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 communication is a Final Office Action in response to Applicant’s amendment for application number 18/812,236 received on 05/25/2026.
In accordance with Applicant’s amendment, claims 8/9 have been canceled. Claims 1-7, 10-20 are amended, currently pending and have been examined.
Priority
Applicants claim for the benefit of a prior-filed application under 35 U.S.C. 119 and/or 35 U.S.C. 120 is acknowledged.
Response to Amendment
The amendment filed on 05/25/2026 has been entered.
Applicant’s amendment necessitated the new ground(s) of rejection set forth in this Office Action.
Upon review of amendment, the §112 (b) rejection previously applied to claim 10 is withdrawn.
Upon review of amendment, the §102 rejections previously applied are withdrawn. The claims previously rejected under §102 remain rejected by new grounds of rejections under §103.
Response to Arguments
Response to §101 arguments – Applicant’s arguments with respect to the §101 rejections previously applied to the claims have been considered and are not persuasive.
Applicant argues (Remarks at pg. 9): “The Examiner asserts that the claims fall under "certain methods of organizing human activity" and "mental processes." Applicant respectfully disagrees. The claims are not directed to any mere economic principle or practice, commercial or legal interaction, or management of personal behaviors or relationships or interactions between people, as described in MPEP 2106.04(a)(2). Further, the amended independent claims, as a whole, cannot be practically performed by the human mind. Specifically, the Examiner's step-by-step characterization of the claim limitations does not account for the amended claim language requiring: (1) generating a graph from product sales information with nodes representing products and edges between nodes representing products having been sold in a single order; and (2) training a graph neural network (GNN) using the generated graph. These are computationally intensive processes that have no mental analog. The human mind cannot generate a product co-purchase graph with complex node-edge relationships from large-scale product sales data, nor can it train a GNN by iteratively updating weights through message passing between nodes. Unlike simple observation or evaluation, GNN training involves mathematical optimization across potentially millions of parameters, which is fundamentally incapable of being performed mentally or with pen and paper.”. In response, Examiner respectfully disagrees and notes that the claims, as currently recited, fall under the “Certain Methods of Organizing Human Activity” and “Mental Processes” abstract idea groupings. For example, in order to fulfill the limitations “present the at least one product bundles to a merchant; receiving an indication of one or more of the product bundles selected to offer for sale; posting the selected one or more product bundles.”, at least one product bundle has to be posted to an online sales channel, which constitutes a commercial interaction or sales activity through an online sales channel between the entity posting the product bundle, and potential customers using the sales channel. Therefore, these limitations fall under the “Certain Methods of Organizing Human Activity”, directed to commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations. Moreover, the claims also fall under the “Mental Processes” abstract idea grouping because they recite steps that can be performed in the human mind via observation, evaluation, judgement, and/or opinion, or with the help of pen and paper. For example, one of ordinary skill in the art would be able to generating a graph from the product sales information with nodes of the graph representing products and edges of the graph between nodes representing products represented by the nodes having been sold in a single order with the help of pen and paper, or automatically identifying at least one anchor product based on one or more of the product sales information and product inventory data via observation, or identify at least one product recommended for sale with the respective anchor product in a respective product bundle via evaluation, judgement, or opinion. Therefore, the claims recite steps that fall under the “Mental Processes” abstract idea grouping for actions that could be accomplished mentally, such as by human observation, evaluation, judgement, opinion, or with the help of pen and paper. Therefore, the claims recite abstract ideas, which requires the analysis to continue under Step 2A, Prong 2 to determine if the additional elements integrate the abstract idea into a practical application, and Step 2B to determine if the additional elements add significantly more to the abstract idea. Examiner further reminds Applicant that reciting “computationally intensive processes” is not a consideration under Step 2A, Prong 1. Furthermore, as disclosed in MPEP 2106.05(f), "claiming the improved speed or efficiency inherent with applying the abstract idea on a computer" does not integrate a judicial exception into a practical application or provide an inventive concept.
Applicant argues (Remarks at pgs. 9-10): “It is respectfully noted that claims reciting limitations directed to training and applying a machine learning model are not directed to an abstract idea, for similar reasons detailed in Example 39 of the Subject Matter Eligibility Examples highlighted in the MPEP 2106.04. Like Example 39, which involved training a neural network on a specific data set to produce a trained model capable of classifying data inputs, the present claims recite training a specific type of neural network (a GNN) on a specific type of data structure (a product co-purchase graph) to produce a trained recommendation model capable of identifying product bundles. The claims do not merely invoke a neural network at a high level of abstraction; they specify how the graph is constructed (nodes representing products, edges representing co-purchase relationships within single orders) and how the GNN is trained on that particular graph structure. This level of specificity places the claims squarely within the category of patent-eligible subject matter identified in Example 39.”. In response, Examiner respectfully disagrees and notes that, contrary to the claim in Example 39, which does not recite an abstract idea, Applicant’s claims do recite abstract ideas as noted above, as well as in the updated §101 rejections below. Examiner further reminds Applicant that the mere recitation of machine learning, or any other additional element does not void abstract steps recited in the claim limitations. Furthermore, the recitation of additional elements in combination with abstract steps results in the analysis of the additional elements under Step 2A, Prong 2, and Step 2B of the eligibility inquiry. See updated §101 rejections below for further details.
Applicant argues (Remarks at pgs. 10-11): “The independent claims provide an improvement in the generation of recommended product bundles through a specific technical architecture. A graph neural network (GNN) is trained using a graph generated from product sales information with nodes representing products and edges representing co-purchase relationships from single orders to identify products recommended for sale with an automatically identified anchor product in a respective product bundle. As noted in the specification (see para. [0036]), "Using a graph neural network yields significant improvements in accuracy, precision@K and recall@K. Essentially, the products recommended by the GNN for bundling are more likely to be bought together than products recommended for bundling by other methods, such as matrix factorization." Critically, this improvement is rooted in the claim elements themselves, namely the specific combination of graph construction from co-purchase data, GNN training on that graph, and application of the trained GNN to automatically identified anchor products is the mechanism that produces the accuracy improvement. This is not a case where the specification describes improvements that are absent from the claim language; rather, the claims expressly recite the technical steps that yield the disclosed improvement. Therefore, the claims are not mere instructions to implement an abstract idea on a generic computer and do not merely "apply it," as alleged by the Examiner. Instead, the claimed method improves computer functionality for product bundle recommendation systems. As noted in the memo by Charles Kim dated December 5, 2025 (Advance notice of change to the MPEP in light of Ex Parte Desjardins), "When evaluating a claim as a whole, examiners should not dismiss additional elements as mere 'generic computer components' without considering whether such elements confer a technological improvement to a technical problem, especially as to improvements to computer components or the computer system. See, e.g., Ex Parte Desjardins, Appeal No. 2024-000567 (PTAB September 26, 2025) (Appeals Review Panel Decision)." Here, the GNN and the product graph are not generic computer components - they are specific technical elements that, in combination, confer a demonstrated technological improvement to the technical problem of accurate product bundle recommendation.”. In response, Examiner respectfully disagrees and notes that the present claims do not provide an analogous improvement to the machine learning model (e.g. recommendation model). Examiner respectfully asserts that the claims are unlike the Desjardins decision because the claims are directed to an abstract idea versus being directed to an improvement to computer functionality. The present claims do not provide an analogous technical solution to that of Desjardins because the claims do not “address challenges in continual learning and model efficiency by reducing storage requirements and preserving task performance across sequential training”. The machine learning model of the present claims is merely a tool to perform the abstract process. An improvement to the presented generation of recommended product bundles using a recommendation model with a GNN architecture would be an improvement to the abstract limitations for consideration under Step 2A, Prong 1 and not to recommendation systems technology. MPEP 2106.05(a): “It is important to note, the judicial exception alone cannot provide the improvement. The improvement can be provided by one or more additional elements...” Additionally, as discussed in 2106.05(a)(II) improvements to technology or technical fields, “an improvement in the abstract idea itself … is not an improvement in technology”.
Applicant argues (Remarks at pgs. 12-13): “Furthermore, even assuming arguendo that the claims are directed to an abstract idea (which Applicant does not concede), the claims satisfy Step 2B because they recite elements that, individually and as an ordered combination, amount to significantly more than any alleged abstract idea. The Examiner characterized the additional elements as well-understood, routine, and conventional, citing Symantec, TLI Communications, OIP Technologies, and buySAFE for the proposition that receiving and transmitting data over a network is routine. However, these cases are not appropriate. The claimed receiving step is not generic data transmission over a network but rather is the receipt of structured product sales information specifically comprising order information indicating products sold together and order costs. This is a particular type of data input tied to and required for the specific graph construction and GNN training process claimed, not mere generic data gathering.
Moreover, the ordered combination of claim elements adds significantly more than any individual element. The specific sequence of (1) receiving product sales information with co-purchase and cost data, (2) generating a product graph with nodes and edges reflecting co-purchase relationships from single orders, (3) training a GNN using the generated graph, (4) automatically identifying anchor products based on sales and inventory data, (5) applying the anchor products to the trained GNN to generate product bundles, (6) presenting bundles to a merchant via a user interface, and (7) posting merchant-selected bundles to an online sales channel is unconventional and not routine in the product recommendation art. This ordered combination provides an inventive concept because it integrates graph-based machine learning with automated anchor product identification and merchant-curated bundle posting in a manner not previously known. The Examiner's assertion that the "ordered combination adds nothing that is not already present as when the elements are taken individually" does not account for the synergistic interaction between the graph construction, GNN training, automatic anchor identification, and merchant-facing presentation steps, each of which feeds into and depends upon the preceding steps to achieve the improved product bundle recommendations.”. In response, Examiner respectfully disagrees and notes that Applicant’s claims are nothing significantly more because they rely on additional elements that amount to using generic computing elements (computer hardware) or instructions/software (engine) to perform the abstract idea, similar to adding the words “apply it” (or an equivalent), which merely serves to link the use of the judicial exception to a particular technological environment (network computing environment, the internet, online) and do not amount to significantly more than the abstract idea itself, and to insignificant extra-solution activity (e.g., mere data gathering), which does not add significantly more. Additionally, the mere data gathering extra-solution activity have been recognized as well-understood, routine, and conventional, and thus insufficient to add significantly more to the abstract idea. Therefore, the additional elements, when taken individually or in combination, fail to integrate the abstract idea into a practical application, fail to add significantly more, and fail to represent an improvement to technology. See updated §101 section below for details.
Response to §102 arguments – Applicant’s arguments with respect to the §102 rejections previously applied to the claims are moot based on the new grounds of rejections set forth in the instant office action, as necessitated by the amendment. Nevertheless, Examiner notes the following:
Regarding Applicant’s argument (Remarks at pg. 14) “In contrast, Das Gupta
, Examiner respectfully disagrees and notes that the action to identify the seed product disclosed in Das Gupta is based on the customer add to cart data. See [Fig. 2] Add To Cart Data 210, and [0003] The memory stores instructions executable by the processor to: receive, from a customer device, a selection of a first product offered for sale via the retailer website; access the data store to retrieve customer add to cart data, product price data, and degree of diversity data, wherein: the customer add to cart data is based on historical customer shopping activity of a plurality of customers and includes data describing online shopping sessions in which products are selected for addition to historical shopping carts to which the first product was previously added, and the customer add to cart data further includes data describing an order in which the products are added to the historical shopping carts to which the first product was previously added; based on the selection of the first product and the customer add to cart data, generate a list of bundled products having a complementary relationship with the first product, wherein, in the historical customer shopping activity, the bundled products have been selected after the selection of the first product more than a threshold number of times; select at least one bundled product from the list of bundled products; and present the at least one bundled product on a user interface on the customer device as a recommendation for purchase with the first product. One of ordinary skill in the art would reasonably consider Add To Cart Data 210 to be equivalent to product sales information. See also pars. [0005] and [0022].
Response to §103 arguments – Applicant’s arguments with respect to the §103 rejections previously applied to the claims have been considered and are not persuasive.
Regarding Applicant’s argument (Remarks at pgs. 16-17) “Das Gupta requires a user to select a seed product, and then presents recommended products based on the selected seed product. There is no disclosure or suggestion of any automatic identification of an anchor product based on at least one of product sales information and inventory data, nor of any posting of recommended bundles on an online sales channel without requiring a user to select a product.”. In response, Examiner notes that as discussed above, one of ordinary skill in the art would reasonably consider the Add To Cart Data 210 from Das Gupta (Fig. 2) as equivalent to product sales information.
Regarding Applicant’s argument (Remarks at pgs. 16-17) “Moreover, the Examiner's combination of up to four references for certain dependent claims (Das Gupta + Ding + Tong + Liu for claim 11) is indicative of impermissible hindsight reconstruction. See In re McLaughlin, 443 F.2d 1392, 1395 (CCPA 1971) ("Any judgment on obviousness is in a sense necessarily a reconstruction based upon hindsight reasoning, but so long as it takes into account only knowledge which was within the level of ordinary skill at the time the claimed invention was made and does not include knowledge gleaned only from applicant's disclosure, such a reconstruction is proper."). Here, the Examiner has not articulated a coherent rationale explaining why a person of ordinary skill would have combined four disparate references including a customer-facing product bundling system (Das Gupta), a product classification GNN (Ding), a domain-transfer graph updating system (Tong), and a graph data storage format (Liu) to arrive at the specific claimed invention. The motivations stated for each combination are generic (e.g., "so that low-level product categories are used as class labels" for Ding; "in order to allow for quick traversals" for Liu) and do not explain why a skilled artisan would have combined these specific teachings to arrive at a merchant- facing GNN-based product bundling system with automatic anchor identification. This level of piecemeal combination, without a clear unified rationale, strongly suggests that the claims themselves were used as the roadmap to combine the references which is- the hallmark of impermissible hindsight.”. In response, Examiner notes that although Ding and Tong have been withdrawn, as necessitated by Applicant’s amendment, in pars. [0038, 0041, 0044, 0047, 0051, 0055, 0060, 0065, and 0070] of the office action mailed 02/25/2026, Examiner provides obviousness statements with clear motivations representing a rationale from the perspective of one of ordinary skill in the art that establish prima facie cases for combining each of the references cited, which are all in the same or similar field of endeavor to Applicant’s claims before Applicant’s effective filing date. Therefore, the combination of said references does not constitute impermissible hindsight.
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 1-7, 10-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-patentable subject matter. The claims are directed to an abstract idea without significantly more. The judicial exception is not integrated into a practical application. The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception as further set forth in MPEP 2106.
Step 1: The claimed invention is analyzed to determine if it falls outside one of the four statutory categories of invention. See MPEP 2106.03
Claim(s) 1-7, 10-12 is/are directed to a method (i.e., Process, claims 13-16 are directed to a system (i.e., Machine), and claims 17-20 are directed to non-transitory computer readable memory (i.e., Item of Manufacture). Therefore, the claims are directed to patent eligible categories of invention. Accordingly, the claims satisfy Step 1 of the eligibility inquiry.
Step 2A, Prong 1: In prong one of step 2A, the claim(s) is/are analyzed to evaluate whether they recite a judicial exception. See MPEP 2106.04
Independent claims 1, 13, and 17 recite a method, a system, and a non-transitory computer readable memory for providing products for sale in an online store. As drafted, the limitations recited by the claims fall under the “Certain methods of organizing human activity” abstract idea group, directed to commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations (see MPEP § 2106.04(a)(2), subsection II), and “Mental Processes” abstract idea grouping by setting forth activities that could be performed mentally by a human (including an observation, evaluation, judgment, opinion) (see MPEP § 2106.04(a)(2), subsection III).
Independent claims 1, 13, and 17 recite the following abstract limitations:
“generating a graph from the product sales information with nodes of the graph representing products and edges of the graph between nodes representing products represented by the nodes having been sold in a single order; automatically identifying at least one anchor product based on one or more of the product sales information and product inventory data; identify at least one product recommended for sale with the respective anchor product in a respective product bundle”. But for the additional elements recited, these limitations could be accomplished mentally, such as by human observation, evaluation, judgement, opinion, or with the help of pen and paper.
“present the at least one product bundles to a merchant; receiving an indication of one or more of the product bundles selected to offer for sale; posting the selected one or more product bundles.”. But for the additional elements recited, these limitations fall under the “Certain Methods of Organizing Human Activity”, directed to commercial or legal interactions (including agreements in the form of contracts; legal obligations; advertising, marketing or sales activities or behaviors; business relations.
The dependent claims further narrow the abstract idea and introduce the following additional elements for consideration:
From claim 12: wherein training the recommendation model comprises learning product embeddings
Dependent claims 2-7, 10-11, 14-16, and 18-20 further narrow the abstract idea and do not introduce further additional elements for consideration.
Step 2A, Prong 2: An evaluation is made whether a claim recites any additional element, or combination of additional elements, that integrate the judicial exception into a practical application of the exception. See MPEP 2106.04(d).
Regarding the computing additional elements, namely at least one processor, at least one memory storing instructions from claim 13, and a non-transitory computer readable memory storing instructions from claim 17, these additional elements have been evaluated but fail to integrate the abstract idea into a practical application because they amount to using generic computing elements or instructions (software) to perform the abstract idea, similar to adding the words “apply it” (or equivalent), which merely serves to link the use of the judicial exception to a particular technological environment (generic computing environment). See MPEP 2106.05(f) and 2106.05(h). In addition, these limitations fail to provide an improvement to the functioning of a computer or to any other technology or technical field, fail to apply the exception with a particular machine, fail to apply the judicial exception to effect a particular treatment or prophylaxis for a disease or medical condition, fail to effect a transformation of a particular article to a different state or thing, and fail to apply/use the abstract idea in a meaningful way beyond generally linking the use of the judicial exception to a particular technological environment (generic computing environment).
With respect to the limitations training a recommendation model using the product sales information to generate a trained recommendation model, wherein the recommendation model comprises a graph neural network (GNN), and training the GNN using the generated graph, for each of the at least one anchor products, applying the respective anchor product to the trained model, generating a user interface, and to the online sales channel from the independent claims (1/13/17), and wherein training the recommendation model comprises learning product embeddings from dependent claim 12, these limitations fail to integrate the abstract idea into a practical application because they provide nothing more than mere instructions to implement an abstract idea on a generic computer. See MPEP 2106.05(f). MPEP 2106.05(f) provides the following considerations for determining whether a claim simply recites a judicial exception with the words “apply it” (or an equivalent), such as mere instructions to implement an abstract idea on a computer: (1) whether the claim recites only the idea of a solution or outcome i.e., the claim fails to recite details of how a solution to a problem is accomplished; (2) whether the claim invokes computers or other machinery merely as a tool to perform an existing process; and (3) the particularity or generality of the application of the judicial exception.
With respect to the limitations for receiving product sales information comprising order information indicating products sold together and order costs from the independent claims (1/13/17), these activities at most amount to insignificant extra-solution activity (e.g., mere data gathering), which does not integrate the abstract idea into a practical application, as noted in MPEP 2106.05(g).
Accordingly, because the Step 2A Prong One and Prong Two analysis resulted in the conclusion that the claims are directed to an abstract idea, additional analysis under Step 2B of the eligibility inquiry must be conducted in order to determine whether any claim element or combination of elements amount to significantly more than the judicial exception.
Step 2B: The claims are analyzed to determine whether any additional element, or combination of additional elements, is/are sufficient to ensure that the claims amount to significantly more than the judicial exception. This analysis is also termed a search for "inventive concept." See MPEP 2106.05.
Regarding the computing additional elements, namely at least one processor, at least one memory storing instructions from claim 13, and a non-transitory computer readable memory storing instructions from claim 17, these additional element(s) has/have been evaluated, but fail to add significantly more to the claims because they amount to using generic computing elements (computer hardware) or instructions/software (engine) to perform the abstract idea, similar to adding the words “apply it” (or an equivalent), which merely serves to link the use of the judicial exception to a particular technological environment (network computing environment, the internet, online) and does not amount to significantly more than the abstract idea itself. Applicant’s specification recites the computing additional elements at a high level of generality.
With respect to the limitations training a recommendation model using the product sales information to generate a trained recommendation model, wherein the recommendation model comprises a graph neural network (GNN), and training the GNN using the generated graph, for each of the at least one anchor products, applying the respective anchor product to the trained model, generating a user interface, and to the online sales channel from the independent claims (1/13/17), and wherein training the recommendation model comprises learning product embeddings from dependent claim 12, these limitations fail to add significantly more to the abstract idea because they provide nothing more than mere instructions to implement an abstract idea on a generic computer. See MPEP 2106.05(f). MPEP 2106.05(f) provides the following considerations for determining whether a claim simply recites a judicial exception with the words “apply it” (or an equivalent), such as mere instructions to implement an abstract idea on a computer: (1) whether the claim recites only the idea of a solution or outcome i.e., the claim fails to recite details of how a solution to a problem is accomplished; (2) whether the claim invokes computers or other machinery merely as a tool to perform an existing process; and (3) the particularity or generality of the application of the judicial exception. Therefore, the additional elements merely describe generic computing elements or computer-executable instructions (software) merely serve to tie the abstract idea to a particular operating environment, which does not add significantly more to the abstract idea. See, e.g., Alice Corp., 134 S. Ct. 2347, 110 USPQ2d 1976; Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015).
With respect to the limitations for receiving product sales information comprising order information indicating products sold together and order costs from the independent claims (1/13/17), these activities at most amount to insignificant extra-solution activity (e.g., mere data gathering), which does not add significantly more, as noted in MPEP 2106.05(g). Additionally, the mere data gathering extra-solution activity have been recognized as well-understood, routine, and conventional, and thus insufficient to add significantly more to the abstract idea. See MPEP 2106.05(d) - Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); TLI Communications LLC v. AV Auto. LLC, 823 F.3d 607, 610, 118 USPQ2d 1744, 1745 (Fed. Cir. 2016) (using a telephone for image transmission); OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network)).
In addition, when taken as an ordered combination, the ordered combination adds nothing that is not already present as when the elements are taken individually. Their collective functions merely provide generic computer implementation. Therefore, when viewed as a whole, these additional claim elements do not provide meaningful limitations to amount to significantly more than the abstract idea itself. The ordered combination of elements in the claims (including the limitations inherited from the parent claim(s)) add nothing that is not already present as when the elements are taken individually. There is no indication that the combination of elements improves the functioning of a computer or improves any other technology. Their collective functions merely provide generic computer implementation. Accordingly, the subject matter encompassed by the dependent claims fails to amount to significantly more than the abstract idea itself.
Claim Rejections - 35 USC § 103
This application currently names joint inventors. In considering patentability of the claims the examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicant is advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1, 10, 12, 13, and 17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Das Gupta et al. (US 20230162258 A1, hereinafter “Das Gupta”), in view of Frazer et al. (US 20100049538 A1, hereinafter “Frazer”).
Regarding claims 1/13/17: Das Gupta teaches a computer implemented method ([Abstract] Methods and systems for providing a product bundle recommendation are disclosed.), a system comprising at least one processor and at least one memory storing instructions ([0003] An example system includes a computing system including a data store, a processor, and a memory communicatively coupled to the processor. The memory stores instructions executable by the processor), and a non-transitory computer readable memory ([0058] The mass storage device 1314 is connected to the one or more processors 1302 through a mass storage controller (not shown) connected to the system bus 1322. The mass storage device 1314 and its associated computer-readable data storage media provide non-volatile, non-transitory storage for the computing system 1300. Although the description of computer-readable data storage media contained herein refers to a mass storage device, such as a hard disk or solid state disk, it should be appreciated by those skilled in the art that computer-readable data storage media can be any available non-transitory, physical device or article of manufacture from which the central display station can read data and/or instructions.) with limitations for:
receiving product sales information comprising order information indicating products sold together and order costs; ([Abstract] Data relating to historical customer shopping activity and data describing online shopping sessions in which products are selected for addition to historical shopping carts to which the first product was previously added is accessed.; [0005] In a third aspect, example systems for providing a product recommendation are described. An example system includes a computing system including a data store, a processor, and a memory communicatively coupled to the processor. The memory stores instructions executable by the processor to: receive, from a customer device, a selection of a first product offered for sale via a retailer website; access the data store to retrieve customer add to cart data, product price data, and degree of diversity data, wherein: the customer add to cart data is based on historical customer shopping activity and includes data describing online shopping sessions in which products are selected for addition to a historical shopping cart to which the first product was previously added, the customer add to cart data further includes data describing an order in which the products are added to the historical shopping cart to which the first product was previously added, and the degree of diversity data includes product similarity data having a product similarity score between two products based on a comparison of product attributes; based on the selection of the first product and the customer add to cart data, generate a list of bundled products, wherein the list of bundled products includes one or more additional products that, in the historical customer shopping activity, have been selected after the selection of the first product more than a threshold number of times; based on a price of the first product, generate a product recommendation price range; based on a product category of the first product, determine a degree of diversity for the product recommendation; select from the list of bundled products at least one bundled product, the selected at least one bundled product having a price within the product recommendation price range and meeting the determined degree of diversity; and present the at least one bundled product on a user interface on the customer device as a recommendation for purchase with the first product.);
training a recommendation model using the received product sales information to generate a trained recommendation model, ([0003] In general, the present disclosure relates to a product bundles recommendation server. In a first aspect, example systems for providing a product recommendation on a retailer website are described. An example system includes a computing system including a data store, a processor, and a memory communicatively coupled to the processor. The memory stores instructions executable by the processor to: receive, from a customer device, a selection of a first product offered for sale via the retailer website; access the data store to retrieve customer add to cart data, product price data, and degree of diversity data, wherein: the customer add to cart data is based on historical customer shopping activity of a plurality of customers and includes data describing online shopping sessions in which products are selected for addition to historical shopping carts to which the first product was previously added, and the customer add to cart data further includes data describing an order in which the products are added to the historical shopping carts to which the first product was previously added; based on the selection of the first product and the customer add to cart data, generate a list of bundled products having a complementary relationship with the first product, wherein, in the historical customer shopping activity, the bundled products have been selected after the selection of the first product more than a threshold number of times; select at least one bundled product from the list of bundled products; and present the at least one bundled product on a user interface on the customer device as a recommendation for purchase with the first product.; [0039] a model may be trained using existing edges or links determined from add-to-cart data in order to predict the probability of an edge or link between two nodes within the network for which an edge is not otherwise derived directly from the add to cart data.);
wherein the recommendation model comprises a graph neural network (GNN) ([0039] FIG. 5 illustrates an example flowchart 500 for inferring or predicting product relationships within a product relationship network using a Graph Neural Network (GNN).);
and training the GNN using the generated graph; ([0039] FIG. 5 illustrates an example flowchart 500 for inferring or predicting product relationships within a product relationship network using a Graph Neural Network (GNN). The example flowchart as shown in FIG. 5 may be carried out by the infer complementary product engine 220. In the examples describe herein, a machine-learning model may include one or more GNNs configured to learn an edge function between nodes within a fully-connected network. For example, a model may be trained using existing edges or links determined from add-to-cart data in order to predict the probability of an edge or link between two nodes within the network for which an edge is not otherwise derived directly from the add to cart data. A hidden state of each node evolves over time by exchanging information with its neighboring nodes via message passing. Weights in the model are shared across nodes, which gives the model the ability to handle a different number of inputs, which is relevant because the number of products in a recommended bundle can vary.)
automatically identifying at least one anchor product based on one or more of the product sales information and product inventory data; ([Fig. 2] Add To Cart Data 210; [0003] access the data store to retrieve customer add to cart data, product price data, and degree of diversity data, wherein: the customer add to cart data is based on historical customer shopping activity of a plurality of customers and includes data describing online shopping sessions in which products are selected for addition to historical shopping carts to which the first product was previously added, and the customer add to cart data further includes data describing an order in which the products are added to the historical shopping carts to which the first product was previously added; based on the selection of the first product and the customer add to cart data, generate a list of bundled products having a complementary relationship with the first product; [0021] the product bundles recommendation server allows a retailer to recommend complementary or bundled products to a user or a customer based on a user viewing a first item or product, also called a “seed product,” that is offered for sale on a retailer website. Additionally, the product bundles recommendation server allows a retailer to apply limiting factors or filters to further control the products within a bundle recommended to a user.; [0022] Complementary relationships between products may be identified and used to generate a list of bundled products to present to a user based on a user first selecting a seed product. For example, recommended product bundles may be meaningful in terms of pricing by including a price filter to ensure all recommended products within a bundle are priced similarly to (or in some logical relation to) a seed product that has been selected. [0025] The product bundles recommendation server 106 communicates with one or more databases 112 to access product data and data associated with the order in which users add products to an online shopping cart.; Based on the user selecting the seed product 712 by adding the seed product 712 to the online shopping cart 710, a product bundle 714 is presented. See also pars. [0005] and [0022]. Examiner notes that one of ordinary skill in the art would reasonably consider add to cart data as product sales information.);
for each of the at least one anchor products, applying the respective anchor product to the trained model to identify at least one product recommended for sale with the respective anchor product in a respective product bundle; ([0003] access the data store to retrieve customer add to cart data, product price data, and degree of diversity data, wherein: the customer add to cart data is based on historical customer shopping activity of a plurality of customers and includes data describing online shopping sessions in which products are selected for addition to historical shopping carts to which the first product was previously added, and the customer add to cart data further includes data describing an order in which the products are added to the historical shopping carts to which the first product was previously added; based on the selection of the first product and the customer add to cart data, generate a list of bundled products having a complementary relationship with the first product; [0025] In the example shown, the retail server 108 may access a product bundles recommendation server 106. The product bundles recommendation server 106 operates to provide product recommendations to be displayed on a retailer website after a user has selected a seed product. The product bundles recommendation server 106 communicates with one or more databases 112 to access product data and data associated with the order in which users add products to an online shopping cart. Databases 112 may be external and updated in real time as new data and information becomes available. Based on data from databases 112, the product bundles recommendation server 106 generates a list of bundled products that are complementary to a seed product that a user has selected.);
generating a user interface to present the at least one product bundles to a merchant; ([0034] In the example shown, after a list of complementary products has been identified and any applicable filters applied by the filter application engine 224, at least one complementary product from the list may be transmitted over a network 204 to a customer device 202, where the at least one complementary product may be presented on a user interface of the customer device 202 as a recommendation of a product to bundle for purchase with a seed product.);
receiving an indication of one or more of the product bundles selected to offer for sale on an online sales channel; ([0021] the product bundles recommendation server allows a retailer to apply limiting factors or filters to further control the products within a bundle recommended to a user.);
and posting the selected one or more product bundles to the online sales channel. ([0011] FIG. 6 illustrates an example user interface displaying a recommended product bundle).
Das Gupta doesn’t explicitly teach:
and wherein training the recommendation model comprises: generating a graph from the product sales information with nodes of the graph representing products and edges of the graph between nodes representing products represented by the nodes having been sold in a single order;
Frazer teaches
and wherein training the recommendation model comprises: generating a graph from the product sales information with nodes of the graph representing products and edges of the graph between nodes representing products represented by the nodes having been sold in a single order; ([0121] The most natural representation of pair-wise relationships between entities abstraction is a structure called Graph. Formally, a graph contains: [0122] a set of Nodes representing entities (products or customers); and [0123] a set of Edges representing strength of relationships between pairs of nodes (entities).; [0124] FIG. 6 shows an example of a insight/relationship determination module Consistency Graph created using the transaction data from a Grocery retailer. In FIG. 6, nodes represent products and edges represent consistency relationships between pairs of nodes. This graph has one node for each product at a category level of the product hierarchy. These nodes are further annotated or colored by department level. In general, these nodes could be annotated by a number of product properties, such as total revenue, margin per customers, and the like. There is a weighted edge between each pair of nodes. The weight represents the consistency with which the products in those categories are purchased together.).
It would have been obvious to one of ordinary skill in the art, at the time of applicant’s invention, to combine Das Gupta with Frazer’s feature(s) listed above. One would’ve been motivated to do so in order to create co-promotion campaigns (Frazer; [0136]). By incorporating the teachings of Frazer, one would’ve been able to generate a graph where nodes represent products and edges represent a relationship between products (i.e., sold together).
Regarding Claim 10: Das Gupta teaches
receiving additional product sales information, the additional product sales information comprising recent product sales information obtained more recently than the product sales information; ([Abstract] Data relating to historical customer shopping activity and data describing online shopping sessions in which products are selected for addition to historical shopping carts to which the first product was previously added is accessed.; [0003] In general, the present disclosure relates to a product bundles recommendation server. In a first aspect, example systems for providing a product recommendation on a retailer website are described. An example system includes a computing system including a data store, a processor, and a memory communicatively coupled to the processor. The memory stores instructions executable by the processor to: receive, from a customer device, a selection of a first product offered for sale via the retailer website; access the data store to retrieve customer add to cart data, product price data, and degree of diversity data, wherein: the customer add to cart data is based on historical customer shopping activity of a plurality of customers and includes data describing online shopping sessions in which products are selected for addition to historical shopping carts to which the first product was previously added, and the customer add to cart data further includes data describing an order in which the products are added to the historical shopping carts to which the first product was previously added; based on the selection of the first product and the customer add to cart data, generate a list of bundled products having a complementary relationship with the first product, wherein, in the historical customer shopping activity, the bundled products have been selected after the selection of the first product more than a threshold number of times; select at least one bundled product from the list of bundled products; and present the at least one bundled product on a user interface on the customer device as a recommendation for purchase with the first product.).
Das Gupta doesn’t explicitly teach:
and updating the graph using the additional product sales information.
Frazer teaches:
and updating the graph using the additional product sales information. ([0058] Due to expected changes in the environment (economy, competitors), changes in customer behavior (fashions), and the increasing information collected over time, it is important to occasionally update (some of) the underlying predictive models, both in terms of their structure and their parameters.; [0132] Node based Sub-graphs are created by selecting a subset of the nodes and therefore, by definition, keeping only the edges between selected nodes. For example, in a product graph, one might be interested in analyzing sub-graph of all products within the electronics department or clothing merchandise, or only the top 10% high value products, or products from a particular manufacturer, etc. Similarly, in a customer graph, one might be interested in analyzing customers in a certain segment, or high value customers, or most recent customers, etc.; [0133] Edge based Sub-graphs are created by pruning a set of edges from the graph and therefore, by definition, removing all nodes that are rendered disconnected from the graph. For example, one might be interested in removing low consistency strength edges (to remove noise), and/or high consistency strength edges (to remove obvious connections), or edges with a support less than a threshold, etc.; [0158] Adaptive: As noted before, both the product space and the customer space is very dynamic. New products are added, customers change over time, new customers get added to the market place and purchase trends change over time. To cope up with these dynamics of the modern day retail market, one needs a system that can quickly assimilate the newly generated transaction data and adapt its models accordingly. The insight/relationship determination module 320 is very adaptive as it can update its graph structures quickly to reflect any changes in the transaction data.).
It would have been obvious to one of ordinary skill in the art, at the time of applicant’s invention, to combine modified Das Gupta with Frazer’s feature(s) listed above. One would’ve been motivated to do so, so that graphs are intuitive and easy to interpret by store managers and corporate executives both to explain results and make decisions (Frazer; [0160]). By incorporating the teachings of Frazer, one would’ve been able to update a graph when additional data is received.
Regarding claims 12: Das Gupta teaches:
wherein training the recommendation model comprises learning product embeddings. ([0040] The GNN model may further include a product embedding layer 504 to learn vector representations of nodes in the graph and improve prediction of links.).
Claim(s) 2-4, 6, 14-16, and 18-20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Das Gupta et al. (US 20230162258 A1, hereinafter “Das Gupta”), in view of Frazer et al. (US 20100049538 A1, hereinafter “Frazer”), as applied to claims 1/13/17 above, in further view of Kondapaneni (US 12314985 B1, hereinafter “Kondapaneni”).
Regarding claims 2/14/18: Das Gupta doesn’t explicitly teach:
wherein identifying the at least one anchor product is based on inventory data.
Kondapaneni teaches:
wherein identifying the at least one anchor product is based on inventory data. ([Column 3, Lines 31-37] embodiments of the following disclosure provide a system and method to analyze assortment planning and potential combinations of product attributes, determine hindsights and useful information from the analysis, and to recommend potential combinations of product attributes (such as, for example, preparing additional red T-shirts with a particular design logo and collar style)… Embodiments utilize the attribute strengths to generate recommendations for product-store and/or product/attribute combinations based on the one or more metrics.; [Column 3, Lines 15-21] Server 122 is configured to receive and transmit inventory data, including item identifiers, pricing data, attribute data, inventory levels, and other like data about one or more items or products at one or more locations in supply chain network 100. Server 122 stores and retrieves inventory data from database 124 or from one or more locations in supply chain network 100.; [Column 13, Lines 5-6] hindsight engine 208 filters the attributes according to individual attribute strength).
It would have been obvious to one of ordinary skill in the art, at the time of applicant’s invention, to combine Das Gupta with Kondapaneni’s feature(s) listed above. One would’ve been motivated to do so in order to include current or projected inventory quantities or states, order rules, or explanatory variables (Kondapaneni; [Column 3, Lines 22-24]). By incorporating the teachings of Kondapaneni, one would’ve been able to select an anchor using inventory data.
Regarding claims 3/15/19: Das Gupta doesn’t explicitly teach:
wherein identifying the at least one anchor product uses a turnover ratio for each product.
Kondapaneni teaches:
wherein identifying the at least one anchor product uses a turnover ratio for each product. ([Abstract] The system provides for classifying each of one or more products in a product display area of a retail entity as high-performers or low-performers according to selected metrics.; [Column 10, Lines 43-48] Metrics data 230 may store data related to one or more metrics used to measure the performance of one or more products, stores, and/or attributes. Example metrics include revenue, profit, inventory turnover, or any other metric according to the preference of a supply chain planner or other user of assortment planner 110.).
It would have been obvious to one of ordinary skill in the art, at the time of applicant’s invention, to combine modified Das Gupta with Kondapaneni’s additional feature(s) listed above. One would’ve been motivated to do so, so that aggregated data 232 may be a set of product-store combinations associated with a metric (Kondapaneni; [Column 10, Lines 51-52]). By incorporating the teachings of Kondapaneni, one would’ve been able to select an anchor using inventory turnover data.
Regarding claims 4/16/20: Das Gupta doesn’t explicitly teach:
wherein the identified at least one anchor product comprise one or more of: products having a highest turnover ratio; and products having a lowest turnover ratio.
Kondapaneni teaches:
wherein the identified at least one anchor product comprise one or more of: products having a highest turnover ratio; and products having a lowest turnover ratio. ([Column 10, Line 59 – Column 11, Line 4] Classification thresholds data 234 may store data specifying the qualities, elements, or transition thresholds between classifications of products (including, but not limited to, classifying products as “high-performing” and “low-performing”). For example, one classification threshold may be 50%, that is, a top half of product-store combinations, according to the metric selected, will be classified as “high-performing” while the corresponding bottom half will be classified as “low-performing.” Other classification thresholds, such as 80% (meaning only the top 20% is classified as high-performing) or 20% (meaning the top 80% is classified as high-performing) may be used according to particular needs or preferences.).
It would have been obvious to one of ordinary skill in the art, at the time of applicant’s invention, to combine modified Das Gupta with Kondapaneni’s additional feature(s) listed above. One would’ve been motivated to do so in order to classify high-performing and low-performing products (Kondapaneni; [Column 12, Line 7]). By incorporating the teachings of Kondapaneni, one would’ve been able to identify products having a highest turnover ratio, and/or lowest turnover ratio.
Regarding claim 6: Das Gupta doesn’t explicitly teach:
wherein the at least one anchor product is identified as a popular or unpopular product.
Kondapaneni teaches:
wherein the at least one anchor product is identified as a popular or unpopular product. ([Column 11, Line 65 – Column 12, Line 8] At activity 302, assortment planner 110 loads historical data 220 into the input data of database 114. Aggregation engine 206 accesses historical data 220, which may include product placement data 222, product attributes data 224, product sales history data 226, and/or planogram dimension data 228, into the input data of database 114. At activity 304, user action processor 202 detects input to one or more input devices 152, and in response to the detected input, user action processor 202 selects one or more metrics with which to aggregate data and classify high-performing and low-performing products.).
It would have been obvious to one of ordinary skill in the art, at the time of applicant’s invention, to combine modified Das Gupta with Kondapaneni’s additional feature(s) listed above. One would’ve been motivated to do so in order to select product margin, product revenue, total profit, and/or any other metric or group of metrics (Kondapaneni; [Column 12, Lines 9-11]). By incorporating the teachings of Kondapaneni, one would’ve been able to identify popular and/or unpopular products.
Claim(s) 5 is/are rejected under 35 U.S.C. 103 as being unpatentable over Das Gupta et al. (US 20230162258 A1, hereinafter “Das Gupta”), in view of Frazer et al. (US 20100049538 A1, hereinafter “Frazer”), in further view of Kondapaneni (US 12314985 B1, hereinafter “Kondapaneni”), as applied to claim 4 above, in further view of Mari Kylänlahti-Harmaala, The product assortment strategy from a store manager’s point of view, Master’s Thesis, Häme University of Applied Sciences, Finland, 2020 (hereinafter “Kylänlahti-Harmaala”).
Regarding claim 5: Das Gupta doesn’t explicitly teach:
wherein the turnover ratio for a product is determined by: determining an average stock level of the product over a time period; determine a cost of goods for the product by multiplying a merchant’s product cost by a total number of the products sold over the time period; and determining the product’s turnover ratio by dividing the cost of goods by the average stock level.
Kondapaneni teaches:
wherein the turnover ratio for a product is determined by: determining an average stock level of the product over a time period; determine a cost of goods for the product by multiplying a merchant’s product cost by a total number of the products sold over the time period; and determining the product’s turnover ratio by dividing the cost of goods by the average stock level. ([Page 4, Inventory Turnover] The inventory turnover is calculated from cost of goods sold / average inventory= inventory turnover).
It would have been obvious to one of ordinary skill in the art, at the time of applicant’s invention, to combine modified Das Gupta with Kylänlahti-Harmaala’s feature(s) listed above. One would’ve been motivated to do so in order to measure how many times the inventory is replaced over a certain time period (Kylänlahti-Harmaala; [Page 4, Inventory Turnover]). By incorporating the teachings of Kylänlahti-Harmaala, one would’ve been able to determine the turnover ratio as described by Applicant’s claim.
Claim(s) 7 is/are rejected under 35 U.S.C. 103 as being unpatentable over Das Gupta et al. (US 20230162258 A1, hereinafter “Das Gupta”), in view of Frazer et al. (US 20100049538 A1, hereinafter “Frazer”), in further view of Kondapaneni (US 12314985 B1, hereinafter “Kondapaneni”) as applied to claim 6 above, in further view of Hendrick et al. (US 20130346234 A1, hereinafter “Hendrick”).
Regarding claim 7: Das Gupta doesn’t explicitly teach:
wherein a popularity of the at least one anchor product is determined for first-time customers and returning customers.
Hendrick teaches:
wherein a popularity of the at least one anchor product is determined for first-time customers and returning customers. ([0038] A new customer can be configured for the system using the shared infrastructure. In the disclosed recommendations system, it is useful to consider the distinction between a system that only supports a single customer and a system that requires additional infrastructure to support new customers. The recommendation system supports the configuration of "organizations", where an organization corresponds to a collection of catalogs, user activity data, and models computed from the catalog and user activity data. An organization is an abstraction that represents a business entity that operates one or more storefronts or mobile applications that will interact with the recommendations system. A realistic example is that ringtones and applications may be handled by different storefronts, but these storefront are operated by a single business entity and serve the same base of customers. The system supports configuration of multiple organizations. The recommendation system scales to support an arbitrary number of organizations by simply adding storage capacity (e.g., Hadoop and HBase nodes) and FE service capacity. Configuring a new customer involves: configuring ingestion of the catalogs for the organization (what they are, where they are stored, how are they parsed), configuring ingestion of the user activity data (what the sources are, where it's stored, how it's parsed), and configuring a Web service API provider for the organization (the API methods available to the organization, the storefronts, and mobile applications that are supported). Each organization is also configured with "channels" which roughly correspond to the storefront and mobile applications that call for recommendations. For example, an organization might support two storefronts, each with its own channel. A channel configuration defines what mix of catalogs are available to be recommended. While an organization may support several catalogs over all, each channel may only be configured to support a subset of the catalogs.; [0046] The system supports building two types of models: associations models and popularity models. These models are used in the recommendations processing, described further below. The associations models include item-to-item associations, which are represented in general terms as:: src1->(tgt1,w1), (tgt2,w2), . . . , meaning that a Source 1 is associated with an n-tuple comprising items (Target 1, Weight 1), (Target 2, Weight 2), and so forth, where the weighting is a value assigned to each respective target. The popularity models comprise a ranking of the most popular items, such as the top "N" most popular items, and are represented in general terms as: (item1,w1), (item2,w2), . . . , meaning that the items are represented in a data file as multiple pairs of items and their respective popularity rankings or weights. The two types of models are used in the online recommendations processing.).
It would have been obvious to one of ordinary skill in the art, at the time of applicant’s invention, to combine modified Das Gupta with Hendrick’s feature(s) listed above. One would’ve been motivated to do so in order to compute a set of recommendations to fulfill a request (Hendrick; [0046]). By incorporating the teachings of Hendrick, one would’ve been able to determine product popularity for first-time customers and existing customers.
Claim(s) 11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Das Gupta et al. (US 20230162258 A1, hereinafter “Das Gupta”), in view of Frazer et al. (US 20100049538 A1, hereinafter “Frazer”), as applied to claim 10 above, in further view of Liu et al. (US 20230128180 A1, hereinafter “Liu”).
Regarding claims 11: Das Gupta teaches:
and a parameter indicating a number of recommended products to provide. ([0004] selecting at least one bundled product from the list of bundled products; and presenting the at least one bundled product on a user interface on the customer device as a recommendation for purchase with the first product.).
Das Gupta doesn’t teach:
providing a tensor of edge indices of the updated graph; providing a tensor of edge weights; a tensor of one or more node indices of the one or more anchor products;
Liu teaches:
providing a tensor of edge indices of the updated graph; providing a tensor of edge weights; a tensor of one or more node indices of the one or more anchor products; ([0032] With the CSR format, nodes and edges of a graph may be stored in separate arrays, with the indices of these arrays corresponding to node identifiers and edge identifiers.; [0033] With the COO format, the edges of a graph may be stored as a list of tuples, where the tuple of each edge may include source node identifier, destination node identifier, suitable attribute information of the edge, or any combination thereof. For example, as shown in FIG. 1, the edge e121 may be represented as [n111, n112]. In some embodiments, if the edges are weighted, an extra entry in the tuple may be created to record the value of the weight for each edge. For example, as shown in FIG. 1, if the edge 121 has a weight value of 3, the tuple representation is [n111, n112, 3]. In some embodiments, graph data stored in the COO format is unsorted. For example, the tuples for all the edges may be stored in any order. The COO format may be suitable for graphs that are frequently updated. For example, if a new node or a new edge is added to the graph, the COO format simply needs to add one or more tuples for the newly added data. Examiner notes that one of ordinary skill in the art would reasonably consider the tuples and the arrays disclosed by Liu as equivalent to a tensor.).
It would have been obvious to one of ordinary skill in the art, at the time of applicant’s invention, to combine modified Das Gupta with Liu’s feature(s) listed above. One would’ve been motivated to do so in order to allow for quick traversals of the graph’s nodes and edges (Liu; [0032]). By incorporating the teachings of Liu, one would’ve been able to provide a tensor of edge weights, edge indices, and node indices.
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 GABRIEL J TORRES CHANZA whose telephone number is (571)272-3701. The examiner can normally be reached Monday thru Friday 8am - 5pm 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, Brian Epstein can be reached on (571)270-5389. 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.
/G.J.T./Examiner, Art Unit 3625
/BRIAN M EPSTEIN/Supervisory Patent Examiner, Art Unit 3625
Prosecution Insights