective contribution functions and a plurality of constraints comprising constraint contribution functions ([0017], control the combination of input rates of input materials to the system, and process rates of processes within the system, such as integrated optimized machine learning for controlling such complex physical systems in novel ways to achieve novel advantages in desired system outputs and outcomes [the recommendation]; [0071], FIG. 4, constraint parser 414 passes parsed constraints as outputs to optimization model generator and solver 418, in parallel with the ML models and model analyses passed from ML model analyzer 410 to optimization model generator and solver 418, all of which optimization model generator and solver 418 may then use as inputs for generating and solving physical system control optimization models based on the ML models, the ML model analyses, and the parsed input constraints);
extracting input data from a data source, the input data comprising a plurality of individual instances of data and a plurality of attributes describing the individual instances of data ([0021], receiving one or more user-indicated preference inputs from a user, indicative of desired outputs, desired output characteristics, or other desired outcomes to be optimized, such as incorporating the received user-indicated preference inputs as part of the constraints among which to optimize the control actions for optimizing the desired outputs, desired output characteristics, or other desired outcomes to be optimized; [0071], FIG. 4, Constraint parser 414 parses machine learning optimization input constraints, and optimization model generator);
based upon the plurality of individual instances of data, identifying a context of the optimization problem, the context relating to a behavior of the input data given the decision objectives and the constraints ([0071], FIG. 4, optimization model generator and solver 418 generates solved machine learning optimization models based at least in part on the parsed machine learning optimization input constraints);
solving the optimization problem by satisfying the plurality of decision objectives and the plurality of constraints, in the context, to generate a solution; and providing the recommendation based on the solution ([0068], FIG. 4, Solution quality checker and UI 420 facilitates user interaction with potential solutions and engages in gauging the quality of generated solutions, and to generate solution outputs 422).
Regarding Claim 2, Phan discloses the method of claim 1, wherein the satisfying the plurality of constraints comprises:
applying the respective constraint contribution functions to the input data ([0017], combining and integrating the separate fields of machine learning and optimization, and iterative improvement feedback loops between machine learning and optimization, to achieve novel advantages in desired system outputs and outcomes, such as maximized output rates of desired physical outputs given the constraints of the physical inputs and the system);
aggregating the constraint contribution functions to an aggregated constraint contribution value; and comparing the aggregated constraint contribution value to a pre-specified constraint contribution limit value ([0071] Optimized machine learning system 400 may further include input constraints 416, which may be outputted to constraint parser 414. Constraint parser 414 may pass parsed constraints as outputs to optimization model generator and solver 418, in parallel with the ML models and model analyses passed from ML model analyzer 410 to optimization model generator and solver 418, all of which optimization model generator and solver 418 may then use as inputs for generating and solving physical system control optimization models based on the ML models, the ML model analyses, and the parsed input constraints, in various examples. Constraint parser 414 may thus parse machine learning optimization input constraints, and optimization model generator and solver 418 may thus generate solved machine learning optimization models based at least in part on the parsed machine learning optimization input constraints).
Regarding Claim 3, Phan discloses the method of claim 2, wherein the extracting the input data comprises:
fetching the input data from the data source; arranging the input data in a canonical data structure; and providing the input data, arranged in the canonical data structure, to an optimization solver, wherein the optimization problem is solved by the optimization solver ([0021], receiving one or more user-indicated preference inputs from a user, indicative of desired outputs, desired output characteristics, or other desired outcomes to be optimized, such as incorporating the received user-indicated preference inputs as part of the constraints among which to optimize the control actions for optimizing the desired outputs, desired output characteristics, or other desired outcomes to be optimized; [0071], FIG. 4, Constraint parser 414 parses machine learning optimization input constraints, and optimization model generator).
Regarding Claim 4, Phan discloses the method of claim 3, wherein the input data and the optimization solver are independent of each other ([0021], receiving one or more user-indicated preference inputs from a user, indicative of desired outputs, desired output characteristics, or other desired outcomes to be optimized, such as incorporating the received user-indicated preference inputs as part of the constraints among which to optimize the control actions for optimizing the desired outputs, desired output characteristics, or other desired outcomes to be optimized; [0071], FIG. 4, Constraint parser 414 parses machine learning optimization input constraints, and optimization model generator).
Regarding Claim 5, Phan discloses the method of claim 3, wherein the optimization solver optimizes the objective contribution functions and the constraint contribution functions using an iterative approach based on metaheuristics of the optimization solver ([0017], control the combination of input rates of input materials to the system, and process rates of processes within the system, such as integrated optimized machine learning for controlling such complex physical systems in novel ways to achieve novel advantages in desired system outputs and outcomes [the recommendation]; [0071], FIG. 4, constraint parser 414 passes parsed constraints as outputs to optimization model generator and solver 418, in parallel with the ML models and model analyses passed from ML model analyzer 410 to optimization model generator and solver 418, all of which optimization model generator and solver 418 may then use as inputs for generating and solving physical system control optimization models based on the ML models, the ML model analyses, and the parsed input constraints).
Regarding Claim 6, Phan discloses the method of claim 3, further comprising storing the solution, wherein the solution is accessed directly in real time or at a different time during a solution consumption stage ([0068], FIG. 4, Solution quality checker and UI 420 facilitates user interaction with potential solutions and engages in gauging the quality of generated solutions, and to generate solution outputs 422).
Regarding Claim 7, Phan discloses the method of claim 6, further comprising:
pairing the solution to a corresponding individual instance of data; and applying the solution to the corresponding individual instance of data ([0021], receiving one or more user-indicated preference inputs from a user, indicative of desired outputs, desired output characteristics, or other desired outcomes to be optimized, such as incorporating the received user-indicated preference inputs as part of the constraints among which to optimize the control actions for optimizing the desired outputs, desired output characteristics, or other desired outcomes to be optimized; [0071], FIG. 4, Constraint parser 414 parses machine learning optimization input constraints, and optimization model generator; [0071], FIG. 4, optimization model generator and solver 418 generates solved machine learning optimization models based at least in part on the parsed machine learning optimization input constraints).
Regarding Claim 8, Phan discloses the method of claim 1, wherein the input data comprises dynamic data with a dynamic context, the dynamic context having a nonpredetermined behavior ([0079], ], FIG. 4, generate control inputs and predicted outputs in the form of real-time, near-real-time, delayed, and/or online controls; short-term scheduled controls; long-term planning; and/or offline controls, and/or any other types of control).
Regarding Claims 11-15, system claims 11-15 of using the corresponding method claimed in claims 1-3, 5 & 8 and the rejections of which are incorporated herein for the same reasons as used above.
Regarding Claims 18-23, Computer media claims 18-23 of using the corresponding method claimed in claims 1-3, 5, 7 & 8, and the rejections of which are incorporated herein for the same reasons as used above.
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 of this title, 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 9-10, 16-17 & 24-25 are rejected under 35 U.S.C. 103 as being unpatentable over Phan et al. (US 20230316150, hereinafter Phan) in view of Kobayashi et al. (US 20240061997, hereinafter Kobayashi).
Regarding Claim 9, Phan discloses the method of claim 1, buy does not explicitly disclose wherein the dynamic context is configured to populate an abstract syntax tree structure conforming to a corresponding formula grammar.
Kobayashi teaches wherein the input data comprises dynamic data with a dynamic context, the dynamic context having a nonpredetermined behavior ([0024], a machine learning (ML) explainability (MLX) approach in which a natural language explanation is generated based on analysis of a parse tree with scored nodes such as for a suspicious database query or JavaScript. By analyzing relevance scores assigned to individual nodes in an abstract syntax tree (AST) of source code).
Therefore, it would have been obvious to one ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of abstract syntax tree structure as taught by Kobayashi ([0121]) into the machine learning system of Pham in order to provide systems for reducing a time and space for generating an explanation of a model inference and avoiding wastage of processing irrelevant nodes in an effective manner (Kobayashi, [0011]).
Regarding Claim 10, Phan in view of Kobayashi discloses the method of claim 9, Phan discloses wherein a performance of the abstract syntax tree structure is optimized using an optimization strategy based on a mathematical formula of the abstract syntax tree structure ([0074], The optimized machine learning system 500 may then analyze physical process system abstraction 552 such as in a regression analysis such as shown in Equation 1).
Regarding Claims 16-17, system claims 16-17 of using the corresponding method claimed in claims 9-10, and the rejections of which are incorporated herein for the same reasons as used above.
Regarding Claims 24-25, Computer media claims 24-25 of using the corresponding method claimed in claims 9-10, and the rejections of which are incorporated herein for the same reasons as used above.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Samuel D Fereja whose telephone number is (469)295-9243. The examiner can normally be reached 8AM-5PM.
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, DAVID CZEKAJ can be reached at (571) 272-7327. 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.
/SAMUEL D FEREJA/Primary Examiner, Art Unit 2487