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 CORRESPONDENCE
Priority
Applicant’s claim for the benefit of a prior-filed application is acknowledged.
Status of Claims
Claims 6, 7, 8, 9, 10, 14, 15 have been amended in the preliminary amendments received on 11/22/2024.
Claims 14, 15 have been amended in the preliminary amendments received on 12/11/2024.
No claims have been cancelled.
No claims have been added.
Information Disclosure Statement
The information disclosure statement filed November 22, 2024 fails to comply with 37 CFR 1.98(a)(3)(i) because it does not include a concise explanation of the relevance, as it is presently understood by the individual designated in 37 CFR 1.56(c) most knowledgeable about the content of the information, of each reference listed that is not in the English language. It has been placed in the application file, but the information referred to therein has not been considered. Contrary to the statement made on item 23 of the IDS, the written opinion does not have an English translation accompanying it.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation(s) is/are:
Claim 12:
generating device configured refer … generate … connecting
Claim 13:
collecting system configured to collect
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof. ¶ 65 of the specification recites, “The generating device 10 and the execution device 40 are computers.” See also ¶ 109, 117 and Fig. 35.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 112(b)
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 13 – 15 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 13 recites the limitation "the generating device" in line 2. There is insufficient antecedent basis for this limitation in the claim.
Claim 14 recites the limitation "the computer" and “the processing method” in lines 1, 2, respectively. There is insufficient antecedent basis for this limitation in the claim.
Claim 15 recites the limitation "the computer" and “the processing method” in line 2. There is insufficient antecedent basis for this limitation in the claim.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claim 14 is rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claimed invention is directed to a program and is, consequently, comprised of software (data) not claimed as embodied in non-transitory computer-readable media. Therefore, claim 14 is descriptive material per se and is not statutory because they are not capable of causing function change in a computer. See In re Warmerdam, 33 F.3d at 1361, 31 USPQ2d at 1760. The limitations set forth in the claims are directed to software and software, per se, is not statutory.
Claims 1 – 15 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claims recite:
(Claim 13) collect quality data regarding the product
refer to an individual model indicating a measure against a cause of a defect mode in a product
generate consolidated models for a plurality of the defect modes respectively, each consolidated model of the consolidated models being generated by rearranging and connecting a plurality of the individual models in accordance with a plurality of weights respectively set for a plurality of the causes
The invention is directed towards the abstract idea of asset maintenance and management, which corresponds to “Mental Processes” and “Certain Methods of Organizing Human Activities” as it is directed towards steps that can be performed by a human(s), in the human mind, and/or with the aid of pen and paper and fundamental economic principles or practices (including hedging, insurance, mitigating risk); commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations); managing personal behavior or relationships or interactions between people (including social activities, teaching, and following rules or instructions), e.g., observing and collecting performance information of an asset and, based on a rule(s), determine if the asset has a defect and determining the defect mode.
The limitations of:
(Claim 13) collect quality data regarding the product
refer to an individual model indicating a measure against a cause of a defect mode in a product
generate consolidated models for a plurality of the defect modes respectively, each consolidated model of the consolidated models being generated by rearranging and connecting a plurality of the individual models in accordance with a plurality of weights respectively set for a plurality of the causes,
are processes that, under its broadest reasonable interpretation, covers performance of the limitation performed by a human(s), in the human mind, and/or with the aid of pen and paper, but for the recitation of generic models, generic computing devices, and generic non-transitory computer-readable storage medium. That is, other than reciting generic models, generic computing devices, and generic non-transitory computer-readable storage medium nothing in the claim element precludes the step from practically being performed in the mind. For example, but for the generic models, generic computing devices, and generic non-transitory computer-readable storage medium in the context of this claim encompasses a user can observe and collecting performance information of an asset and, based on the collected information and applicable rule(s), determining a defect mode. If a claim limitation, under its broadest reasonable interpretation, covers performance of the limitation in the mind but for the recitation of a generic models, generic computing devices, and generic non-transitory computer-readable storage medium, then it falls within the “Mental Processes” and “Certain Methods of Organizing Human Activities” groupings of abstract ideas. Accordingly, the claims recite an abstract idea.
This judicial exception is not integrated into a practical application. In particular, the claim only recites additional elements – generic computing devices and generic non-transitory computer-readable storage medium, as well as performing operations that a human can perform in their mind and/or pen and paper, i.e. determining a defect mode based on collected information and applicable rule(s). The generic computing devices and generic non-transitory computer-readable storage medium in the steps are recited at a high-level of generality (i.e., as a generic computing devices and generic non-transitory computer-readable storage medium can perform the insignificant extra solution steps of referring to stored information (See MPEP 2106.05(g) while also reciting that the a generic computing devices and generic non-transitory computer-readable storage medium are merely being applied to perform the steps that can be performed by a human(s), in the human mind, and/or with the aid of pen and paper; "[use] of a computer or other machinery in its ordinary capacity for economic or other tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., a fundamental economic practice or mathematical equation) does not integrate a judicial exception into a practical application or provide significantly more.” Therefore, according to the MPEP, this is not solely limited to computers but includes other technology that, recited in an equivalent to “apply it,” is a mere instruction to perform the abstract idea on that technology (See MPEP 2106.05(f)) such that it amounts no more than mere instructions to apply the exception using generic computing devices and generic non-transitory computer-readable storage medium.
Although the claim recite “an individual model” and “consolidated models” the claims and specification fail to provide sufficient disclosure regarding an improvement to how a machine learning algorithm can be trained, but simply recites a high-level generic recitation that a plurality of models are trained and generically combined/consolidated. There is insufficient evidence from the specification to indicate that the use of the machine learning algorithm involves anything other than the generic application of a known technique or that the claimed invention purports to improve the functioning of the computer itself or the machine learning algorithm. None of the limitations reflects an improvement in the functioning of a computer, or an improvement to other technology or technical field, applies or uses a judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition, implements a judicial exception with, or uses a judicial exception in conjunction with, a particular machine or manufacture that is integral to the claim, effects a transformation or reduction of a particular article to a different state or thing, or applies or uses the judicial exception in some other meaningful way beyond generally linking the use of the judicial exception to a particular technological environment, such that the claim as a whole is more than a drafting effort designed to monopolize the exception.
Even training and applying machine learning models are simply an application of a computer model, itself an abstract idea manifestation. Further, such training and applying of a model is no more than putting data into a black box machine learning operation. The nomination as being a consolidation of machine learning models is a functional label, devoid of technological implementation and application details. The specification does not contend it invented any of these activities, or the creation and use of such machine learning models. In short, each step does no more than require a generic computer to perform generic computer functions. As to the data operated upon, "even if a process of collecting and analyzing information is 'limited to particular content' or a particular 'source,' that limitation does not make the collection and analysis other than abstract." SAP America, Inc. v. InvestPic LLC, 898 F.3d 1161, 1168 (Fed. Cir. 2018).
The Examiner asserts that the scope of the disclosed invention, as presented in the originally filed specification, is not directed towards the improvement of machine learning, but directed towards collecting and analyzing information to determine an asset’s defect mode. The specification’s disclosure on machine learning is nothing more than a high general explanation of generic technology and applying it to the abstract idea. Referring to MPEP § 2106.05(f), the individual and consolidated models are merely being used to facilitate the tasks of the abstract idea, which provides nothing more than a results-oriented solution that lacks detail of the mechanism for accomplishing the result and is equivalent to the words “apply it,” per MPEP § 2106.05(f). The Examiner asserts that in light of the 2024 Guidance Update on Patent Subject Matter Eligibility, Including on Artificial Intelligence, the claimed invention is analogous to Example 47, Claim 2.
Further, the combination of these elements is nothing more than a generic computing system with machine learning model(s). Because the additional elements are merely instructions to apply the abstract idea to a computer, as described in MPEP § 2106.05(f), they do not integrate the abstract idea into a practical application.
Accordingly, these additional elements do not integrate the abstract idea into a practical application because they do not impose any meaningful limits on practicing the abstract idea. The claims are directed to an abstract idea.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above with respect to integration of the abstract idea into a practical application, the additional element of using generic models, generic computing devices, and generic non-transitory computer-readable storage medium to perform the steps of:
(Claim 13) collect quality data regarding the product
refer to an individual model indicating a measure against a cause of a defect mode in a product
generate consolidated models for a plurality of the defect modes respectively, each consolidated model of the consolidated models being generated by rearranging and connecting a plurality of the individual models in accordance with a plurality of weights respectively set for a plurality of the causes,
amounts to no more than mere instructions to apply the exception using a generic computer component. Mere instructions to apply an exception using a generic computer component cannot provide an inventive concept.
Additionally:
Claim 2 is directed towards the extra-solution activity of collecting information and reciting generic technology, i.e. consolidated models, at a high level of generality and applying it to the abstract idea, as well as collect and comparing information and, based on a rule(s), identify options.
Claim 3 is directed towards the extra-solution activity of collecting information.
Claims 4, 5 are directed towards updating information and reciting generic technology, updating weights of a model, at a high level of generality and applying it to the abstract idea.
Claim 6 is directed towards the extra-solution activity of displaying information and receiving information provided by a human.
Claim 7 is directed towards the extra-solution activity of displaying information.
Claims 8, 9, 11 are directed towards reciting generic technology at a high level of generality and applying it to the abstract idea.
Claim 10 is directed towards descriptive subject matter.
In summary, the dependent claims are simply directed towards providing additional descriptive factors that are considered for determining an asset’s defect mode. Accordingly, the claims are not patent eligible.
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 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.
Claims 1 – 5, 8 – 15 are rejected under 35 U.S.C. 103 as being unpatentable over Chengdu Shulian Cloud Computing Technology CO Ltd. (CN 113781485), hereinafter referred to as Chengdu, in view of Gunes (Why do stacked ensemble models win data science competitions).
In regards to claims 1, 12, 13, 14, 15, Chengdu discloses (Claim 1) a processing method causing a computer to; (Claim 12) a generating device configured to: (Claim 13) a processing system comprising; (Claim 14) a program causing the computer to perform the processing method according to claim 1; (Claim 15) a non-transitory computer-readable storage medium storing a program causing the computer to perform the processing method according to claim 1:
(Claim 13) the generating device according to claim 12 (see below); and
(Claim 13) a collection system configured to collect quality data regarding the product (Page 5, ¶ 6 “An embodiment of the present disclosure provides an intelligent detection method for a PCB defect category.”
Page 6, ¶ 6 “In the embodiments of the present disclosure, the image detection model includes a deep learning model, which is configured to detect a defect category and a confidence level of the PCB image, and the defect category may be understood as a category defect existing in a PCB corresponding to the PCB image, and the confidence level may be understood as a credibility of the detected defect category.”
Page 7, ¶ 5 “Further, in the embodiment of the present disclosure, the PCB image is input into the trained image detection model to detect the defect category and confidence of the PCB image through the image detection model.”
Page 7, ¶ 8 “In an embodiment of the present disclosure, the image category decision model includes a machine learning model configured to perform secondary learning on an output result of the image detection model to ensure a detection accuracy of a defect category of the PCB image.”)
refer to an individual model indicating a measure against a cause of a defect mode in a product (Page 5, ¶ 6 “An embodiment of the present disclosure provides an intelligent detection method for a PCB defect category.”
Page 6, ¶ 6 “In the embodiments of the present disclosure, the image detection model includes a deep learning model, which is configured to detect a defect category and a confidence level of the PCB image, and the defect category may be understood as a category defect existing in a PCB corresponding to the PCB image, and the confidence level may be understood as a credibility of the detected defect category.”
Page 7, ¶ 5 “Further, in the embodiment of the present disclosure, the PCB image is input into the trained image detection model to detect the defect category and confidence of the PCB image through the image detection model.”
Page 7, ¶ 8 “In an embodiment of the present disclosure, the image category decision model includes a machine learning model configured to perform secondary learning on an output result of the image detection model to ensure a detection accuracy of a defect category of the PCB image.”
Page 7, ¶ 9 “As an embodiment of the present disclosure, referring to FIG. 3, the inputting the defect category, the confidence level, and the defect feature into a trained image category decision model for processing to obtain a final defect category of the PCB image includes:”); and
generate [a] model[…] for a plurality of the defect modes respectively, […] the […] model[…] being generated by rearranging and connecting [an] individual model[…] in accordance with a […] weight[…] [of a] set for a plurality of the causes (Abstract “According to the invention, automatic identification of PCB image defect categories can be realized, and the accuracy of PCB image defect classification is improved.”
Page 3, 6 “Based on this, there is an urgent need for a PCB defect classification method to realize automatic identification of PCB image defect categories and improve accuracy of PCB image defect classification.”
Page 4, ¶ 5 “It can be learned that, in this embodiment of the present disclosure, the defect confidence of the candidate box area in the training image is calculated by using the image detection model, and the candidate box area with the largest defect confidence is used as the predicted defect area of the training image, so that the candidate box area and the defect category that meet the maximum weight proportion can be screened out, to ensure the accuracy of obtaining the candidate box area and the defect category.”
Page 7, ¶ 4 “Further, in an optional embodiment of the present disclosure, a parameter of the image detection model includes a weight and a bias, and parameter adjustment of the model may be implemented by a gradient descent algorithm, for example, a stochastic gradient descent algorithm.”).
Chengdu discloses:
Page 3, ¶ 4 “The current PCB board manufacturer finds a defect substrate, usually uses a camera to capture a PCB, and then performs manual judgment by a special judgment personnel, the effect of the manual judgment is dependent on the proficiency of the employee technology, and the manual judgment also requires a high labor cost. Therefore, an intelligent substitute artificial judgment is used to realize automatic identification and classification of the production defect of the PCB, which is an urgent need for PCB board production.”
Page 3, ¶ 5 “In addition, the intelligent judgment technology in the prior art mainly uses deep learning to detect and classify images of defects, and since the accuracy of the judgment is limited to the defect distribution of the training set image and cannot learn the structured information of the image, the accuracy of PCB image defect classification cannot be guaranteed when performing PCB image defect classification by using deep learning.”
Accordingly, Chengdu discloses a system and method that leverages machine learning to analyze a plurality of printed circuit boards to determine if a detect exists and, if so, classify the defect into a particular category based on the features of the defect, as well as weights. Despite this, Chengdu fails to explicitly disclose whether it would have been obvious to one of ordinary skill in the art to consolidate a plurality of models.
To be more specific, Chengdu fails to explicitly disclose:
generate consolidated models for a plurality of the defect modes respectively, each consolidated model of the consolidated models being generated by rearranging and connecting a plurality of the individual models in accordance with a plurality of weights respectively set for a plurality of the causes
However, Gunes teaches:
“Ensemble methods are commonly used to boost predictive accuracy by combining the predictions of multiple machine learning models. The traditional wisdom has been to combine so-called "weak" learners. However, a more modern approach is to create an ensemble of a well-chosen collection of strong yet diverse models.”
(Page 2)
“Building powerful ensemble models has many parallels with building successful human teams in business, science, politics, and sports. Each team member makes a significant contribution and individual weaknesses and biases are offset by the strengths of other members.”
(Page 2)
“More generally, you can think about using weighted averages. For example, you might believe that some of the models are better or more accurate and you want to manually assign higher weights to them. But an even better approach might be to estimate these weights more intelligently by using another layer of learning algorithm. This approach is called model stacking.”
(Page 2)
“The figure shows that a diverse set of 64 single models were used to build the model library. These models are trained by using various machine learning algorithms. For example, the green boxes represent gradient boosting models (GBM), pink boxes represent neural network models (NN), and orange boxes represent factorization machines models (FM). You can see that there are multiple gradient boosting models in the model library; they probably vary in their use of different hyperparameter settings and/or feature sets.”
(Page 4)
“In order to build a powerful predictive model like the one that was used to win the 2015 KDD Cup, building a diverse set of initial models plays an important role! There are various ways to enhance diversity such as using:
• Different training algorithms.
• Different hyperparameter settings.
• Different feature subsets.
• Different training sets.”
(Page 4)
“For the same machine learning algorithm, you can enhance diversity by using different hyperparameter settings and subsets of variables. If you have many features, one efficient method is to choose subsets of the variables by simple random sampling. Choosing subsets of variables could be done in more principled fashion that is based on some computed measure of importance which introduces the large and difficult problem of feature selection.”
(Page 5)
“Applying stacked models to real -world big data problems can produce greater prediction accuracy and robustness than do individual models. The model stacking approach is powerful and compelling enough to alter your initial data mining mindset from finding the single best model to finding a collection of really good complementary models.”
(Page 6)
One of ordinary skill would have been motivated to aggregate a plurality of models each configured with their own respective algorithms and, by extension, weights because it allows for the selection of the “best” or “strongest” models to work together to provide more accurate solutions.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to incorporate into the machine learning based asset detection system and method of Chengdu with the ability to consolidate a plurality of models and weights, as taught by Gunes, because consolidating various diverse models and corresponding weights would result in a boost to predictive accuracy.
In regards to claim 2, the combination of Chengdu and Gunes discloses the processing method according to claim 1, further causing the computer to:
obtain quality data regarding the product; and
execute, when the quality data indicates an occurrence of any defect mode among the plurality of defect modes, one of the consolidated models corresponding to the occurring defect mode
(Chengdu:
Page 5, ¶ 6 “An embodiment of the present disclosure provides an intelligent detection method for a PCB defect category.”
Page 6, ¶ 6 “In the embodiments of the present disclosure, the image detection model includes a deep learning model, which is configured to detect a defect category and a confidence level of the PCB image, and the defect category may be understood as a category defect existing in a PCB corresponding to the PCB image, and the confidence level may be understood as a credibility of the detected defect category.”
Page 7, ¶ 5 “Further, in the embodiment of the present disclosure, the PCB image is input into the trained image detection model to detect the defect category and confidence of the PCB image through the image detection model.”
Page 7, ¶ 8 “In an embodiment of the present disclosure, the image category decision model includes a machine learning model configured to perform secondary learning on an output result of the image detection model to ensure a detection accuracy of a defect category of the PCB image.”
Gunes:
Pages 2, 4, 5, 6 wherein a plurality of models are consolidated to predict a more accurate solution.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to incorporate into the machine learning based asset detection system and method of Chengdu with the ability to consolidate a plurality of models and weights, as taught by Gunes, because consolidating various diverse models and corresponding weights would result in a boost to predictive accuracy.).
In regards to claim 3, the combination of Chengdu and Gunes discloses the processing method according to claim 2, further causing the computer to receive check data regarding a part of the plurality of causes, the check data being obtained upon execution of the consolidated model (Chengdu:
Page 4, ¶ 2 “It can be seen that in the embodiments of the present disclosure, the real defect area of the training image and the corresponding real defect category are labeled for comparison with the result of subsequent prediction of the image detection model, so as to realize supervised learning of the image detection model and ensure the detection capability of the image detection model.”
Page 4, ¶ 7 “It can be learned that, in this embodiment of the present disclosure, the training loss of the image detection model is calculated by using the first loss and the second loss, so that the computational comprehensiveness of the training loss can be ensured, so that the learning ability of the image detection model can be supervised in multiple dimensions, and the robustness of the image detection model is improved.”
Page 6, ¶ 5 “S2, inputting the PCB image into a trained image detection model for processing to obtain a defect category and a confidence level of the PCB image.”
Page 6, ¶ 8 “S202, inputting the training image into the image detection model to calculate a defect confidence of a candidate box area in the training”… “S205, if the training loss does not satisfy the preset condition, adjusting parameters of the image detection model, and returning to the step of inputting the training image into the image detection model; S206, if the training loss meets the preset condition, obtaining a trained image detection model.”
Page 7, ¶ 8 “In an embodiment of the present disclosure, the image category decision model includes a machine learning model configured to perform secondary learning on an output result of the image detection model to ensure a detection accuracy of a defect category of the PCB image.”
Page 8, ¶ 2 “It can be seen that, in the embodiments of the present disclosure, the defect category and the confidence of the PCB image are detected by the image detection model, and the defect feature of the PCB image is combined, so that the image category decision model performs secondary learning on the PCB image to correct the defect category judgment result of the image detection model, effectively combines the image detection model with the image category decision model, realizes automatic identification of the defect category of the PCB image, and solves the problem in the prior art that the image detection model is limited by the defect distribution of the training set image and cannot ensure the accuracy of PCB image defect classification. Therefore, the intelligent detection method for the PCB defect category provided by the embodiments of the present disclosure can realize automatic identification of the PCB image defect category, and improve the accuracy of PCB image defect classification.”
Page 8, ¶ 8 “The training image refers to a PCB image that has been subjected to defect category detection. Therefore, in the embodiments of the present disclosure, the real defect area of the training image and the corresponding real defect category are labeled for comparison with the result of subsequent prediction by using the image detection model, so as to realize supervised learning of the image detection model and ensure the detection capability of the image detection model.”).
In regards to claim 4, the combination of Chengdu and Gunes discloses the processing method according to claim 3, further causing the computer to: update a weight of at least a part of the plurality of causes by using the check data (Chengdu:
Page 4, ¶ 2 “It can be seen that in the embodiments of the present disclosure, the real defect area of the training image and the corresponding real defect category are labeled for comparison with the result of subsequent prediction of the image detection model, so as to realize supervised learning of the image detection model and ensure the detection capability of the image detection model.”
Page 4, ¶ 5 “It can be learned that, in this embodiment of the present disclosure, the defect confidence of the candidate box area in the training image is calculated by using the image detection model, and the candidate box area with the largest defect confidence is used as the predicted defect area of the training image, so that the candidate box area and the defect category that meet the maximum weight proportion can be screened out, to ensure the accuracy of obtaining the candidate box area and the defect category.”
Page 6, ¶ 8 “S202, inputting the training image into the image detection model to calculate a defect confidence of a candidate box area in the training”… “S205, if the training loss does not satisfy the preset condition, adjusting parameters of the image detection model, and returning to the step of inputting the training image into the image detection model; S206, if the training loss meets the preset condition, obtaining a trained image detection model.”
Page 7, ¶ 4 “Further, in an optional embodiment of the present disclosure, a parameter of the image detection model includes a weight and a bias, and parameter adjustment of the model may be implemented by a gradient descent algorithm, for example, a stochastic gradient descent algorithm.”
Page 8, ¶ 8 “The training image refers to a PCB image that has been subjected to defect category detection. Therefore, in the embodiments of the present disclosure, the real defect area of the training image and the corresponding real defect category are labeled for comparison with the result of subsequent prediction by using the image detection model, so as to realize supervised learning of the image detection model and ensure the detection capability of the image detection model.”
Gunes:
Page 2 “More generally, you can think about using weighted averages. For example, you might believe that some of the models are better or more accurate and you want to manually assign higher weights to them. But an even better approach might be to estimate these weights more intelligently by using another layer of learning algorithm. This approach is called model stacking.”
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to incorporate into the machine learning based asset detection system and method of Chengdu with the ability to consolidate a plurality of models and weights, as taught by Gunes, because consolidating various diverse models and corresponding weights would result in a boost to predictive accuracy.).
In regards to claim 5, the combination of Chengdu and Gunes discloses the processing method according to claim 4, further causing the computer to: update one or more of the consolidated models in accordance with the updated weight (Chengdu:
Page 4, ¶ 2 “It can be seen that in the embodiments of the present disclosure, the real defect area of the training image and the corresponding real defect category are labeled for comparison with the result of subsequent prediction of the image detection model, so as to realize supervised learning of the image detection model and ensure the detection capability of the image detection model.”
Page 4, ¶ 5 “It can be learned that, in this embodiment of the present disclosure, the defect confidence of the candidate box area in the training image is calculated by using the image detection model, and the candidate box area with the largest defect confidence is used as the predicted defect area of the training image, so that the candidate box area and the defect category that meet the maximum weight proportion can be screened out, to ensure the accuracy of obtaining the candidate box area and the defect category.”
Page 6, ¶ 8 “S202, inputting the training image into the image detection model to calculate a defect confidence of a candidate box area in the training”… “S205, if the training loss does not satisfy the preset condition, adjusting parameters of the image detection model, and returning to the step of inputting the training image into the image detection model; S206, if the training loss meets the preset condition, obtaining a trained image detection model.”
Page 7, ¶ 4 “Further, in an optional embodiment of the present disclosure, a parameter of the image detection model includes a weight and a bias, and parameter adjustment of the model may be implemented by a gradient descent algorithm, for example, a stochastic gradient descent algorithm.”
Page 8, ¶ 8 “The training image refers to a PCB image that has been subjected to defect category detection. Therefore, in the embodiments of the present disclosure, the real defect area of the training image and the corresponding real defect category are labeled for comparison with the result of subsequent prediction by using the image detection model, so as to realize supervised learning of the image detection model and ensure the detection capability of the image detection model.”
Gunes:
Page 2 “More generally, you can think about using weighted averages. For example, you might believe that some of the models are better or more accurate and you want to manually assign higher weights to them. But an even better approach might be to estimate these weights more intelligently by using another layer of learning algorithm. This approach is called model stacking.”
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to incorporate into the machine learning based asset detection system and method of Chengdu with the ability to consolidate a plurality of models and weights, as taught by Gunes, because consolidating various diverse models and corresponding weights would result in a boost to predictive accuracy.).
In regards to claim 8, the combination of Chengdu and Gunes discloses the processing method according to claim 2, wherein each of the consolidated models is generated by using one or more individual models of the plurality of individual models, the one or more individual models corresponding to one or more causes of the plurality of causes, the one or more causes having weights greater than a threshold value (Chengdu:
Page 4, ¶ 5 “It can be learned that, in this embodiment of the present disclosure, the defect confidence of the candidate box area in the training image is calculated by using the image detection model, and the candidate box area with the largest defect confidence is used as the predicted defect area of the training image, so that the candidate box area and the defect category that meet the maximum weight proportion can be screened out, to ensure the accuracy of obtaining the candidate box area and the defect category.”
Page 5, ¶ 6 “An embodiment of the present disclosure provides an intelligent detection method for a PCB defect category.”
Page 6, ¶ 6 “In the embodiments of the present disclosure, the image detection model includes a deep learning model, which is configured to detect a defect category and a confidence level of the PCB image, and the defect category may be understood as a category defect existing in a PCB corresponding to the PCB image, and the confidence level may be understood as a credibility of the detected defect category.”
Page 7, ¶ 4 “Further, in an optional embodiment of the present disclosure, a parameter of the image detection model includes a weight and a bias, and parameter adjustment of the model may be implemented by a gradient descent algorithm, for example, a stochastic gradient descent algorithm.”
Page 7, ¶ 5 “Further, in the embodiment of the present disclosure, the PCB image is input into the trained image detection model to detect the defect category and confidence of the PCB image through the image detection model.”
Page 7, ¶ 8 “In an embodiment of the present disclosure, the image category decision model includes a machine learning model configured to perform secondary learning on an output result of the image detection model to ensure a detection accuracy of a defect category of the PCB image.”
Gunes:
Pages 2, 4, 5, 6 wherein a plurality of models are consolidated to predict a more accurate solution.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to incorporate into the machine learning based asset detection system and method of Chengdu with the ability to consolidate a plurality of models and weights, as taught by Gunes, because consolidating various diverse models and corresponding weights would result in a boost to predictive accuracy.).
In regards to claim 9, the combination of Chengdu and Gunes discloses the processing method according to claim 2 wherein upon execution of the consolidated model, only one or more individual models of the plurality of individual models are executed, the one or more individual models corresponding to one or more causes of the plurality of causes, the one or more causes having weights greater than a threshold value (Chengdu:
Page 4, ¶ 5 “It can be learned that, in this embodiment of the present disclosure, the defect confidence of the candidate box area in the training image is calculated by using the image detection model, and the candidate box area with the largest defect confidence is used as the predicted defect area of the training image, so that the candidate box area and the defect category that meet the maximum weight proportion can be screened out, to ensure the accuracy of obtaining the candidate box area and the defect category.”
Page 5, ¶ 6 “An embodiment of the present disclosure provides an intelligent detection method for a PCB defect category.”
Page 6, ¶ 6 “In the embodiments of the present disclosure, the image detection model includes a deep learning model, which is configured to detect a defect category and a confidence level of the PCB image, and the defect category may be understood as a category defect existing in a PCB corresponding to the PCB image, and the confidence level may be understood as a credibility of the detected defect category.”
Page 7, ¶ 4 “Further, in an optional embodiment of the present disclosure, a parameter of the image detection model includes a weight and a bias, and parameter adjustment of the model may be implemented by a gradient descent algorithm, for example, a stochastic gradient descent algorithm.”
Page 7, ¶ 5 “Further, in the embodiment of the present disclosure, the PCB image is input into the trained image detection model to detect the defect category and confidence of the PCB image through the image detection model.”
Page 7, ¶ 8 “In an embodiment of the present disclosure, the image category decision model includes a machine learning model configured to perform secondary learning on an output result of the image detection model to ensure a detection accuracy of a defect category of the PCB image.”
Gunes:
Pages 2, 4, 5, 6 wherein a plurality of models are consolidated to predict a more accurate solution.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to incorporate into the machine learning based asset detection system and method of Chengdu with the ability to consolidate a plurality of models and weights, as taught by Gunes, because consolidating various diverse models and corresponding weights would result in a boost to predictive accuracy.).
In regards to claim 10, the combination of Chengdu and Gunes discloses the processing method according to claim 1, wherein the plurality of weights are determined by using one or more selected from a first group consisting of a score based on a number of occurrences of each of the plurality of causes, a score based on a configuration of a production line of the product, a score based on a scale of staff related to production of the product, a score based on a repair man-hours for each of the plurality of defect modes, a score based on a price of a component of the product, a score based on a difficulty of production of the product, a score based on a type of a component included in the product, a score based on a production environment of the product, a score based on a number of years elapsed of manufacturing equipment for the product, and a score based on a time of an occurrence of each of the plurality of causes (Chengdu: Page 4, ¶ 5 “It can be learned that, in this embodiment of the present disclosure, the defect confidence of the candidate box area in the training image is calculated by using the image detection model, and the candidate box area with the largest defect confidence is used as the predicted defect area of the training image, so that the candidate box area and the defect category that meet the maximum weight proportion can be screened out, to ensure the accuracy of obtaining the candidate box area and the defect category.”
Page 4, ¶ 7 “It can be learned that, in this embodiment of the present disclosure, the training loss of the image detection model is calculated by using the first loss and the second loss, so that the computational comprehensiveness of the training loss can be ensured, so that the learning ability of the image detection model can be supervised in multiple dimensions, and the robustness of the image detection model is improved.”
Page 6, ¶ 8 “S202, inputting the training image into the image detection model to calculate a defect confidence of a candidate box area in the training”… “S205, if the training loss does not satisfy the preset condition, adjusting parameters of the image detection model, and returning to the step of inputting the training image into the image detection model; S206, if the training loss meets the preset condition, obtaining a trained image detection model.”
Page 7, ¶ 15 “For example, the selected sampling feature is a board surface of the PCB image, and the information dimension of the sampling feature includes a defect area, a defect height/width, a board surface residual copper, and a board surface scratch/scratch, and the board surface residual copper segmentation point of the sampling feature may be determined as whether the board surface residual copper exceeds a preset threshold, so as to obtain whether the feature segmentation node of the sampling feature is whether the board surface residual copper exceeds a preset threshold”).
In regards to claim 11, the combination of Chengdu and Gunes discloses the processing method according to claim 10, wherein the plurality of weights are obtained by inputting the one or more selected from the first group into a neural network (Chengdu: Page 6, ¶ 10 “Further, in another optional embodiment of the present disclosure, the extraction of the feature image may be implemented by a convolution kernel in the convolutional layer, and is configured to extract feature information of the candidate box region, improve a processing speed of a subsequent image, and the pooling processing of the feature image may be implemented by an activation function of the pooling layer, for example, a relu function, which is configured to implement dimensionality reduction of the feature image, and defect recognition of the feature image may be implemented by a fully convolutional neural network of the fully connected layer, where
the activation function includes a softmax function.”).
______________________________________________________________________
Claims 6, 7 are rejected under 35 U.S.C. 103 as being unpatentable over Chengdu Shulian Cloud Computing Technology CO Ltd. (CN 113781485), hereinafter referred to as Chengdu, in view of Gunes (Why do stacked ensemble models win data science competitions) in further view of Vasquez-Ingelmo et al. (Taking advantage of the software product line paradigm to generate customized user interfaces for decision-making processes a case study on university employability).
In regards to claim 6, the combination of Chengdu and Gunes discloses a system and method for identifying defects in a product using machine learning and for consolidating, training, and updating a plurality of machine learning models so that the results produced are more accurate. Although one of ordinary skill in the art would have found that machine learning models exist as computer programs and the combination of Chengdu and Gunes discloses that a user can mix and match different computer models, as well as providing an electronic device with a display to allow a user to interact with the system (Page 8 ¶ 9, 13), thereby making it obvious that a display device and GUI would be inherently included, the combination of Chengdu and Gunes fails to explicitly disclose a display device or user interface to manage the plurality of machine learning models and data.
To be more specific, the combination of Chengdu and Gunes fails to explicitly disclose:
the processing method according to claim 1, further causing the computer to:
cause a display device to display a user interface showing data regarding any of the plurality of individual models; and
accept an edit of the data on the user interface.
However, Vasquez-Ingelmo teaches:
“Information dashboards play a key role in analyzing and visually exploring data about a specific topic or domain, but end users can present several necessities that differ from each other, regarding the displayed information itself, design features and even functionalities. By applying a domain engineering approach (within the software product line paradigm), it is possible to produce customized dashboards to fit into particular requirements, by the identification of commonalities and singularities of every product that could be part of the product line. Software product lines increase productivity, maintainability and traceability regarding the evolution of the requirements, among other benefits.”
(Abstract)
“Nevertheless, even after performing data analysis, identifying patterns and indicators derived from the analysis outcomes remains a complex challenge. That is why it is crucial to assist decision-makers with powerful tools that allow reaching insights about the domain of the problem, to support decisions with complete and quality information (especially in the academic context, where these processes might have a series of social implications), that is, information and knowledge that has been extracted through visual analysis.”
(Page 2)
“Information dashboards are one of the most commonly used software products for visual data analysis and knowledge extraction (Few, 2006; Sarikaya et al., 2018).”
(Page 2)
“The main idea behind software product lines (SPLs) is that the final products can be derived from a set of configurable core assets, allowing their adaptation to fit specific requirements. These core assets are developed during the domain engineering phase, in which commonality and variability of the target product domain are identified to build a common base of components. Core assets are developed with variability points in which specific functionalities could be injected to obtain new products. Functionalities in SPLs are seen as features; the combination of the defined features within the scope of the line (generally following a feature model (Kang et al., 1990) allow stakeholders to build personalized products by reusing and assembling software components.”
(Page 4)
“These practical applications have proved the benefits of this paradigm. However, features usually refer to the software's logic, deflecting attention to the presentation layer. The idea of generating customized dashboards can be seen as a specific case of graphical user interfaces (GUI) automatic generation within SPLs. User interfaces require additional work regarding their implementation; they not only need to be functional but also usable to allow users to complete their tasks efficiently and achieve their goals. That is why the design of user interfaces is present through the whole development process, being time and resource-consuming job.”
(Page 4)
“A control panel will rely on its underlying logic, and it can count on different optional features, like data selectors to dynamically change the visualization's presented data; in case of the X and Y axes, these selectors could be located within the control panel space or in-place controls (i.e., situated near the scatter diagram axes).”
(Page 9)
“Dashboards are complex software solutions that could be highly beneficial when adequately designed and tailored for specific users. These products can support decision-making processes, assisting visual analysis by presenting information in an understandable manner. However, the variety of profiles involved in these processes and their different definitions of “understandable'' makes the implementation of dashboards a time- and resource-consuming task, since a dashboard configuration that is highly useful for one user could be pointless for the rest of them. What is more, dashboards can be composed of several elements, from simple visualizations to different linked views, cross-filtering capabilities, interaction methods, handlers, etc., thus making the dashboards' domain a complex domain not only because of the different profiles of potential users, but because of the great quantity of feasible combinations of these “dashboard elements'' to build a proper solution. In addition, these features can be very fine-grained; in user-centered systems, a slight modification on visualization types, interaction patterns, layouts, color palettes, etc. could be crucial regarding the final perceived usability of the product.”
(Page 21)
“Relying on a framework to easily generate information dashboards would allow stakeholders to focus on the information requirements and their refinement to provide better results when seeking valuable insights on large datasets. Also, it opens up the possibility to automatically adapt the dashboards' configurations to match dynamic requirements based on the device used (Cruz-Benito et al., 2018b) or other factors.”
(Page 21)
“Having powerful visualization tools allow reaching insights about patterns or factors to guide the execution of more complex analyses and make decisions about the actions to take or the future research directions, like developing machine learning (ML) models (García-Peñalvo et al., 2018). Regarding this last field, having visualization tools to explore the input data before training any ML model could help to build better and more accurate models through an appropriate feature selection phase guided by the previously reached insights (Hall, 1999).”
(Page 23
“Automating the generation of dashboards given their goal, their context, their end users, etc. could be extremely beneficial due to the vast potential of impact that these tools have (Sarikaya et al., 2018).”
(Page 23)
“A domain engineering approach has been applied to the dashboards' domain to obtain a SPL of this type of software solution. By the identification of commonalities and variability points, a dashboard meta-model has been developed as well as a feature model to capture the different customization dimensions of the SPL.”
(Page 23)
One of ordinary skill in the art would have found it obvious to provide a user device and user interface that allows a user to visualize and dynamically modify computer code, such as, machine-learning models, because “Having powerful visualization tools allow reaching insights about patterns or factors to guide the execution of more complex analyses and make decisions about the actions to take or the future research directions, like developing machine learning (ML) models (García-Peñalvo et al., 2018),” and “Automating the generation of dashboards given their goal, their context, their end users, etc. could be extremely beneficial due to the vast potential of impact that these tools have (Sarikaya et al., 2018).”
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to incorporate into the consolidation machine learning product detection system and method of the combination of Chengdu and Gunes with the ability to provide a display device and user interface to allow for the visualization and modification of computer code, such as machine learning, as taught by Vasquez-Ingelmo, because “Having powerful visualization tools allow reaching insights about patterns or factors to guide the execution of more complex analyses and make decisions about the actions to take or the future research directions, like developing machine learning (ML) models (García-Peñalvo et al., 2018),” and “Automating the generation of dashboards given their goal, their context, their end users, etc. could be extremely beneficial due to the vast potential of impact that these tools have (Sarikaya et al., 2018).”
In regards to claim 7, the combination of Chengdu and Gunes discloses the combination of Chengdu and Gunes discloses a system and method for identifying defects in a product using machine learning and for consolidating, training, and updating a plurality of machine learning models so that the results produced are more accurate. Although one of ordinary skill in the art would have found that machine learning models exist as computer programs and the combination of Chengdu and Gunes discloses that a user can mix and match different computer models, thereby making it obvious that a display device and GUI would be inherently included, the combination of Chengdu and Gunes fails to explicitly disclose a display device or user interface to manage the machine learning models and data.
To be more specific, the combination of Chengdu and Gunes fails to explicitly disclose:
the processing method according to claim 1, further causing the computer to: upon execution of the consolidated model, cause a display device to display one or more individual models of the plurality of individual models, the one or more individual models being rearranged.
However, Vasquez-Ingelmo teaches:
“Information dashboards play a key role in analyzing and visually exploring data about a specific topic or domain, but end users can present several necessities that differ from each other, regarding the displayed information itself, design features and even functionalities. By applying a domain engineering approach (within the software product line paradigm), it is possible to produce customized dashboards to fit into particular requirements, by the identification of commonalities and singularities of every product that could be part of the product line. Software product lines increase productivity, maintainability and traceability regarding the evolution of the requirements, among other benefits.”
(Abstract)
“Nevertheless, even after performing data analysis, identifying patterns and indicators derived from the analysis outcomes remains a complex challenge. That is why it is crucial to assist decision-makers with powerful tools that allow reaching insights about the domain of the problem, to support decisions with complete and quality information (especially in the academic context, where these processes might have a series of social implications), that is, information and knowledge that has been extracted through visual analysis.”
(Page 2)
“Information dashboards are one of the most commonly used software products for visual data analysis and knowledge extraction (Few, 2006; Sarikaya et al., 2018).”
(Page 2)
“The main idea behind software product lines (SPLs) is that the final products can be derived from a set of configurable core assets, allowing their adaptation to fit specific requirements. These core assets are developed during the domain engineering phase, in which commonality and variability of the target product domain are identified to build a common base of components. Core assets are developed with variability points in which specific functionalities could be injected to obtain new products. Functionalities in SPLs are seen as features; the combination of the defined features within the scope of the line (generally following a feature model (Kang et al., 1990) allow stakeholders to build personalized products by reusing and assembling software components.”
(Page 4)
“These practical applications have proved the benefits of this paradigm. However, features usually refer to the software's logic, deflecting attention to the presentation layer. The idea of generating customized dashboards can be seen as a specific case of graphical user interfaces (GUI) automatic generation within SPLs. User interfaces require additional work regarding their implementation; they not only need to be functional but also usable to allow users to complete their tasks efficiently and achieve their goals. That is why the design of user interfaces is present through the whole development process, being time and resource-consuming job.”
(Page 4)
“A control panel will rely on its underlying logic, and it can count on different optional features, like data selectors to dynamically change the visualization's presented data; in case of the X and Y axes, these selectors could be located within the control panel space or in-place controls (i.e., situated near the scatter diagram axes).”
(Page 9)
“Dashboards are complex software solutions that could be highly beneficial when adequately designed and tailored for specific users. These products can support decision-making processes, assisting visual analysis by presenting information in an understandable manner. However, the variety of profiles involved in these processes and their different definitions of “understandable'' makes the implementation of dashboards a time- and resource-consuming task, since a dashboard configuration that is highly useful for one user could be pointless for the rest of them. What is more, dashboards can be composed of several elements, from simple visualizations to different linked views, cross-filtering capabilities, interaction methods, handlers, etc., thus making the dashboards' domain a complex domain not only because of the different profiles of potential users, but because of the great quantity of feasible combinations of these “dashboard elements'' to build a proper solution. In addition, these features can be very fine-grained; in user-centered systems, a slight modification on visualization types, interaction patterns, layouts, color palettes, etc. could be crucial regarding the final perceived usability of the product.”
(Page 21)
“Relying on a framework to easily generate information dashboards would allow stakeholders to focus on the information requirements and their refinement to provide better results when seeking valuable insights on large datasets. Also, it opens up the possibility to automatically adapt the dashboards' configurations to match dynamic requirements based on the device used (Cruz-Benito et al., 2018b) or other factors.”
(Page 21)
“Having powerful visualization tools allow reaching insights about patterns or factors to guide the execution of more complex analyses and make decisions about the actions to take or the future research directions, like developing machine learning (ML) models (García-Peñalvo et al., 2018). Regarding this last field, having visualization tools to explore the input data before training any ML model could help to build better and more accurate models through an appropriate feature selection phase guided by the previously reached insights (Hall, 1999).”
(Page 23
“Automating the generation of dashboards given their goal, their context, their end users, etc. could be extremely beneficial due to the vast potential of impact that these tools have (Sarikaya et al., 2018).”
(Page 23)
“A domain engineering approach has been applied to the dashboards' domain to obtain a SPL of this type of software solution. By the identification of commonalities and variability points, a dashboard meta-model has been developed as well as a feature model to capture the different customization dimensions of the SPL.”
(Page 23)
One of ordinary skill in the art would have found it obvious to provide a user device and user interface that allows a user to visualize and dynamically modify computer code, such as, machine-learning models, because “Having powerful visualization tools allow reaching insights about patterns or factors to guide the execution of more complex analyses and make decisions about the actions to take or the future research directions, like developing machine learning (ML) models (García-Peñalvo et al., 2018),” and “Automating the generation of dashboards given their goal, their context, their end users, etc. could be extremely beneficial due to the vast potential of impact that these tools have (Sarikaya et al., 2018).”
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the invention to incorporate into the consolidation machine learning product detection system and method of the combination of Chengdu and Gunes with the ability to provide a display device and user interface to allow for the visualization and modification of computer code, such as machine learning, as taught by Vasquez-Ingelmo, because “Having powerful visualization tools allow reaching insights about patterns or factors to guide the execution of more complex analyses and make decisions about the actions to take or the future research directions, like developing machine learning (ML) models (García-Peñalvo et al., 2018),” and “Automating the generation of dashboards given their goal, their context, their end users, etc. could be extremely beneficial due to the vast potential of impact that these tools have (Sarikaya et al., 2018).”
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure can be found in the attached PTO-892 Notice of References Cited.
Zhou et al. (US PGPub 2021/0397170 A1); Barhak (US PGPub 2021/0183523 A1); Paguio (US PGPub 2012/0303562 A1); Hsiao (TW 202127014 A): Hsiao (JP 2021119341 A); Lu et al. (CN 114065687 A); Zhuo et al. (CN 112669275 A); Luo et al. (CN 105372581 A); Huang et al. (A PCB Dataset for Defects Detection and Classification); Kim et al. (Printed Circuit Board Defect Detection Using Deep Learning via A Skip-Connected Convolutional Autoencoder); Wu et al. (PCB Defect Detection Using Deep Learning Methods) – which are directed towards monitoring the condition of an asset, e.g., printed circuit board (PCB), with the aid of machine learning
Mohandas (Dashboard) – which discusses the use of a dashboards
Any inquiry concerning this communication or earlier communications from the examiner should be directed to GERARDO ARAQUE JR whose telephone number is (571)272-3747. The examiner can normally be reached Monday - Friday 8-4:30.
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, Sarah Monfeldt can be reached at 571-270-1833. 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.
GERARDO ARAQUE JR
Primary Examiner
Art Unit 3629
/GERARDO ARAQUE JR/Primary Examiner, Art Unit 3629 10/5/2026
