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 .
Status of Claims
This action is in response to the amendments filed 29 April 2025.
Claims 1-20 are pending.
Of the pending claims, 1, 8, and 15 are amended.
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 1-20 are rejected under 35 U.S.C. 103.
Claims 1, 2, 8, 14, 15, and 20
Claims 1-2, 8, 14-15, and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Hsiao et al. ("Applying a Hybrid Approach Based On Fuzzy Neural Network and Genetic Algorithm to Product Form Design", hereinafter "Hsiao") in view of Aggarwal et al. (U.S. 20230012650 hereinafter Aggarwal) in further view of Gandhi et al. ("Spatio-temporal Multi-graph Networks for Demand Forecasting in Online Marketplaces", hereinafter "Gandhi").
Claim 1
Regarding claim 1, Hsiao teaches:
a computer-implemented method of creating a prototype (Abstract, “The proposed method provides an automatic design system, which gives designers the ability to rapidly obtain a product form and its corresponding image”), the method comprising:
retrieving product image data corresponding to the one or more input design parameters from a data source (Hsiao (paragraph above figure 9), “the customized interface used for constructing a 3-D model of an electronic door lock is shown in Fig. 9. In the parameter-setting window, the designer is invited to specify the dimensions of the required form by regulating the scroll bar or by selecting the appropriate options for each form parameter. When all of the parameters have been set, the designer presses the ‘‘Show 3-D models’’ button to dynamically browse the exported VRMLfile”), wherein the data source comprises at least one of an internet source or a user provided source (Hsiao (paragraph above figure 9), “the customized interface used for constructing a 3-D model of an electronic door lock is shown in Fig. 9. In the parameter-setting window, the designer is invited to specify the dimensions of the required form by regulating the scroll bar or by selecting the appropriate options for each form parameter), and wherein the retrieving is based on a result of a data search (Hsiao (paragraph above figure 9, fig. 10, product form search, fig. 11, VRML file represented on the Cosmo Player), “when all of the parameters have been set, the designer presses the ‘‘Show 3-D models’’ button to dynamically browse the exported VRMLfile under I-Deas batch-mode processing (result of a data search)”)
receiving, by a first neural network, one or more input design parameters (Hsiao, pg. 415, section 2.2, third paragraph, “Prior to using the BPN [backpropagating neural network] in practice, it must first undergo a training process, in which it is assumed that the product form is generated by an input parameter set: v1, v2, …, vn (i.e. the input-node code) and a corresponding fuzzy number, ~tout, for the related image sensation”; fig. 1 shows a concept which uses input form variables, an input goal image, and experimental image data that the BPN receives, which would be equivalent to design parameters);
generating, using the first neural network comprising a discriminator, a plurality of prototypes based on the one or more input design parameters (Hsiao, Abstract, “a feature-based hierarchical computer-aided design (CAD) model is constructed, in which the related form parameters are thoroughly defined in applicable domains to facilitate the automatic generation of new product forms. A fuzzy neural network algorithm is then applied to establish the relationships between the input form parameters and a series of adjectival image words”; pg. 415, section 2.2, first paragraph, “In the product image prediction mechanism, a fuzzy BPN is employed to model the non-linear relationships between the given form parameters (i.e. the network input) and the product image evaluation (i.e. the network output)”; pg. 415, section 2.2, third paragraph, “During the training process, a specific number of verified training data sets are input to the BPN. The network is supervised such that if it outputs an incorrect answer, a correction mechanism is activated which back-propagates through the network, adjusting the network’s weights and biases iteratively so as to progressively reduce the degree of error”; Examiner interprets a “discriminator” to be any type of classifier that can predict labels for input data, and the reference’s BPN classifies input data and produces a predictive output); and
wherein the first neural network is trained on at least one of the product image data or historical user preference data (pg. 415, section 2.2, second paragraph, “The input nodes of the BPN are used to describe the defined parameters for the product form, where each form parameter is constrained within an applicable range. The 21 output nodes are used to describe the corresponding fuzzy membership function… This function can then be related to a specified linguistic variable to indicate the degree of intensity of the product image sensation”; such data that describes human sensations as it relates to images that the BPN is trained on can be considered historical user preference data).
While Hsiao teaches generating prototypes with neural networks, Hsiao does not explicitly teach generating decoy prototypes. However, Aggarwal teaches
randomly initializing a decoy prototype by a second neural network trained on market place data (Aggarwal (¶0054, fig. 6 item 606), “calculating, by the machine learning framework, an environmental impact score and a demand impact score associated with each of the created refurbished designs (decoy prototype), wherein the demand impact score is based at least in part on the location-specific demand data”) based on at least one prototype of the plurality of prototype (Aggarwal (¶0054, fig. 6 item 606), “includes creating, using the machine learning framework, one or more refurbished designs of each given one of the plurality of products based on the initial image of the given product and one or more design constraints”).
presenting, by an electronic display, a report comprising at least a portion of the plurality of prototypes and at least one of the one or more decoy prototypes (Aggarwal (¶0030, fig. 3), “this figure shows a counterfactual formulation process in accordance with exemplary embodiments. In the FIG. 3 example, the counterfactual formulation process 302 is performed with respect to an image of a first product 300 and an image of a refurbished product 304. As can be seen from FIG. 3, the refurbished product 304 includes some modifications 306 relative to product 300”)
Hsiao and Aggarwal are considered analogous to the claimed invention as they are in the field of prototype design and development. Aggarwal teaches the development of decoy prototypes with particular design parameters, while Hsiao teaches the development of prototypes with particular design parameters through use of a neural network. It would have been obvious to a person having ordinary skill in the art (hereinafter “PHOSITA”), before the effective filing date of the invention, to incorporate the teachings of Aggarwal into the design of Hsiao. The benefit of making this change is to “an exemplary embodiment includes an artificial intelligence-based system that is configured to discover products and refurbishing schemes for products in unsold inventory” (Aggarwal (¶0016 line 1-4)).
The combination of Hsiao and Herweg does not teach the limitation backpropagating through the second neural network to modify the decoy prototype to derive one or more decoy prototypes. However, Gandhi teaches:
backpropagating through the second neural network to modify the decoy prototype to derive one or more decoy prototypes (Gandhi, pg. 12, section 4.1, first paragraph, “In each minibatch, we […] feed [embeddings] into a sequential network for generating predictions, and finally backpropagate the loss to train the network”; graph neural network feeds into sequential network, and backpropagation takes place within both at once during training).
Gandhi is considered analogous to the claimed invention since they are in the same field of endeavor of utilizing machine learning techniques to achieve a stated goal. It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Gandhi into that of Hsiao and Herweg; it is inherent that a neural network will output data related to that which it is trained on, so using a particular neural network such as a graph neural network or using marketplace data as training data is considered a simple substitution leading to a predictable result.
Claim 2
Regarding claim 2, the rejection of claim 1 is incorporated. Further, the combination of Hsiao, Herweg, Gandhi, and Teppan teaches the computer-implemented method of claim 1, further comprising fabricating at least one select prototype of the plurality of prototypes and the one or more decoy prototypes (Hsiao, pg. 426, left col., first paragraph, “The CAD models… can then be input to the RP (Rapid Prototyping) machine to produce physical prototypes”; Hsiao explains, in the Abstract, “a feature-based hierarchical computer-aided design (CAD) model is constructed, in which the related form parameters are thoroughly defined in applicable domains to facilitate the automatic generation of new product forms”).
Claim 5
Regarding claim 5, the rejection of claim 1 is incorporated. Further, the combination of Hsiao, Herweg, Gandhi, and Teppan teaches the computer-implemented method of claim 1 further comprising receiving, by the second neural network, competitive data relating to a product corresponding to the one or more input design parameters (Gandhi, pg. 11, “Experimental Results” paragraph, “This section outlines our experiments using different GNN architectures for demand forecasting in online marketplaces… We use a random sample of the demand data (weekly sales of products from sellers) on Amazon from January 2016 to July 2018 for evaluating our models”; Amazon marketplace data is considered competitive data). It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Gandhi into that of Hsiao in view of Aggarwal; it is inherent that a neural network will output data related to that which it is trained on, so using historical product data as training data is considered a simple substitution leading to a predictable result.
Claim 8
Regarding claim 8, Hsiao teaches:
a system for creating a prototype (Hsiao, Abstract, “The proposed method provides an automatic design system, which gives designers the ability to rapidly obtain a product form and its corresponding image”), the system comprising:
one or more processors (Hsiao, pgs. 423-4, figs. 9-11 show screenshots from a computer display, indicating the display is part of a system containing a processor);
an electronic display (Hsiao, pgs. 423-4, figs. 9-11 show screenshots from a computer display, indicating the presence of an electronic display);
one or more non-transitory memory modules storing computer-readable instructions (Hsiao, pgs. 423-4, figs. 9-11 show screenshots from a computer display, indicating the display is part of a system containing a processor and inherent non-transitory memory units containing computer instructions for running the process shown in the screenshots) that, when executed, cause the one or more processors to
The rest of the limitation(s) are rejected for the same reasons as Claim 1.
Claim 11
Regarding claim 11, the rejection of claim 8 is incorporated. Further, the combination of Hsiao, Herweg, Gandhi, and Teppan teaches the system of claim 8, wherein the computer-readable instructions further cause the one or more processors to receive, by the second neural network, competitive data relating to a product corresponding to the one or more input design parameters (Gandhi, pg. 11, “Experimental Results” paragraph, “This section outlines our experiments using different GNN architectures for demand forecasting in online marketplaces… We use a random sample of the demand data (weekly sales of products from sellers) on Amazon from January 2016 to July 2018 for evaluating our models”; Amazon marketplace data is considered competitive data, and use of the neural network is done via computer instructions).
It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Gandhi into that of Hsiao in view of Aggarwal; it is inherent that a neural network will output data related to that which it is trained on, so using historical product data as training data is considered a simple substitution leading to a predictable result.
Claim 14
Regarding claim 14, the rejection of claim 8 is incorporated. Further, the combination of Hsiao, in view of Aggarwal in further view of Gandhi teaches the system of claim 8, further comprising a three-dimensional printer, wherein the three-dimensional printer is configured to fabricate at least a portion of at least one prototype of the plurality of prototypes (Hsiao, pg. 426, left col., first paragraph, “The CAD models… can then be input to the RP (Rapid Prototyping) machine to produce physical prototypes”; Hsiao explains, in the Abstract, “a feature-based hierarchical computer-aided design (CAD) model is constructed, in which the related form parameters are thoroughly defined in applicable domains to facilitate the automatic generation of new product forms”).
Claim 15
Regarding claim 15, Hsiao teaches:
a computer-implemented method of fabricating a prototype (Abstract, “The proposed method provides an automatic design system, which gives designers the ability to rapidly obtain a product form and its corresponding image”), the method comprising:
The rest of the limitation(s) are rejected for the same reasons as Claim 1.
Claim 18
Regarding claim 18, the rejection of claim 15 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the computer-implemented method of claim 15, but does not teach the limitation further comprising receiving, by the second neural network, competitive data relating to a product corresponding to the one or more input design parameters. However, Gandhi teaches this limitation (Gandhi, pg. 11, “Experimental Results” paragraph, “This section outlines our experiments using different GNN architectures for demand forecasting in online marketplaces… We use a random sample of the demand data (weekly sales of products from sellers) on Amazon from January 2016 to July 2018 for evaluating our models”; Amazon marketplace data is considered competitive data).
It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Gandhi into that of Hsiao in view of Aggarwal; it is inherent that a neural network will output data related to that which it is trained on, so using historical product data as training data is considered a simple substitution leading to a predictable result.
Claim 20
Regarding claim 20, the rejection of claim 15 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches wherein receiving the plurality of prototypes comprises receiving a report comprising ranked prototypes. However, Teppan teaches this limitation (Aggarwal (¶0029 last 4 lines), “The counterfactual optimization module 128 may generate a ranked list of refurbished designs using population-based search algorithms (such as NSGA-II, for example).”).
Claims 3, 9, and 16
Claims 3, 9, and 16 are rejected under 35 U.S.C. 103 as being unpatentable over in view of Aggarwal in further view of Gandhi as applied to claims 1, 8, and 15 above, and further in view of Krahe et al. ("Deep Learning for Automated Product Design", hereinafter "Krahe").
Claim 3
Regarding claim 3, the rejection of claim 1 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the computer-implemented method of claim 1, but does not teach the limitation wherein the first neural network comprises a generative adversarial network (GAN). However, Krahe teaches this limitation (Krahe, pg. 4, right col., last paragraph, “The approach developed is based on the generative model presented in [4] consisting of an AE [autoencoder] and a GAN. The AE converts the 3D objects into a latent vector representation. Since training and use of the GAN is fully based on these latent vectors, the GAN is called l-GAN (latent GAN) and is independent of the input data format”; the approach is specifically in relation to product design).
Krahe is considered analogous to the claimed invention since they are in the field of product design via machine learning. It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Krahe into the teachings of Hsiao in view of Aggarwal in further view of Gandhi. The benefit of making this change is that “more recent approaches, as presented in [4,20–26], focus on generative and discriminative representations for geometry… They make use of a combination of AEs and generative models, such as GANs [3], to create new 3D objects” (Krahe, pg. 4, right col., second paragraph).
Claim 9
Regarding claim 9, the rejection of claim 8 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the system of claim 8, but does not teach the limitation wherein the first neural network comprises a generative adversarial network (GAN). However, Krahe teaches this limitation (Krahe, pg. 4, right col., last paragraph, “The approach developed is based on the generative model presented in [4] consisting of an AE [autoencoder] and a GAN. The AE converts the 3D objects into a latent vector representation. Since training and use of the GAN is fully based on these latent vectors, the GAN is called l-GAN (latent GAN) and is independent of the input data format”; the approach is specifically in relation to product design, and use of the neural network is done via computer instructions).
It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Krahe into the teachings of Hsiao in view of Aggarwal in further view of Gandhi. The benefit of making this change is that “more recent approaches, as presented in [4,20–26], focus on generative and discriminative representations for geometry… They make use of a combination of AEs and generative models, such as GANs [3], to create new 3D objects” (Krahe, pg. 4, right col., second paragraph).
Claim 16
Regarding claim 16, the rejection of claim 15 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the computer-implemented method of claim 15, but does not teach the limitation wherein the first neural network comprises a generative adversarial network (GAN). However, Krahe teaches this limitation (Krahe, pg. 4, right col., last paragraph, “The approach developed is based on the generative model presented in [4] consisting of an AE [autoencoder] and a GAN. The AE converts the 3D objects into a latent vector representation. Since training and use of the GAN is fully based on these latent vectors, the GAN is called l-GAN (latent GAN) and is independent of the input data format”; the approach is specifically in relation to product design).
Krahe is considered analogous to the claimed invention since they are in the field of product design via machine learning. It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Krahe into the teachings of Hsiao in view of Aggarwal in further view of Gandhi. The benefit of making this change is that “more recent approaches, as presented in [4,20–26], focus on generative and discriminative representations for geometry… They make use of a combination of AEs and generative models, such as GANs [3], to create new 3D objects” (Krahe, pg. 4, right col., second paragraph).
Claims 4, 10, and 17
Claims 4, 10, and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Hsiao in view of Aggarwal in further view of Gandhi as applied to claims 1, 8, and 15 above, and further in view of Edwards et al. (U.S. Patent No. 10,650,358; hereinafter "Edwards").
Claim 4
Regarding claim 4, the rejection of claim 1 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the computer-implemented method of claim 1, but does not teach the limitation further comprising receiving, by the first neural network, historical product data for training the first neural network. However, Edwards teaches this limitation (Edwards, paragraph 36, “assume the document management platform trained or received a data model that had been trained on historical warranty data. The historical warranty data may include… historical product data for the product to which the warranty applies, and/or the like. The data model may use one or more machine learning techniques… The one or more machine learning techniques may include… a technique using a neural network, and/or the like”; paragraph 37, “the document management platform may provide the transaction information… as input to the data model”; Edwards discloses a neural network model trained on historical data such as warranty data or transaction data, which can be considered product data).
Edwards is considered analogous to the claimed invention since they are in the same field of endeavor of utilizing machine learning techniques to achieve a stated goal. It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Edwards into that of Hsiao in view of Aggarwal in further view of Gandhi; it is inherent that a neural network will output data related to that which it is trained on, so using historical product data as training data is considered a simple substitution leading to a predictable result.
Claim 10
Regarding claim 10, the rejection of claim 8 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the system of claim 8, but does not teach the limitation wherein the computer-readable instructions further cause the one or more processors to receive, by the first neural network, historical product data for training the first neural network. However, Edwards teaches this limitation (Edwards, paragraph 36, “assume the document management platform trained or received a data model that had been trained on historical warranty data. The historical warranty data may include… historical product data for the product to which the warranty applies, and/or the like. The data model may use one or more machine learning techniques… The one or more machine learning techniques may include… a technique using a neural network, and/or the like”; paragraph 37, “the document management platform may provide the transaction information… as input to the data model”; Edwards discloses a neural network model trained on historical data such as warranty data or transaction data, which can be considered product data, and use of the neural network is done via computer instructions).
Edwards is considered analogous to the claimed invention since they are in the same field of endeavor of utilizing machine learning techniques to achieve a stated goal. It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Edwards into that of Hsiao in view of Aggarwal in further view of Gandhi; it is inherent that a neural network will output data related to that which it is trained on, so using historical product data as training data is considered a simple substitution leading to a predictable result.
Claim 17
Regarding claim 17, the rejection of claim 15 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the computer-implemented method of claim 15, but does not teach the limitation further comprising receiving, by the first neural network, historical product data for training the first neural network. However, Edwards teaches this limitation (Edwards, paragraph 36, “assume the document management platform trained or received a data model that had been trained on historical warranty data. The historical warranty data may include… historical product data for the product to which the warranty applies, and/or the like. The data model may use one or more machine learning techniques… The one or more machine learning techniques may include… a technique using a neural network, and/or the like”; paragraph 37, “the document management platform may provide the transaction information… as input to the data model”; Edwards discloses a neural network model trained on historical data such as warranty data or transaction data, which can be considered product data).
Edwards is considered analogous to the claimed invention since they are in the same field of endeavor of utilizing machine learning techniques to achieve a stated goal. It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Edwards into that of Hsiao in view of Aggarwal in further view of Gandhi; it is inherent that a neural network will output data related to that which it is trained on, so using historical product data as training data is considered a simple substitution leading to a predictable result.
Claims 6, 7, 12, 13, and 19
Claims 6-7, 12-13, and 19 are rejected under 35 U.S.C. 103 as being unpatentable over Hsiao in view of Aggarwal in further view of Gandhi as applied to claims 1, 8, and 15 above, and further in view of Felfernig et al. ("A Dominance Model for the Calculation of Decoy Products in Recommendation Environments", hereinafter "Felfernig").
Claim 6
Regarding claim 6, the rejection of claim 1 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the computer-implemented method of claim 1, but does not teach the limitation further comprising filtering the plurality of prototypes to eliminate one or more prototypes of the plurality of prototypes. However, Felfernig teaches this limitation (Felfernig, pg. 46, right col., first paragraph, “It is important to note that the application of the SDM [simple dominance model] happens after the recommender has filtered out completely unsuitable items (e.g.: items exceeding a minimum/maximum threshold on a certain attribute)”; Felfernig explains in the previous paragraph that these items are grouped with decoy items that increase their dominance value, and items not within a certain attribute or quality range are discarded).
Felfernig is considered analogous to the claimed invention since they are in the same field of product and decoy design. It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Felfernig into the teachings of Hsiao in view of Aggarwal in further view of Gandhi. As the Hsiao combination does not include any form of filtering prototypes, Felfernig’s teachings can be added as an improvement. This would be applying a known technique to a known device (method or product) ready for improvement to yield predictable results.
Claim 7
Regarding claim 7, the rejection of claim 1 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the computer-implemented method of claim 1, but does not teach the limitation further comprising ranking the plurality of prototypes. However, Felfernig teaches this limitation (Felfernig, pg. 43, right col., list item 3, “After the utility calculation of the products the system typically presents the top ranked products to the user”; the products are a combination of items and decoy items as explained on page 46).
It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Felfernig into the teachings of Hsiao in view of Aggarwal in further view of Gandhi. As the Hsiao combination does not include any form of ranking prototypes, Felfernig’s teachings can be added as an improvement. This would be applying a known technique to a known device (method or product) ready for improvement to yield predictable results.
Claim 12
Regarding claim 12, the rejection of claim 8 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the system of claim 8, but does not teach the limitation wherein the computer-readable instructions further cause the one or more processors to filter the plurality of prototypes to eliminate one or more prototypes of the plurality of prototypes. However, Felfernig teaches this limitation (Felfernig, pg. 46, right col., first paragraph, “It is important to note that the application of the SDM [simple dominance model] happens after the recommender has filtered out completely unsuitable items”; Felfernig explains in the previous paragraph that these items are grouped with decoy items that increase their dominance value, and items not within a certain attribute or quality range are discarded).
It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Felfernig into that of Hsiao in view of Aggarwal in further view of Gandhi. As the Hsiao combination does not include any form of filtering prototypes, Felfernig’s teachings can be added as an improvement. This would be applying a known technique to a known device (method or product) ready for improvement to yield predictable results. Furthermore, writing instructions into non-transitory memory to perform a filtering task is an obvious solution for a PHOSITA, which would be considered a combination of prior art elements.
Claim 13
Regarding claim 13, the rejection of claim 8 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the system of claim 8, but does not teach the limitation wherein the computer-readable instructions further cause the one or more processors to rank the plurality of prototypes. However, Felfernig teaches this limitation (Felfernig, pg. 43, right col., list item 3, “After the utility calculation of the products the system typically presents the top ranked products to the user”; the products are a combination of items and decoy items as explained on page 46).
It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Felfernig into that of Hsiao in view of Aggarwal in further view of Gandhi. As the Hsiao combination does not include any form of ranking prototypes, Felfernig’s teachings can be added as an improvement. This would be applying a known technique to a known device (method or product) ready for improvement to yield predictable results. Furthermore, writing instructions into non-transitory memory to perform a ranking task is an obvious solution for a PHOSITA, which would be considered a combination of prior art elements.
Claim 19
Regarding claim 19, the rejection of claim 15 is incorporated. Further, the combination of Hsiao in view of Aggarwal in further view of Gandhi teaches the computer-implemented method of claim 15, but does not teach the limitation further comprising filtering the plurality of prototypes to eliminate one or more prototypes of the plurality of prototypes. However, Felfernig teaches this limitation (Felfernig, pg. 46, right col., first paragraph, “It is important to note that the application of the SDM [simple dominance model] happens after the recommender has filtered out completely unsuitable items”; Felfernig explains in the previous paragraph that these items are grouped with decoy items that increase their dominance value, and items not within a certain attribute or quality range are discarded).
Felfernig is considered analogous to the claimed invention since they are in the same field of product and decoy design. It would have been obvious to a PHOSITA, before the effective filing date of the claimed invention, to incorporate the teachings of Felfernig into the teachings of Hsiao in view of Aggarwal in further view of Gandhi. As the Hsiao combination does not include any form of filtering prototypes, Felfernig’s teachings can be added as an improvement. This would be applying a known technique to a known device (method or product) ready for improvement to yield predictable results.
Response to Arguments
The following is responsive to the remarks filed 2 September 2025.
35 U.S.C. 101
Applicant’s arguments regarding the rejection of claim 1 under 35 U.S.C. 101, starting on page 2 of the remarks, have been fully considered. Applicant argues, on page 5 of the remarks, that “backpropagting through the second neural network to modify the one or more decoy prototypes” cannot be performed in human mind.
PNG
media_image1.png
229
703
media_image1.png
Greyscale
Applicant’s arguments are persuasive. Therefore, 35 U.S.C. §101 rejections are respectfully withdrawn.
35 U.S.C. 103
Applicants argue that cited references does not disclose amended Claim 1 (last paragraph of page 11 in the remarks).
PNG
media_image2.png
146
672
media_image2.png
Greyscale
Applicants’ arguments are moot because new reference in combination with Hsiao and Gandhi teaches the current claim amendment(s).
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to TEVIN T NIANG-TROST whose telephone number is (571)272-2424. The examiner can normally be reached Mon-Fri, 7:30 a.m. - 5:00 p.m. CDT.
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, Viker Lamardo can be reached on (571) 270-5871. 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.
/NHAT HUY T NGUYEN/Primary Examiner, Art Unit 2147