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 .
This action is in response to amendments filed December 5th, 2025, in which claims 1, 3 and 13-17 have been amended. No claims have been cancelled nor added. The amendments have been entered, and claims 1-17 are currently pending in the case. Claims 1, 13 and 15 are independent claims.
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.
Claims 1-17 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
Regarding claim 1:
Step 1: Claim 1 is directed to a method, therefore it falls under the statuary category of a process.
Step 2A Prong 1: The claim recites, in part:
“monitoring…terminal resources of the plurality of terminals” this encompasses the mental monitoring of observed terminals
“allocating subclassification models…according to monitoring results of the terminal resources of the plurality of terminals” this encompasses the mental allocation of models to terminals according to observed monitoring results.
“generate confidence values for classes allocated to the subclassification models as classification data for the target data” this encompasses the mental creation of confidence values for observed classes.
“determining…a final class for the target data using the classification data” this encompasses the mental determination of a class from observed data.
“classes allocated for classification…include a subset of a plurality of target classes allocated to a main classification model” this encompasses the mental allocation of classes and target classes.
“the number of classes allocated…is less than the number of target classes allocated to a main classification model” This encompasses the mental allocation of fewer classes than target classes.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “performed by one of the plurality of terminals or a server” in lines 5, 12 and 15 of the claim the limitations are an additional element that amounts to adding the words “apply it” (or an equivalent) with the judicial exception, or merely uses a computer in its ordinary capacity as a tool to perform an existing process. See MPEP § 2106.05(f)(2). “receiving classification data for the target data generated through the subclassification models each distributed to a plurality of terminals” the limitation is an additional element that amounts to adding insignificant extra-solution activity to the judicial exception. See MPEP §§ 2106.04(d), 2106.05(g). “wherein the terminal resources include an amount of computation, a memory size and a network usage of the plurality of terminals”, “the subclassification models are artificial neural networks”, “to each of the plurality of terminals”, “to each of the subclassification models, which have been trained in advance”, “to each of the subclassification models” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “performed by one of the plurality of terminals or a server” in lines 5, 12 and 15 of the claim the limitations are an additional element that amounts to adding the words “apply it” (or an equivalent) with the judicial exception, or merely uses a computer in its ordinary capacity as a tool to perform an existing process. See MPEP § 2106.05(f)(2). “wherein the terminal resources include an amount of computation, a memory size and a network usage of the plurality of terminals”, “the subclassification models are artificial neural networks”, “to each of the subclassification models, which have been trained in advance”, “to each of the subclassification models” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 2, the rejection of claim 1 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“classes allocated…further include other class that is a class other than the target classes” this encompasses the mental allocation of classes other than a target class.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “to each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “to each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 3, the rejection of claim 1 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“classes allocated for classification…include the plurality of target classes” this encompasses the mental allocation of class.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “the subclassification models are lightweight models from the main classification model”, “the main classification model” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “the subclassification models are lightweight models from the main classification model”, “the main classification model” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 4, the rejection of claim 1 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“the classes allocated…include different target classes.” This encompasses the mental allocation of classes which include different target classes.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 5, the rejection of claim 1 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“the classes allocated…are determined according to an upper concept of the target classes” this encompasses the mental allocation of classes according to an observed upper concept.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 6, the rejection of claim 1 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“classes allocated…include at least one overlapping target class” this encompasses the mental allocation of classes including overlapping target classes.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “first and second subclassification models among the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “first and second subclassification models among the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 7, the rejection of claim 6 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“the classification data includes confidence values for the classes allocated” this limitation is a mathematical concept.
“the determining of the final class comprises determining the final class using a largest one of the confidence values of the overlapping target class or an average value of the confidence values of the overlapping target class” this encompasses the mental determination of a final class based on the largest observed confidence value, or an average of them. Further, this limitation is a mathematical concept.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 8, the rejection of claim 2 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“the classification data includes confidence values for the classes allocated” this limitation is a mathematical concept.
“the determining of the final class comprises determining a class corresponding to a largest one of the confidence values as the final class.” this encompasses the mental determination of a final class based on the largest observed confidence value. Further, this limitation is a mathematical concept.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 9, the rejection of claim 8 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“wherein the classification data includes confidence values for the target classes among the classes allocated” this encompasses the mental allocation of confidence values for observed target classes. Further, this limitation is a mathematical concept.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitation “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 10, the rejection of claim 2 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“the classification data includes a largest one of confidence values for the classes allocated” this encompasses the mental allocation of confidence values for classes, including a largest value. Further, this limitation is a mathematical concept.
“the determining of the final class comprises determining the final class using the largest value” this encompasses the mental determination of a final class based on observed confidence values, further this limitation is a mathematical concept.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitation “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 11, the rejection of claim 10 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“the classification data includes confidence values for the target classes among the classes allocated” this encompasses the mental allocation of confidence values for target classes. Further, this limitation is a mathematical concept.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “each of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 12, the rejection of claim 1 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“the classification data includes confidence values for classes allocated” this encompasses the mental allocation of confidence value for classes. Further, this limitation is a mathematical concept.
“determining of the final class comprises determining a target class not included in the classification data as the final class when the confidence values are smaller than a threshold value” this encompasses the mental determination of a final class not included amongst observed target classes when an observed confidence value is less than a threshold value. Further, this limitation is a mathematical concept.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “some of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “some of the subclassification models” the limitation is an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 13:
Step 1: Claim 13 is directed to a method, therefore it falls under the statuary category of a process.
Step 2A Prong 1: The claim recites, in part:
“monitoring resources” this encompasses the mental monitoring of resources.
“allocating subclassification models…according to a monitoring result” this encompasses the mental allocation of observed models according to an observed monitoring result.
“the allocation of the subclassification models…is dynamically updated in response to changes in the monitored resources” this encompasses changing the mental allocation of observed models according to observed resources.
“Generating…classification data for the target image, wherein the classification data includes confidence values for classes allocated to the subclassification models” this encompasses the mental classification of observed including confidence values.
“determining…a final class for the target image using the classification data” this encompasses the mental determination of a final class for an observed image using observed classification data.
“the classes allocated for classification…include a subset of a plurality of target classes allocated to a main classification model” this encompasses the mental allocation of classes which include target classes.
“a number of classes allocated…is less than a number of target classes allocated to a main classification model” this encompasses the mental allocation of classes which are fewer than the number of target classes.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “by one of the plurality of terminals or a server”, “by the subclassification models”, “by at least one of the plurality of terminals or the server”, “by one of the plurality of terminals or the server” the limitations are an additional element that amounts to adding the words “apply it” (or an equivalent) with the judicial exception, or merely uses a computer in its ordinary capacity as a tool to perform an existing process. See MPEP § 2106.05(f)(2). “wherein the resources of the plurality of terminals include an amount of computation, a memory size and a network usage”, “wherein the subclassification models are artificial neural networks”, “the plurality of terminals”, “the terminals”, “each of the subclassification models, which have been trained in advance”, “each of the subclassification models” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h). “receiving…the classification data for the target image generated through the subclassification models each distributed to the plurality of terminals” the limitation is an additional element that amounts to adding insignificant extra-solution activity to the judicial exception. See MPEP § 2106.05(g).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “by one of the plurality of terminals or a server”, “by the subclassification models”, “by at least one of the plurality of terminals or the server”, “by one of the plurality of terminals or the server” the limitations are an additional element that amounts to adding the words “apply it” (or an equivalent) with the judicial exception, or merely uses a computer in its ordinary capacity as a tool to perform an existing process. See MPEP § 2106.05(f)(2). “wherein the resources of the plurality of terminals include an amount of computation, a memory size and a network usage”, “wherein the subclassification models are artificial neural networks”, “the plurality of terminals”, “the terminals”, “each of the subclassification models, which have been trained in advance”, “each of the subclassification models” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h). “receiving…the classification data for the target image generated through the subclassification models each distributed to the plurality of terminals” the limitation is an additional element that amounts to adding insignificant extra-solution activity to the judicial exception. See MPEP § 2106.05(g). Furthermore the additional element is directed to receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d.
Regarding claim 14, the rejection of claim 13 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“the number of classes allocated…is determined according to available resources of the plurality of terminals” this encompasses the mental allocation of classes according to a observed available resources.
“the classes allocated…further include other class that is a class other than the target classes” this encompasses the mental allocation of classes other than a target class.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “each of the subclassification models”, “each of the subclassification models” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “each of the subclassification models”, “each of the subclassification models” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 15:
Step 1: Claim 15 is directed to a method, therefore it falls under the statuary category of a process.
Step 2A Prong 1: The claim recites, in part:
“monitoring resources” this encompasses the mental monitoring of resources.
“allocating subclassification models…based on the monitoring of the resources” this encompasses the mental allocation of models based on observed resources.
“generate classification data for a target image, and the classification data includes confidence values for classes allocated to the subclassification models” this encompasses the mental creation of classification data for an observed image as well as the creation of related confidence values.
“determining classification terminals for classifying the target image…according to a monitoring result” this encompasses the mental determination of classification terminals for further classification of observed target image according to an observed monitoring result.
“wherein the classes allocated for classification…include a subset of a plurality of target classes allocated to the main classification model” this encompasses the mental allocation of classes which include target classes.
“a number of classes allocated…is less than a number of the target classes allocated to the main classification model” This encompasses the mental allocation of fewer classes then a number of target classes.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “wherein the resources of the plurality of candidate terminals include an amount of computation, a memory size and a network usage”, “a plurality of candidate terminals”, “to the plurality of candidate terminals”, “wherein the subclassification models are artificial neural networks”, “each of the subclassification models, which have been trained in advance”, “each of the subclassification models” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h). “using subclassification models among the candidate terminals” the limitation is an additional element that amounts to adding the words “apply it” (or an equivalent) with the judicial exception, or merely uses a computer in its ordinary capacity as a tool to perform an existing process. See MPEP § 2106.05(f)(2). “receiving the classification data for the target image generated through the subclassification models each distributed to the plurality of candidate terminals” the limitation is an additional element that amounts to adding insignificant extra-solution activity to the judicial exception. See MPEP § 2106.05(g).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “a plurality of candidate terminals”, “each of the subclassification models, which have been trained in advance”, “each of the subclassification models” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 16, the rejection of claim 15 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“determining a transmission terminal…according to the monitoring results” this encompasses the mental determination of a transmission terminal according to an observed monitoring result.
“the classes allocated… further include other class that is a class other than the target classes” This encompasses the mental allocation of an other class, which is other than the target classes.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “for transmitting the target image to” the limitation is an additional element that amounts to adding insignificant extra-solution activity to the judicial exception. See MPEP § 2106.05(g). “the classification terminals among the plurality of candidate terminals”, “each of the subclassification models” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “for transmitting the target image to” the limitation is an additional element that amounts to adding insignificant extra-solution activity to the judicial exception. See MPEP § 2106.05(g). Furthermore the additional element is directed to receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d. See MPEP § 2106.05(d)/(II). “the classification terminals among the plurality of candidate terminals”, “each of the subclassification models” these limitations are an additional element that generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP §2106.05(h).
Regarding claim 17, the rejection of claim 16 is incorporated and further:
Step 2A Prong 1: The claim recites, in part:
“determines a final class for the target image” this encompasses the mental determination of a final class for an observed target image.
“determining a class determination terminal…according to the monitoring results” this encompasses the mental determination of a terminal based on observed monitoring results.
Step 2A Prong 2: The judicial exception is not integrated into a practical application; the remaining limitations of the claim are as follows: “which receives the classification data”, “each distributed to the classification terminals among the candidate terminals” these limitations are an additional element that amounts to adding insignificant extra-solution activity to the judicial exception. See MPEP § 2106.05(g). “generated for the target image through the subclassification models” The limitation is an additional element that amounts to adding the words “apply it” (or an equivalent) with the judicial exception, or merely uses a computer in its ordinary capacity as a tool to perform an existing process. See MPEP § 2106.05(f)(2).
Step 2B: The claim does not contain significantly more than the judicial exception. The limitations “which receives classification the data”, “each distributed to the classification terminals among the candidate terminals” these limitations are an additional element that amounts to adding insignificant extra-solution activity to the judicial exception. See MPEP § 2106.05(g). Furthermore the additional element is directed to receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d. See MPEP § 2106.05(d)/(II). “generated for the target image through the subclassification models” The limitation is an additional element that amounts to adding the words “apply it” (or an equivalent) with the judicial exception, or merely uses a computer in its ordinary capacity as a tool to perform an existing process. See MPEP § 2106.05(f)(2).
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-2 and 4-17 are rejected under 35 U.S.C. 103 as being unpatentable over Ono et al. (U.S. Patent 8,892.494 B2) hereinafter Ono in view of 정부은 et al. (Jung, Boo Eun et al.) (KR 20120121143 A) hereinafter Jung in view of Liu et al. (U.S. Patent 7,809,723 B2) hereinafter Liu in further view of Chong et al. (US 20040064552 A1) hereinafter Chong.
Regarding claim 1:
Ono teaches A computer-implemented classification method…the classification method comprising (Ono, claim 1 “A defect classification apparatus which classifies”):
allocating subclassification models to each of the plurality of terminals according to monitoring results of the terminal resources of the plurality of terminals, wherein the subclassification models are artificial neural networks (Ono, col 1, lines 30-33 “Representative methods of the learning classification technology include discriminant analysis based on a neural network and the Bayes discriminant theory and the like.”),
receiving, …the classification data for the target data (Ono, col 6, lines 31-32 “Then, the classification unit 149 inputs those pieces of attribute data to a main classification model.” The attribute data can be considered the classification data for target data) generated through the subclassification models (Ono, col 6, lines 55-58 “On the other hand, the sub classification model has the role of detailed classification, and hence the learning classification model is suitable.”)…; and
determining, …a final class for the target data using the classification data (Ono, col 6, lines 38-40 “Next, the classification unit 149 inputs the attribute data to a sub classification model that is modeled for each of the main classes, and classifies the defects into detailed sub classes.” here, the detailed sub class for the data can be considered the final class for the target data, in light of the specification, page 11, ¶3 “The target data for which the final class is determined may be used for training a subclassification model to which the final class is allocated.”),
wherein the classes allocated for classification to each of the subclassification models, which have been trained in advance (Ono, col 1, lines 27-30 “In the learning classification technology, image data for learning is collected in advance and learned, to thereby optimize a classification model.” Here, model learning done in advance can be considered training), include a subset at least one of a plurality of target classes (Ono, Fig. 7 shows subclassification models, and each classifies data into a plurality of sub classes which can be considered the target classes) allocated to a main classification model, and
wherein the number of classes allocated to each of the subclassification models is less than the number of target classes allocated to the main classification model (Ono, Fig. 7 shows subclassification models, each of which classifies data into a number of sub classes which can be considered the target classes, the number of classes allocated to each subclassification model (A-D) is less than a number of classes allocated to the main classification model).
Ono does not teach "…for enhancing classification accuracy of target data across network computing terminals by processing computation of a classification model in a distributed manner by splitting the computation to a plurality of terminals,
monitoring, performed by one of the plurality of terminals or a server, terminal resources of the plurality of terminals, wherein the terminal resources include an amount of computation…;
…performed by at least one of the plurality of terminals or the server,
…performed by at least one of the plurality of terminals or the server,
…performed by one of the plurality of terminals or the server,
…each distributed to the plurality of terminals"
However, Jung teaches for enhancing classification accuracy of target data across network computing terminals by processing computation of a classification model in a distributed manner by splitting the computation to a plurality of terminals (Jung, ¶48 “In this way, the present invention divides one inspection target image into a plurality of sub-images and allocates the divided sub-images to each unit processing unit (430a to 430n) in consideration of the resource usage of the unit processing units (430a to 430n), thereby improving the accuracy of inspection target image analysis and shortening the time required for analysis of the inspection target image.”),
monitoring, performed by one of the plurality of terminals or a server, terminal resources of the plurality of terminals, wherein the terminal resources include an amount of computation (Jung, ¶51 “In another embodiment, the resource monitoring unit (640) can monitor the CPU usage of each unit processing unit (430a to 430n) and provide the result to the subimage allocation unit (630).”)…;
…performed by at least one of the plurality of terminals or the server (Jung, ¶41 “Referring again to FIG. 4, the distributed processing server (420) divides the inspection target image transmitted from the analysis server (410) into a plurality of sub-images and assigns each sub-image to a unit processing unit (430a to 430n), thereby enabling analysis of the inspection target image to be distributedly processed by a plurality of unit processing units (430a to 430n)” here, the unit processing units can be considered the terminals),
…performed by at least one of the plurality of terminals or the server (Jung, ¶41 “Referring again to FIG. 4, the distributed processing server (420) divides the inspection target image transmitted from the analysis server (410) into a plurality of sub-images and assigns each sub-image to a unit processing unit (430a to 430n), thereby enabling analysis of the inspection target image to be distributedly processed by a plurality of unit processing units (430a to 430n)” here, the unit processing units can be considered the terminals),
…performed by one of the plurality of terminals or the server (Jung, ¶41 “Referring again to FIG. 4, the distributed processing server (420) divides the inspection target image transmitted from the analysis server (410) into a plurality of sub-images and assigns each sub-image to a unit processing unit (430a to 430n), thereby enabling analysis of the inspection target image to be distributedly processed by a plurality of unit processing units (430a to 430n)” here, the unit processing units can be considered the terminals),
…each distributed to the plurality of terminals (Jung, ¶41 “Referring again to FIG. 4, the distributed processing server (420) divides the inspection target image transmitted from the analysis server (410) into a plurality of sub-images and assigns each sub-image to a unit processing unit (430a to 430n), thereby enabling analysis of the inspection target image to be distributedly processed by a plurality of unit processing units (430a to 430n)” here, the unit processing units can be considered the terminals)
Ono and Jung are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono’s method to incorporate the terminals taught by Jung. The motivation for doing so would have been to improve the classification speed by processing on distributed terminals as stated in Jung, ¶6 “based on distributed processing that can minimize the time required for diagnosis of an image to be examined.”.
Ono in view of Jung does not teach "generate confidence values for classes allocated to the subclassification models as classification data for the target data"
However, Liu teaches generate confidence values for classes allocated to the subclassification models as classification data for the target data (Liu, col 4-5, lines 66-2 “A binary classifier for a classification classifies documents as either being in or not in that classification with a certain confidence.” The certain confidence can be considered the confidence values);
Ono in view of Jung and Liu are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs’s classification method to incorporate the confidence values taught by Liu. The motivation for doing so would have been to incorporate the use of confidence scores and thresholds to ensure data is classified correctly. Liu, col 3, lines 45-47 “The training system may, for each classification, train and cross validate multiple classifiers and select a confidence threshold for each classifier.”
Ono in view of Jung in further view of Liu does not teach "…a memory size and a network usage of the plurality of terminals "
However, Chong teaches …a memory size and a network usage of the plurality of terminals (Chong, ¶45 “System resources indicates such information as the amount of available memory and number of available connections.”)
Ono in view of Jung in further view of Liu and Chong are analogous art because both references concern methods for distributed processing. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs/Liu’s classification method to incorporate the resource monitoring taught by Chong. The motivation for doing so would have been to incorporate the resource monitoring taught by Chong, as stated in Chong, ¶45 “The computer program of the invention is particularly useful for applications running on application servers.”
Regarding claim 2:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 1
Ono in view of Jung does not teach “wherein the classes allocated to each of the subclassification models further include other class that is a class other than the target classes”
However Liu teaches wherein the classes allocated to each of the subclassification models further include other class that is a class other than the target classes (Liu, claim 11 “ The computer system of claim 10 wherein a classifier is a binary classifier for a classification that classifies documents within that classification as being within or not within a sub-classification.” here, by being not within a sub-classification they can be considered an other class).
Ono in view of Jung and Liu are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs’ classification method to incorporate the other class through binary classification thresholds taught by Liu. The motivation for doing so would have been to indicate a minimum confidence of a classification belonging to a class. Liu, col 6, lines 38-40 “A confidence threshold for a classification indicates the minimum confidence score to be used for classifying a document into that classification.”
Regarding claim 4:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 1, wherein the classes allocated to each of the subclassification models include different target classes (Ono, fig. 7 shows that each main class allocated to the sub classification models includes different sub or “target” classes).
Regarding claim 5:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 1, wherein the classes allocated to each of the subclassification models are determined according to an upper concept of the target classes (Ono, col 6, lines 31-40 “Then, the classification unit 149 inputs those pieces of attribute data to a main classification model. The main classification model classifies the defects based on the input attribute data. In the example of FIG. 7, a large number of defects are classified into four types of main classes A, B, C, and D.
Next, the classification unit 149 inputs the attribute data to a sub classification model that is modeled for each of the main classes, and classifies the defects into detailed sub classes” here, the sub classification models are determined according to an “upper concept” which is one of the main classes A-D).
Regarding claim 6:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 1, wherein classes allocated to first and second subclassification models among the subclassification models include at least one overlapping target class (Ono, fig 7 shows sub classification models A and B, “a first and second subclassification model”, include sub classes 2 and 5 “overlapping target classes”).
Regarding claim 7:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 6, wherein the classification data includes confidence values for the classes allocated to each of the subclassification models (Liu, col 4-5, lines 66-2 “A binary classifier for a classification classifies documents as either being in or not in that classification with a certain confidence.” The certain confidence can be considered the confidence values), and
the determining of the final class comprises determining the final class using a largest one of the confidence values of the overlapping target class (Liu, col 5, lines 12-17 “ For example, when a document is classified with the sports classification, the hierarchical classifier applies to the document the binary classifiers for the baseball and football classifications. The hierarchical classifier then selects the classification whose binary classifier indicates the highest confidence level of being within the classification.” Here, baseball and football can be considered the overlapping target class). It is noted the claim recites alternative language, and Liu teaches at least one of the alternatives.
It would have been obvious to combine the teachings of Ono, Jung, Chong and Liu for the reasons set forth in connection with claim 1 above.
Regarding claim 8:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 2, wherein the classification data includes confidence values for the classes allocated to each of the subclassification models (Liu, col 4-5, lines 66-2 “A binary classifier for a classification classifies documents as either being in or not in that classification with a certain confidence.”), and
the determining of the final class comprises determining a class corresponding to a largest one of the confidence values as the final class (Liu, col 5, lines 15-17 “The hierarchical classifier then selects the classification whose binary classifier indicates the highest confidence level of being within the classification.”).
Ono in view of Jung and Liu are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs’ classification method to incorporate the confidence values taught by Liu. The motivation for doing so would have been to incorporate the use of confidence scores and thresholds to ensure data is classified correctly. Liu, col 3, lines 45-47 “The training system may, for each classification, train and cross validate multiple classifiers and select a confidence threshold for each classifier.”
Regarding claim 9:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 8, wherein the classification data includes confidence values for the target classes among the classes allocated to each of the subclassification models (Liu, col 5, lines 9-15 “To determine the correct sub-classification for a document with a certain classification, a hierarchical classifier using binary classifiers applies each classifier for the Sub classifications. For example, when a document is classified with the sports classification, the hierarchical classifier applies to the document the binary classifiers for the baseball and football classifications.”).
It would have been obvious to combine the teachings of Ono in view of Jung and Liu for the reasons set forth in connection with claim 8 above.
Regarding claim 10:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 2, wherein the classification data includes a largest one of confidence values for the classes allocated to each of the subclassification models (Liu, col 4-5, lines 66-2 “A binary classifier for a classification classifies documents as either being in or not in that classification with a certain confidence.” Further, Liu, col 5, lines 9-15 “To determine the correct sub-classification for a document with a certain classification, a hierarchical classifier using binary classifiers applies each classifier for the Sub classifications. For example, when a document is classified with the sports classification, the hierarchical classifier applies to the document the binary classifiers for the baseball and football classifications.” by applying the classifier for all classes, the data will include the largest of those values), and
the determining of the final class comprises determining the final class using the largest value (Liu, col 5, lines 15-17 “The hierarchical classifier then selects the classification whose binary classifier indicates the highest confidence level of being within the classification.”).
Ono in view of Jung and Liu are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs’ classification method to incorporate the confidence values taught by Liu. The motivation for doing so would have been to incorporate the use of confidence scores and thresholds to ensure data is classified correctly. Liu, col 3, lines 45-47 “The training system may, for each classification, train and cross validate multiple classifiers and select a confidence threshold for each classifier.”
Regarding claim 11:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 10, wherein the classification data includes confidence values for the target classes among the classes allocated to each of the subclassification models (Liu, col 5, lines 9-15 “To determine the correct sub-classification for a document with a certain classification, a hierarchical classifier using binary classifiers applies each classifier for the Sub classifications. For example, when a document is classified with the sports classification, the hierarchical classifier applies to the document the binary classifiers for the baseball and football classifications.”).
It would have been obvious to combine the teachings of Ono in view of Jung and Liu for the reasons set forth in connection with claim 10 above.
Regarding claim 12:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 1 wherein the classification data includes confidence values for classes allocated to some of the subclassification models (Liu, col 4-5, lines 66-2 “A binary classifier for a classification classifies documents as either being in or not in that classification with a certain confidence.”), and
the determining of the final class comprises determining a target class not included in the classification data as the final class when the confidence values are smaller than a threshold value (Liu, col 6, lines 38-40 “A confidence threshold for a classification indicates the minimum confidence score to be used for classifying a document into that classification” Here, if the value is smaller than the confidence threshold, that classification is not final class for the data and it is a different target class).
Ono in view of Jung and Liu are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs’ classification method to incorporate the confidence values taught by Liu. The motivation for doing so would have been to incorporate the use of confidence scores and thresholds to ensure data is classified correctly. Liu, col 3, lines 45-47 “The training system may, for each classification, train and cross validate multiple classifiers and select a confidence threshold for each classifier.”
Regarding claim 13:
Ono teaches A computer-implemented classification method…the classification method comprising (Ono, claim 1 “A defect classification apparatus which classifies”):
allocating subclassification models to the plurality of terminals according to a monitoring result wherein the subclassification models are artificial neural networks (Ono, col 1, lines 30-33 “Representative methods of the learning classification technology include discriminant analysis based on a neural network and the Bayes discriminant theory and the like.”),
receiving, …the classification data for the target image (Ono, col 6, lines 31-32 “Then, the classification unit 149 inputs those pieces of attribute data to a main classification model.” The attribute data can be considered the classification data for target image) generated through the subclassification models (Ono, col 6, lines 55-58 “On the other hand, the sub classification model has the role of detailed classification, and hence the learning classification model is suitable.”)…; and
determining, …a final class for the target image using the classification data (Ono, col 6, lines 38-40 “Next, the classification unit 149 inputs the attribute data to a sub classification model that is modeled for each of the main classes, and classifies the defects into detailed sub classes.” here, the detailed sub class for the data can be considered the final class for the target image, in light of the specification, page 11, ¶3 “The target data for which the final class is determined may be used for training a subclassification model to which the final class is allocated.”),
wherein the classes allocated for classification to each of the subclassification models, which have been trained in advance (Ono, col 1, lines 27-30 “In the learning classification technology, image data for learning is collected in advance and learned, to thereby optimize a classification model.” Here, model learning done in advance can be considered training), include a subset at least one of a plurality of target classes (Ono, Fig. 7 shows subclassification models, and each classifies data into a plurality of sub classes which can be considered the target classes) allocated to a main classification model, and
wherein the number of classes allocated to each of the subclassification models is less than the number of target classes allocated to the main classification model (Ono, Fig. 7 shows subclassification models, each of which classifies data into a number of sub classes which can be considered the target classes, the number of classes allocated to each subclassification model (A-D) is less than a number of classes allocated to the main classification model).
Ono does not teach "…for enhancing classification accuracy of a target image across network computing terminals by processing computation of a classification model in a distributed manner by splitting the computation to a plurality of terminals,
the allocation of the subclassification models to the plurality of terminals is dynamically updated in response to changes in the monitored resources
monitoring, performed by one of the plurality of terminals or a server, resources of a plurality of terminals, wherein the resources of the plurality of terminals include an amount of computation…;
…performed by at least one of the plurality of terminals or the server,
…performed by at least one of the plurality of terminals or the server,
…performed by one of the plurality of terminals or the server,
…each distributed to the plurality of terminals"
However, Jung teaches for enhancing classification accuracy of a target image across network computing terminals by processing computation of a classification model in a distributed manner by splitting the computation to a plurality of terminals (Jung, ¶48 “In this way, the present invention divides one inspection target image into a plurality of sub-images and allocates the divided sub-images to each unit processing unit (430a to 430n) in consideration of the resource usage of the unit processing units (430a to 430n), thereby improving the accuracy of inspection target image analysis and shortening the time required for analysis of the inspection target image.”),
the allocation of the subclassification models to the plurality of terminals is dynamically updated in response to changes in the monitored resources (Jung, ¶48 “In this way, the present invention divides one inspection target image into a plurality of sub-images and allocates the divided sub-images to each unit processing unit (430a to 430n) in consideration of the resource usage of the unit processing units (430a to 430n), thereby improving the accuracy of inspection target image analysis and shortening the time required for analysis of the inspection target image.”)
monitoring, performed by one of the plurality of terminals or a server, resources of a plurality of terminals, wherein the resources of the plurality of terminals include an amount of computation (Jung, ¶51 “In another embodiment, the resource monitoring unit (640) can monitor the CPU usage of each unit processing unit (430a to 430n) and provide the result to the subimage allocation unit (630).”)…;
… by at least one of the plurality of terminals or the server (Jung, ¶41 “Referring again to FIG. 4, the distributed processing server (420) divides the inspection target image transmitted from the analysis server (410) into a plurality of sub-images and assigns each sub-image to a unit processing unit (430a to 430n), thereby enabling analysis of the inspection target image to be distributedly processed by a plurality of unit processing units (430a to 430n)” here, the unit processing units can be considered the terminals),
… by at least one of the plurality of terminals or the server (Jung, ¶41 “Referring again to FIG. 4, the distributed processing server (420) divides the inspection target image transmitted from the analysis server (410) into a plurality of sub-images and assigns each sub-image to a unit processing unit (430a to 430n), thereby enabling analysis of the inspection target image to be distributedly processed by a plurality of unit processing units (430a to 430n)” here, the unit processing units can be considered the terminals),
… by one of the plurality of terminals or the server (Jung, ¶41 “Referring again to FIG. 4, the distributed processing server (420) divides the inspection target image transmitted from the analysis server (410) into a plurality of sub-images and assigns each sub-image to a unit processing unit (430a to 430n), thereby enabling analysis of the inspection target image to be distributedly processed by a plurality of unit processing units (430a to 430n)” here, the unit processing units can be considered the terminals),
…each distributed to the plurality of terminals (Jung, ¶41 “Referring again to FIG. 4, the distributed processing server (420) divides the inspection target image transmitted from the analysis server (410) into a plurality of sub-images and assigns each sub-image to a unit processing unit (430a to 430n), thereby enabling analysis of the inspection target image to be distributedly processed by a plurality of unit processing units (430a to 430n)” here, the unit processing units can be considered the terminals)
Ono and Jung are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono’s method to incorporate the terminals taught by Jung. The motivation for doing so would have been to improve the classification speed by processing on distributed terminals as stated in Jung, ¶6 “based on distributed processing that can minimize the time required for diagnosis of an image to be examined.”.
Ono in view of Jung does not teach "generating, by the subclassification models, classification data for the target image, wherein the classification data includes confidence values for classes allocated to the subclassification models"
However, Liu teaches generating, by the subclassification models, classification data for the target image, wherein the classification data includes confidence values for classes allocated to the subclassification models (Liu, col 4-5, lines 66-2 “A binary classifier for a classification classifies documents as either being in or not in that classification with a certain confidence.” The certain confidence can be considered the confidence values);
Ono in view of Jung and Liu are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs’s classification method to incorporate the confidence values taught by Liu. The motivation for doing so would have been to incorporate the use of confidence scores and thresholds to ensure data is classified correctly. Liu, col 3, lines 45-47 “The training system may, for each classification, train and cross validate multiple classifiers and select a confidence threshold for each classifier.”
Ono in view of Jung in further view of Liu does not teach "…a memory size and a network usage"
However, Chong teaches …a memory size and a network usage (Chong, ¶45 “System resources indicates such information as the amount of available memory and number of available connections.”)
Ono in view of Jung in further view of Liu and Chong are analogous art because both references concern methods for distributed processing. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs/Liu’s classification method to incorporate the resource monitoring taught by Chong. The motivation for doing so would have been to incorporate the resource monitoring taught by Chong, as stated in Chong, ¶45 “The computer program of the invention is particularly useful for applications running on application servers.”
Regarding claim 14:
Ono in view of Jung in view of Liu in further view of Chong teaches The classification method of claim 13, wherein the number of classes allocated to each of the subclassification models is determined according to available resources of the plurality of terminals (Jung, ¶56 “according to these embodiments, the present invention allocates sub images determined to be similar to disease images to each unit processing unit (430a to 430n) by considering the resource usage of the unit processing units (430a to 430n)” Here, by allocating similar images to be classified to the processing units according to resource usage, the system is reducing the number of classes per terminal, which can be considered determining a number of classes according to available resources), and the classes allocated to each of the subclassification models further include other class that is a class other than the target classes (Liu, claim 11 “ The computer system of claim 10 wherein a classifier is a binary classifier for a classification that classifies documents within that classification as being within or not within a sub-classification.” here, by being not within a sub-classification they can be considered an other class).
Ono in view of Jung and Liu are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs’ classification method to incorporate the other class through binary classification thresholds taught by Liu. The motivation for doing so would have been to indicate a minimum confidence of a classification belonging to a class. Liu, col 6, lines 38-40 “A confidence threshold for a classification indicates the minimum confidence score to be used for classifying a document into that classification.”
Regarding claim 15:
Ono teaches A non-transitory computer-readable medium storing computer-executable instructions that, when executed by one or more processors, cause a computing system to perform a classification method…, the classification method comprising (Ono, claim 1 “A defect classification apparatus which classifies”):
wherein the subclassification models are artificial neural networks (Ono, col 1, lines 30-33 “Representative methods of the learning classification technology include discriminant analysis based on a neural network and the Bayes discriminant theory and the like.”)
receiving the classification data for the target image generated through the subclassification models (Ono, col 6, lines 55-58 “On the other hand, the sub classification model has the role of detailed classification, and hence the learning classification model is suitable.”)
wherein the classes allocated for classification to each of the subclassification models, which have been trained in advance (Ono, col 1, lines 27-30 “In the learning classification technology, image data for learning is collected in advance and learned, to thereby optimize a classification model.” Here, model learning done in advance can be considered training), include a subset of a plurality of target classes allocated to a main classification model (Ono, Fig. 7 shows subclassification models, and each classifies data into a plurality of sub classes which can be considered the target classes), and
wherein the number of classes allocated to each of the subclassification models is less than the number of the target classes allocated to the main classification model (Ono, Fig. 7 shows subclassification models, each of which classifies data into a number of sub classes which can be considered the target classes, the number of classes allocated to each subclassification model (A-D) is less than a number of classes allocated to the main classification model).
Ono does not teach "by processing computation of a classification model in a distributed manner by splitting the computation to a plurality of terminals
monitoring resources of a plurality of candidate terminals, wherein the resources of the plurality of candidate terminals include an amount of computation…;
allocating subclassification models to the plurality of candidate terminals based on the monitoring of the resources,
each distributed to the plurality of candidate terminals;
determining classification terminals for classifying the target data image using the subclassification models among the plurality of candidate terminals according to a monitoring result"
However, Jung teaches by processing computation of a classification model in a distributed manner by splitting the computation to a plurality of terminals (Jung, ¶48 “In this way, the present invention divides one inspection target image into a plurality of sub-images and allocates the divided sub-images to each unit processing unit (430a to 430n) in consideration of the resource usage of the unit processing units (430a to 430n), thereby improving the accuracy of inspection target image analysis and shortening the time required for analysis of the inspection target image.”)
monitoring resources of a plurality of candidate terminals, wherein the resources of the plurality of candidate terminals include an amount of computation (Jung, ¶51 “In another embodiment, the resource monitoring unit (640) can monitor the CPU usage of each unit processing unit (430a to 430n) and provide the result to the subimage allocation unit (630).”)…;
allocating subclassification models to the plurality of candidate terminals based on the monitoring of the resources(Jung, ¶51 “In another embodiment, the resource monitoring unit (640) can monitor the CPU usage of each unit processing unit (430a to 430n) and provide the result to the sub image allocation unit (630). In accordance with this embodiment, the sub-image allocation unit (630) gives priority to allocating sub-images to unit processing units (430a to 430n) with low CPU usage among the unit processing units (430a to 430n).” here, the result of monitoring CPU usage cam be considered the monitoring result, and terminals are allocated according to that result),
each distributed to the plurality of candidate terminals (Jung, ¶41 “Referring again to FIG. 4, the distributed processing server (420) divides the inspection target image transmitted from the analysis server (410) into a plurality of sub-images and assigns each sub-image to a unit processing unit (430a to 430n), thereby enabling analysis of the inspection target image to be distributedly processed by a plurality of unit processing units (430a to 430n)” here, the unit processing units can be considered the terminals);
determining classification terminals for classifying the target data image using the subclassification models among the plurality of candidate terminals according to a monitoring result (Jung, ¶51 “In another embodiment, the resource monitoring unit (640) can monitor the CPU usage of each unit processing unit (430a to 430n) and provide the result to the sub image allocation unit (630). In accordance with this embodiment, the sub-image allocation unit (630) gives priority to allocating sub-images to unit processing units (430a to 430n) with low CPU usage among the unit processing units (430a to 430n).” here, the result of monitoring CPU usage cam be considered the monitoring result, and terminals are allocated according to that result),
Ono and Jung are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono’s method to incorporate the terminals taught by Jung. The motivation for doing so would have been to improve the classification speed by processing on distributed terminals as stated in Jung, ¶6 “based on distributed processing that can minimize the time required for diagnosis of an image to be examined.”.
Ono in view of Jung does not teach "and generate classification data for a target image, and the classification data includes confidence values for classes allocated to the subclassification models"
However, Liu teaches and generate classification data for a target image, and the classification data includes confidence values for classes allocated to the subclassification models (Liu, col 4-5, lines 66-2 “A binary classifier for a classification classifies documents as either being in or not in that classification with a certain confidence.” The certain confidence can be considered the confidence values); Ono in view of Jung and Liu are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs’s classification method to incorporate the confidence values taught by Liu. The motivation for doing so would have been to incorporate the use of confidence scores and thresholds to ensure data is classified correctly. Liu, col 3, lines 45-47 “The training system may, for each classification, train and cross validate multiple classifiers and select a confidence threshold for each classifier.”
Ono in view of Jung in further view of Liu does not teach "…a memory size and a network usage of the plurality of terminals "
However, Chong teaches …a memory size and a network usage of the plurality of terminals (Chong, ¶45 “System resources indicates such information as the amount of available memory and number of available connections.”)
Regarding claim 16:
Ono in view of Jung teaches The classification method of claim 15, further comprising determining a transmission terminal for transmitting the target image to the classification terminals among the plurality of candidate terminals according to the monitoring results (Jung, ¶46 “Next, the sub-image allocation unit (630) allocates each sub-image to the unit processing units (430a to 430n) in the order of low resource usage rates according to the monitoring results of the resource monitoring unit (640)” here, the sub-image allocation unit can be considered the transmission terminal),
wherein the classes allocated to each of the subclassification models further include other class that is a class other than the target classes (Liu, claim 11 “ The computer system of claim 10 wherein a classifier is a binary classifier for a classification that classifies documents within that classification as being within or not within a sub-classification.” here, by being not within a sub-classification they can be considered an other class).
Ono in view of Jung and Liu are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs’ classification method to incorporate the other class through binary classification thresholds taught by Liu. The motivation for doing so would have been to indicate a minimum confidence of a classification belonging to a class. Liu, col 6, lines 38-40 “A confidence threshold for a classification indicates the minimum confidence score to be used for classifying a document into that classification.”
Regarding claim 17:
Ono in view of Jung in further view of Liu teaches The classification method of claim 16, further comprising determining a class determination terminal, which receives the classification data generated for the target image through the subclassification models each distributed to the classification terminals (Jung, ¶39 “Next, the diagnosis result generation unit (530) generates a diagnosis result for the examination target image by marking an area corresponding to a sub-image determined to be similar to a disease image on the examination target image.” Here, the result generation unit can be considered the “class determination terminal”) and determines a final class for the target image, (Ono, col 6, lines 38-40 “Next, the classification unit 149 inputs the attribute data to a sub classification model that is modeled for each of the main classes, and classifies the defects into detailed sub classes.” here, the detailed sub class for the data can be considered the final class for the target data, in light of the specification, page 11, ¶3 “The target data for which the final class is determined may be used for training a subclassification model to which the final class is allocated.”) among the plurality of candidate terminals according to the monitoring results (Jung, ¶46 “Next, the sub-image allocation unit (630) allocates each sub-image to the unit processing units (430a to 430n) in the order of low resource usage rates according to the monitoring results of the resource monitoring unit (640)”).
It would have been obvious to combine the teachings of Ono and Jung for the reasons set forth in connection with claim 15 above.
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over Ono in view of Jung in view of Liu in view of Chong in further view of Ekambaram et al. (Pub. No. US 2019/0188760 A1) hereinafter Ekambaram.
Regarding claim 3:
Ono in view of Jung teaches The classification method of claim 1, wherein
the classes allocated for classification to the main classification model include the plurality of target classes (Ono, fig 7 shows the main classification model as having main, or “target’, classes A-D allocated to it).
Ono in view of Jung does not teach the subclassification models are lightweight models from a main classification model
However, Ekambaram teaches the subclassification models are lightweight models from a main classification model (Ekambaram, ¶32 “In one embodiment, the decision tree classifier (block 306) is a lightweight version of the classification engine 214 of FIG. 2, which utilizes the learned dataset attribute classification models 216 to classify one or more data categories of the sampled dataset.”).
Ono in view of Jung and Ekambaram are analogous art because both references concern methods for data classification. Accordingly, it would have been obvious to a person of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Ono/Jungs’ classification method to incorporate the lightweight model taught by Ekambaram. The motivation for doing so would have been to have a fast and efficient initial classification, as stated in Ekambaram, ¶48 “an online machine learning API classification process 712-1, which is fast and resource efficient, and a backend machine learning classification process 712-2, which is resource intensive, but provides higher accuracy”.
Response to Arguments
Applicant's arguments filed December 5th, 2025 have been fully considered but they are not persuasive.
Applicant first argues “The revisions clarify that the classification method can be performed in a computing device and further clarifies that such a computer-implemented method is done by processing computation of a classification model in a distributed manner to a plurality of terminals, and is also distributed based on the monitoring of the terminal resources, which may include an amount of computation, a memory size, and a network usage of the plurality of terminals.” The MPEP states “Adding the words "apply it" (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer, e.g., a limitation indicating that a particular function such as creating and maintaining electronic records is performed by a computer, as discussed in Alice Corp., 573 U.S. at 225-26, 110 USPQ2d at 1984” See MPEP § 2106.05(I)(A). The addition of the limitations “performed by one of the plurality of terminals or a server” is merely an additional element that amounts to adding the words “apply it” (or an equivalent) with the judicial exception, or merely uses a computer in its ordinary capacity as a tool to perform an existing process. See MPEP § 2106.05(f)(2). The recitation of the types of resources that may be monitored generally links the use of the judicial exception to a particular technological environment or field of use. See MPEP § 2106.05(h). Further information regarding how the models are processed, such as those found on page 5 of the Specification, would be necessary to not amount to more adding the words “apply it” (or an equivalent) with the judicial exception. Therefore, the claim 1 is rejected under 35 U.S.C. § 101.
Applicant next argues “Applicants believe that such a method could not be achieved by humans as a purely mental process, and accordingly, Applicants respectfully request the 101 rejections against claim 1 be withdrawn.” The MPEP states “If the identified limitation(s) falls within at least one of the groupings of abstract ideas, it is reasonable to conclude that the claim recites an abstract idea in Step 2A Prong One. The claim then requires further analysis in Step 2A Prong Two, to determine whether any additional elements in the claim integrate the abstract idea into a practical application” See MPEP § 2106.04(a). While the method is not a purely mental process, those elements that are considered a mental process were identified in Step 2A Prong One, and additional elements were found to either amount to adding the words “apply it”, generally link the use of the judicial exception to a particular technological environment or adding insignificant extra-solution activity to the judicial exception in Step 2A Prong Two and Step 2B. Therefore, the claims are directed to the abstract idea of a mental process, and rejected under 35 U.S.C. § 101.
Applicant next argues “the classes assigned to its sub-model are different sub-classes (1 ~ 5) of the main model's classes (A ~ D)”. However, the broadest reasonable interpretation of a subset of the main classes includes the subclasses of the main class as shown in Figure 6 of Ono.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JACOB Z SUSSMAN MOSS whose telephone number is (571) 272-1579. The examiner can normally be reached Monday - Friday, 9 a.m. - 5 p.m. ET.
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, Kakali Chaki can be reached at (571) 272-3719. 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.
/J.S.M./Examiner, Art Unit 2122
/KAKALI CHAKI/Supervisory Patent Examiner, Art Unit 2122