DETAILED ACTION
This action is responsive to the claims filed on September, 9, 2025. Claims 1-8, 10-17 and 19-20 are under examination.
The drawings are objected to.
Claims 1-8, 10-17, and 19-20 are rejected under 35 USC 112(a).
Claims 1-8, 10-17, and 19-20 are rejected under 35 USC 112(b).
Claims 1-8, 10-17, and 19-20 are rejected under 35 USC 101.
Claims 1-4, 6, 10-13, 15, and 19 are rejected under 35 USC 103 over Wang, Antipov, and Channer.
Claims 5, 8, 14, 17, and 20 are rejected under 35 USC 103 over Wang, Antipov, Channer, and Brownlee.
Claims 7 and 16 are rejected under 35 USC 103 over Wang, Antipov, Channer, and Lundberg.
Response To Arguments
Drawing Objections: FIGs. 3AC, 3AD, 3C, and 3D are still undecipherable after being rendered into PDF form by the PTO. The “OOB SAMPLE 2” overlapping with the dots in FIGs. 3H and 3AB (making it unclear what the sample 2 is), the graphic on the right side of FIG. 3I, the information displayed on the monitors in FIGs. 3J and 3K. The Applicant is responsible for providing PDF files with drawings that are legible when rendered in the USPTO systems. There is a practical limit to how much information an Applicant can place in a single drawing on a single sheet, in part because the rendering by the USPTO software will the resolution of the images. In order to comply with the drawing rules, an Applicant may be compelled to produce images over multiple sheets with connecting elements in order to accommodate the limited resolution of the images rendered by the USPTO systems. The Applicant is advised to further deconstruct the images into smaller images with an accompanying amendment to the specification to accommodate the composite images. Because the images are presented with individual elements having connections therebetween, there is no need to place so much information on each page. The pages can be further parsed.
Claim Objection: The arguments and amendments have been considered and are persuasive. The claim objection has been withdrawn.
35 USC 112(a) Rejections: The Applicant’s arguments and amendments have been considered but are not persuasive. The Applicant has demonstrated that Shapley value determinations represent the quantify/ing steps of the rejected claims. We will call this (I). The Applicant has also shown that the estimate/ing steps of the rejected claims result in the determination of a product uniqueness index, a product transferable volume, and/or a product lost volume. We will call this (II). This Final Action will demonstrate how there is no link in the specification between (I) and (II). That is, there is no teaching in the entire specification of how to use the result of the quantifying step (I) to conduct the estimating step (II). For simplicity, I will call the link between (I) and (II) the “nexus.”
The Applicant’s specification fails to provide a meaningful nexus that demonstrates possession and enablement of the determination of the product uniqueness index, a product transferable volume, and/or a product lost volume from Shapley values. None of the equations or words of the written description nor any of the figures (the ones sufficiently readable) remedy this. The specification merely teaches the one and then the other without any linking between the two. None of the evidence presented by the Applicant refutes this.
Specifically, on page 16 of the response, the Applicant points to paragraphs [0030]-[0031], [0036], [0049]-[0059], and Figures 1, 2, 3A, 3H, 3I, 3J, and 3K. None of these elements describe, enable, or even MENTION the “nexus.”
In part (A) presented on pages 17-21 of the response, the Applicant demonstrates part (I). There is no demonstration of the nexus. The Applicant merely restates that the Shapley values are described.
In the Support from Drawings section on pages 20-23, the Applicant further demonstrates the determination of Shapley values. When the Applicant finally addresses the “nexus” on page 21, second paragraph, the Applicant states:
The Office Action rejects claims 1-20 on the ground that the specification does not sufficiently describe how Shapley value are used to calculate a product uniqueness index, transferable volume, or lost volume. The Examiner points to Hoberg et al. (2010) and contends that, because Hoberg employed HHI/C4 indices and the specification does not provide explicit guidance for substituting SHAP values, a person of ordinary skill would conclude that Applicant lacks possession. Applicant further submits that converting SHAP weight vectors into a uniqueness or similarity score and then using that score within routine simulation/business-rule logic to allocate transferable v. lost volume is a matter of ordinary skill and does not require adoption of Hoberg’s HHI/C4 metrics. A skilled artisan in retail analytics/data science would routinely apply well-known vector similarity/distance techniques (for example, cosine similarit, Euclidian distance, or weight overlap) or other standard allocation heuristics to SHAP vectors to generate a uniqueness index and then combine that index with business parameters (baseline volumes, elasticity, channel share) to estimate transferred and lost volumes, exactly the workflow the specification discloses in Paragraph [0036], [0049]-[0053] and Figures 3J-3K. Thus the presence of Hoberg on the record does not show lack of possession, the present disclosure itself provides the conceptual bridge and system elements for performing the conversion and simulation.
The Examiner respectfully disagrees. The introduction of Hoberg was to demonstrate that product uniqueness index is esoteric, a scarcely used metric in the art. Further, the math in Hoberg is difficult and does not involve the use of Shapley values. To determine how to use Shapley values in this complicated math would require manipulation beyond what is known in the art and would require unduly burdensome deconstruction at best, and a complete reworking of the math that is in a consumer demand field different from the Shapley determinations. For example, HHI is a measurement of market share, not a measurement of significance in making a purchase decision. The Hoberg reference is evidence that the specification fails to show any conceptual bridge representing the “nexus” with regard to product uniqueness index. If the claim recites “product uniqueness index,” which the independent claims do, then the “nexus” between the Shapley values of (I) and the product uniqueness index (II) must be demonstrated in the specification. The “nexus” is not demonstrated. Neither is the nexus demonstrated for the product transferable volume or a product lost volume. In order to enable and describe the full scope of the claims, the “nexus” between the Shapley values (I) and ALL of the product uniqueness index, the product transferable volume, and the product lost volume, must be explicitly present in the specification.
The Applicant continues on pages 22-23 to state that the possession (written description) is demonstrated through an adequate combination of structure and function and that the specification need not include each line of code for implementing a solution. The Examiner did not assert that code or even pseudo code is required. The Applicant provides further evidence that Shapley values (I) are determined and says that the “quantifying” (I) identifies concrete structural components, data flows, and explainability techniques by which the functions are achieved. Specifically, the Applicant states:
The specification's disclosure of an integrated system also significantly bolsters the written description. FIG. 2 depicts SHAP metadata (236) and demand-transferability modules (238) in the overall architecture, showing where quantified attribute weights are stored and how they are accessed for simulation. The data-preprocessing, ML modeling, and decomposition engines are described as part of a distributed system, enabling complex scenario simulations and reporting to be performed efficiently. This architecture shows possession by tying together (i) the upstream quantification of attribute contributions, (ii) storage of those weights as metadata, and (iii) downstream simulation engines that apply them to demand transferability estimation. Per Vasudevan Software, Inc. v. MicroStrategy, Inc., 782 F.3d 671, 681-83 (Fed. Cir. 2015), disclosure of such functional and structural integration is sufficient without requiring line-by-line code. The Drawings serve as visual evidence of possession: FIG. 2 shows SHAP metadata and demand-transferability modules within the system. FIG. 3A illustrates scenario simulation workflows. FIG. 3J shows computation of uniqueness indices and other attribute-based outputs. FIG. 3K depicts transferred and lost volume projections under alternative scenarios.
The assertion here is that because the (II) is demonstrated downstream of (I) with an expressed purpose, there is sufficient disclosure. Again, none of this information demonstrates the “nexus” necessary. The cited FIGs. show results to be achieved without any “nexus” at all.
The Applicant continues:
A skilled artisan in retail analytics and data science would readily understand from this disclosure how to implement the claimed transferability estimation. The combination of SHAP- based attribute weights, explicit definition of uniqueness/transferable/lost volume outputs, and simulation figures demonstrates possession of the claimed limitation. Applicant respectfully submits that the rejection under 35 U.S.C. §112(a) (written description) should be withdrawn, as the specification, read as a whole, conveys possession of the claimed invention consistent with Ariad Pharm., Inc. v. Eli Lilly & Co., 598 F.3d 1336, 1351 (Fed. Cir. 2010) (en banc) and Vasudevan, 782 F.3d at 682.
The Examiner maintains that the ORDINARILY SKILLED artisan would not be certain as to what a product uniqueness index is, as it is rarely used by anyone for anything. Even if the skilled artisan had heard of the obscure Hoberg paper, that person would now know how to calculate any of a product uniqueness index, transferable product volume, or lost product volume from Shapley values without some guidance. The Applicant appears to admit that this guidance is lacking and relies on the skilled artisan to fill in the gaps. This illustrates that the Applicant’s specification fails to demonstrate the “nexus” in a manner that shows possession or enablement. Further, the Applicant has asserted without evidentiary demonstration that the skilled artisan would know how to fill in the “nexus” without providing an example or even a suggestion (evidence) of how.
With regard to the feature “estimating the demand transferability,” addressed on pages 23-25, the same logic applies. The calculation of Shapley values (I)is separately described from the “estimation of demand transferability” without any demonstration of how. On page 23 of the response, the Applicant looks to paragraph [0036], which merely states a desired outcome. Paragraph [0036] provides a definition of demand transferability (II) but provides no information about how it is determined from Shapley values. This does not address the “nexus.” On page 24 of the response, the Applicant looks to paragraphs [0049]-[0051] that Shapley values (I) are “used as inputs to simulation workflows that model business scenarios such as SKU delisting, new product introduction, or attribute/price modifications (II).” This is a result to be achieved but no “nexus.” The Applicant then provides conclusory logic (“simulations apply attribute weights to assess product similarity and uniqueness, enabling the system to project how much of a delisted product’s sales volume would migrate to other products in the portfolio (transferable volume) versus how much would disappear entirely from the market (lost volume)”) to attempt to demonstrate that there is some “nexus.” First of all, this, if introduced to the specification or claims, would be new matter not present in the original specification. Further, this still leaves significant gaps in how the demand transferability is estimated, based on the Shapley values. This still provides no “nexus.”
The Applicant then refers to paragraphs [0036], [0049]-[0053], most of which have already been discussed, to attempt to demonstrate that the “weights are the foundation for computing uniqueness and projecting volume shifts.” Again, How? The Applicant then states, “[t]he attribute importance arrays act as numerical signatures of each product, allowing the system to assess how similar or distinct products are within the portfolio. A product with a highly unique attribute-importance profile will be less substitutable, leading to a higher proportion of lost volume if removed; conversely, products with overlapping importance profiles are likely to absorb displaced sales.” Again, this subject matter is not in the specification and would be new matter if the Applicant attempted to amend this quote into the specification. However, even if it were supported by the specification, this would still not adequately describe or enable the “nexus” to a person of ordinary skill in the art.
The Drawings, to the extent the text can be discerned do not provide any elements of the “nexus” other than arrows between elements that qualify as (I) or (II). On page 25 of the response, the Applicant attempts to assert that the FIGs. 3A, 3J, and 3K provide support for the “nexus.” FIG. 3A is not so organized as to illustrate any nexus. In fact, there is no connection between “Insight To Action Value Realization Support” 312 and any other element. FIG. 3J illustrates that data quality assessment yields attribute modeling, which yields incrementality analysis. The features illustrated in the monitors are not discernible. This does not provide any description of a nexus between Shapley values and anything. FIG. 3K generically illustrates a simulation set-up and simulation output impact on financial core KPIs. The features illustrated in the monitors are not discernible. This does not provide any description of a nexus between Shapley values and anything.
The Applicant then attempts to rely on the WANDS factors to demonstrate that “nexus,” which is demonstrably absent from the specification, is somehow enabled on page 32 of the response. The examiner will address the most relevant WANDS factors as applied to the claims:
“Nature of invention”: the Applicant asserts the use of SHAP metadata to drive downstream analytics is a straightforward and predictable extension of known techniques, but this is not a demonstration of how to determine the product uniqueness index, a product transferable volume, and/or a product lost volume from Shapley values, the “nexus.” The language the Applicant uses appears to imply that there is no invention at all- that the nature of the claims is a straightforward and predictable extension of known techniques. The Examiner has demonstrated that the techniques are not well-known, especially with regard to the difficult-to-find product uniqueness index. The Applicant has provided no evidence to rebut this, including evidence that rebuts that the product uniqueness index is not well-known, as the Applicant asserts by stating using downstream statistical processes to come up with any sales metric under the sun from Shapley values is conventional knowledge of a person of ordinary skill in the art.
“State of Prior Art”: The Applicant refers to Hoberg to attempt to demonstrate that Hoberg’s HHI/C4 “indices confirm that product similarity and transferability metrics are known and computable. This does not mean there is anything in the prior art about the “nexus” between (I) and (II). The Applicant then states, “comparing vectors for similarity (using cosine similarity, Euclidian distance, or clustering) was well-known at the time of filing, so the art provided abundant tools for implementation.” The comparison of vectors was presented by Applicant with no basis in the specification. Further, these, on their own, would be insufficient to describe and/or enable the “nexus” between (I) and (II).
“Level of predictability in the art”: the Applicant states data science and explainable AI are predictable once the inputs and outputs are defined.” However, this merely relates to the estimation of the Shapley values (I). It has nothing to do with how to determine the product uniqueness index, a product transferable volume, and/or a product lost volume from Shapley values, the “nexus” between (I) and (II). This is not predictable, especially for the esoteric product uniqueness index, the calculation of which is not expressed in terms analogous to the Shapley values, for reasons already stated.
“Direction provided by the inventor”: the Applicant asserts that Paragraphs [0049]-[0051] and FIGs. 3A, 3J, and 3K, demonstrate how to apply weights (e.g., seemingly derived from Shapley values) to compute uniqueness, transferable and lost volume, but there is no example or explanation of how. FIGs. 3A, 3J, and 3K do not so demonstrate.
“Existence of working examples”: the Applicant points to equations 1-5 and FIGs. 3J and 3K. However, equations 1-5 merely demonstrate how to determine Shapley weights ((I) the quantifying step) and FIGs. 3J and 3K do not show any examples of how to determine the product uniqueness index, a product transferable volume, and/or a product lost volume from Shapley values, the “nexus.” This also applies to the nexus between trhe Shapley values and the estimate of the demand transferability.
“Quantity of experimentation” the Applicant has argued, without evidentiary demonstration, that computing similarity indices and allocating volumes in simulations is a conventional application of standard statistical techniques. However, this does not address using Shapley values to determine the product uniqueness index, a product transferable volume, and/or a product lost volume.
The Applicant continues:
Once SHAP weights are available, computing product similarity indices and allocating volumes in simulations is a conventional application of standard statistical and analytics techniques. No undue experimentation is necessary. Accordingly, the record shows that the specification provides both the structural disclosure (system modules, metadata components, and simulation engines) and functional guidance (equations, workflows, and figures) to enable a skilled artisan to make and use the invention without undue experimentation. Per Vasudevan Software, Inc. v. MicroStrategy, Inc., 782 F.3d 671, 682 (Fed. Cir. 2015), detailed source code or one prescribed algorithm is not required where the specification describes the functions and structures in sufficient detail.
The Examiner respectfully disagrees. The Examiner has demonstrated that the Applicant’s specification fails to explicitly state the “nexus” between (I) and (II), and the Applicant has failed to provide evidence to rebut that. The Examiner has demonstrated that the Applicant’s specification has not provided sufficient data for a person of ordinary skill in the art to determine the “nexus” without undue experimentation, and the Applicant has failed to provide evidence to rebut this. The determination of the “nexus” between (I) the Shapley values and (II) the product uniqueness index, product transferable volume, and product lost volume would require the person of ordinary skill in the art to undergo undue experimentation to fill in the gaps that the Applicant’s specification’s failure to address the “nexus” leaves.
For at least these reasons, the Applicant has demonstrated that the Applicant’s specification is not in possession of the claims and has not enabled a person of ordinary skill in the art to make and use the claims without undue experimentation.
35 USC 112(b) Rejections: The Applicant’s arguments and amendments have been considered but are not persuasive. With regard to the “importance of one or more product attributes,” the Applicant appears to state that the importance and the SHAP metadata are equivalent. However, the importance is determined in the retrieve step. The SHAP metadata is determined in the quantify step. It is unclear what role this determined importance plays in the claim because it is never referred to again. This relationship should be clarified or the limitation reciting the importance, removed.
With regard to the overlapping categories, “sales data, product data, product parameters, and financial data.” Contrary to the assertions by the Applicant, FIGs. 3E-3G do not demonstrate a clear distinction between the categories. The Applicant asserts:
PNG
media_image1.png
230
695
media_image1.png
Greyscale
The specification does not state these categorical distinctions. If introduced by amendment to the specification or the claims, this information would be new matter. The actual recitations from the specification show overlapping scope for many if not all of the categories. Here are examples from the Applicant’s specification:
[0005] “an importance of the one or more product attributes being determined based on the product sales data, product data, product parameters, and financial data associated with the product.” (Product sales data v. financial data?)
[0031] “The attribute-based decomposition engine102 using the processor104 may establish for the set of products a relationship between the retrieved one or more product attributes and product sales associated with the product, based on the implementation of non- parametric machine learning (ML) modeling on a data model.” (Product sales is a product attribute under BRI.)
[0032] “In an embodiment, the game theoretic framework may be based on a SHapley Additive exPlanations (SHAP) approach that may process the relationship established by the machine learning (ML) modeling to enable the quantification of the contribution of each product attribute to the product sales. In an embodiment, the one or more product attributes are based on at least one of product sale/price data, product hierarchy, promotion and cost, target consumer, outlet location, product availability, outlet location attributes, competition, product distribution, target audience/market, and product parameters.” (Product parameters could include sale/price data, promotion and cost, and other financial elements, under BRI.)
[0034] “The data ingestion engine 206 may include, data processing/mapping/transformation 218, sales data 208 (Business to Business (B2B), Business to Customer (B2C)), product data 210, financial data 212.” (There is nothing specific here to distinguish between sales and product data.)
The specification does not clearly define or distinguish between sales data, product data, product parameters, and financial data. For example, how is sales data not financial data? MPEP 2173.02(I) states, “A court will not find a patented claim indefinite unless the claim interpreted in light of the specification and the prosecution history fails to ‘inform those skilled in the art about the scope of the invention with reasonable certainty.’ Id. at 899, 110 USPQ2d at 1689.” For the reasons stated, the specification fails to inform those skilled in the art about the scope of the invention with reasonable certainty. As such, a person of ordinary skill in the art would not be able to discern the metes and bounds of the claimed categories individually or as a whole.
With regard to the use of “potential,” potential is a relative term that indicates that a result may or may not happen. Such elements are not given patentable weight. Because these elements are essential to the results of the claim, a person skilled in the art would not be able to determine the metes and bounds of the claims. If the Applicant wishes to attribute a more specific meaning to “potential” or “potentially,” the meaning must be explicit in the claims in terms supported by the specification. The 35 USC 112(b) ejections are maintained.
35 USC 101: With regard to Step 2, Prong 1,The Applicant’s arguments and amendments have been considered but are not persuasive. Contrary to the assertions in the Applicant’s response, the steps identified as evaluations are readily calculable using pen, paper, or a calculator. In reply to the bullets on Mental Processes on the Applicant’s response pages 52-53:
The establish step uses non-parametric ML modeling (e.g., random forest), which can be done with pen, paper, and a calculator. That it is a significant amount of computation is not material to whether the step is practically performable in the mind or with aid of a pen, paper, and/or a calculator. The question of practicality is whether the operations are integrated into computer functionality such that the mind as a processor will not work as a substitute. The Applicant’s assertion that the process would be especially impractical “for millions of product-attribute combinations across multiple markets is also immaterial because those millions of product-attribute combinations across multiple markets are not claimed.
The quantify step employs Shapley, which can also be done on pen and paper. The calculations are also able to be performed with pen paper and a calculator. Again, the quantity and the difficulty of the calculations are immaterial to the practicality of the performance in the mind or with aid of pen, paper, and a calculator. Aside from the burden of the calculations, which is largely immaterial, the Applicant has not demonstrated that these calculations, like and calculations, cannot be practically performed in the mind or with the aid of pen, paper, and/or a calculator.
The estimate step calculates demand transferability. The Applicant states, “[t]his requires vast datasets and statistical modeling, far beyond human mental capacity. Again, the practicality element is not addressed to the difficulty and magnitude of calculations. There is no reason, if a person skilled were sufficiently informed by the specification as to how the calculations are done, that a person skilled in the art could not do the calculations mentally or with the aid of pen, paper, and/or a calculator.
The Applicant also mentions that the Applicant’s claimed recitation of automation by computer means that the elements cannot practically be performed in the mind. However, claiming automation or describing it in the specification does not make the tasks automated any less practically performable in the mind or with aid of pen, paper, and/or a calculator. The independent claims unequivocally recite mental processes.
With regard to Mathematical concepts, the Applicant states that amended claim 1 is not abstract mathematical concepts in isolation but are applied with a specific technological framework to solve a technical problem, citing Enfish. The Applicant then states a practical purpose of the Shap determinations. The Applicant fails to understand that the claim is not rejected as abstract but that the claim is rejected for reciting a mathematical concept, an abstract idea, at Step 2A, Prong 1. Whether additional limitations confer eligibility at Step 2A, Prong 2 or Step 2B is a different inquiry. The Applicant appears to concede that the Shapley determinations, themselves, are mathematical concepts. The Examiner agrees.
With regard to fundamental economic practices, the Applicant has seemingly contradicted the statements made with regard to the 35 USC 112(a) rejection. Here, the Applicant states, “The amended claim 1 does not merely select products to sell but provides a sophisticated, automated system for quantifying attribute contributions and estimating demand transferability to optimize product portfolios. Unlike Hoberg et al., which focuses on product similarity metrics (pages 20-22), amended claim 1 integrates ML and SHAP to produce fine-grained, product-specific insights (e.g., SHAP metadata, product uniqueness index) for dynamic portfolio management across markets (paragraph [0025]). This is not a longstanding economic practice but a novel, data-driven approach to retail analytics, akin to the technological improvements in McRO, Inc. v. Bandai Namco Games America Inc., 837 F.3d 1299 (Fed. Cir. 2016).” In response to the 35 USC 112(a) rejection, the Applicant stated on page 14 of the response, “[a] skilled artisan in retail analytics would readily recognize that applying quantified attribute weights to estimate demand transferability involves comparing attribute-importance profiles across products and simulating redistribution of volume, which is a routine practice in demand forecasting and portfolio analytics.” Page 15: “A person of ordinary skill in retail analytics would readily understand that SHAP vectors can be compared using standard similarity techniques (e.g., cosine similarity, Euclidean distance, clustering), which are routine in the art, to produce uniqueness indices and allocate transferable versus lost volume.” Page 20: “Once SHAP weights are available, computing product similarity indices and allocating volumes in simulations is a conventional application of standard statistical and analytics techniques.” The Applicant appears to be confused as to whether application of Shapley values to product uniqueness or volume determinations is something conventional or is something extraordinarily inventive. As a concept, using measures of product feature importance to determine likely product sales is a fundamental economic practice. As implemented in the specification, while this is a specifically and incompletely demonstrated method, the underlying economic principle is still a fundamental economic practice.
With regard to Step 2A, Prong 2, the amendments presented do not provide any additional limitations that would integrate the abstract idea into a practical application at Step 2A, Prong 2. The inquiry at Step 2A, Prong 2 is whether the claim recites “additional limitations” that integrate the judicial exception/abstract idea into a practical application. The Applicant claims that the following allegedly claimed features integrate the abstract idea into a practical application.
Non-parametric ML- this is a generic computing structure being used for its ordinary inferential purpose and which is also an element of the abstract idea. Even if it were an additional limitation, its generic computing nature would not confer eligibility under MPEP 2106.05(f).
SHAP Framework and metadata – This is also a generic computing element that is being used for its ordinary purpose and is also an element of the abstract idea. Even if it were an additional limitation, its generic computing nature would not confer eligibility under MPEP 2106.05(f).
Azure infrastructure which leverages… NOT CLAIMED – Only elements of the claim can integrate the abstract idea into a practical application.
Data Vault method… NOT CLAIMED – Only elements of the claim can integrate the abstract idea into a practical application.
The claims do not provide any additional limitations, apart from the abstract idea, that would integrate the abstract idea/judicial exception into a practical application, so the claims are directed to the abstract idea/judicial exception at Step 2A, Prong 2. Contrary to the assertions in the response on pages 47-48, the computer itself is not improved by the claim. The claim is merely running an abstract idea that, according to the Applicant, provides a good result. According to the Applicant, the claims recite “an attribute decomposition engine (which they no longer recite), Shap metadata storing attribute importance values across attribute arrays (which merely limit the abstract idea/judicial exception to a particular field of use and, under MPEP 2106.05(h), does not confer eligibility), and demand transferability metrics (which are determined as elements of the abstract idea. The Applicant tries to rebut that the retrieve step is insignificant extra-solution activity by referring to Shapley data that is determined based on that retrieved data, which does not make sense. Retrieving data for subsequent determinations is mere data gathering and is clearly treated as such in MPEP 2106.05(g). Also, contrary to the assertions by the Applicant, and despite conventionality not being a consideration at Step 2A, Prong 2, the arrangement of elements, receiving data, processing data to determine an intermediate, and then processing the intermediate to come to a result, is conventional and similar to Example 47, Claim 2 of the PEG examples. Accordingly, the claims do not integrate the judicial exception/abstract idea into a practical application and are directed to the judicial exception/abstract idea.
With regard Step 2B, the Applicant states,
The amended independent claim 1 recites a unique combination of a "of an attribute-based decomposition engine, non-parametric machine learning (ML) models such as Random Forest, a SHapley Additive exPlanations (SHAP) game-theoretic framework, and a scalable Azure cloud infrastructure with a Data Vault methodology for unified data modeling. This combination is not well-understood, routine, or conventional, as it addresses the specific problem of optimizing Price Pack Architecture (PPA) and product assortment in retail and manufacturing through a tailored machine-learning architecture that quantifies product attribute contributions and estimates demand transferability across millions of product-attribute combinations, producing precise, actionable insights via SHAP metadata and demand transferability metrics (e.g., product uniqueness index, transferable volume, lost volume).
First of all, some of these elements are not claimed (scalable Azure cloud infrastructure and Data Vault methodology). Further, the elements the Applicant claims to be additional limitations are elements of the abstract idea. Even if, arguendo, the elements were not, the elements are generic computing elements performing their routine computing functions. Shapley with random forest determinations are conventionally used to assess the value of particular features/attributes of a product. This means that, even if they were additional limitations, they would not confer eligibility under MPEP 2106.05(f) and 2106.05(d).
The Applicant continues to try and emphasize the collection of generic data used for PPA determinations is other than mere data gathering, but it is clearly analogous to the examples presented in MPEP 2106.05(g). Also, the Applicant asserts that the product uniqueness index (the only calculation for which the Examiner found was from the Hoberg reference) is unconventional, but it is clearly a mathematical concept, an element of the abstract idea that cannot be considered an additional limitation, because the determination of the product uniqueness index is purely mathematical calculation and nothing more.
For at least these reasons, the Applicant’s claims fail to recite any additional limitations that combine with the other elements of the claim to provide significantly more than the judicial exception/abstract idea that would render the claims an inventive concept at Step 2B. None of the other amendments to the dependent claims remedy this. For example, with respect to claims 1, 10, and 19, the first new “wherein clause” is nothing more than a generic data storing step for storing metadata (MPEP 2106.05(d) and 2106.05(g)) with a description of the data that merely limits the abstract idea to a particular field (MPEP 2106.05(h)). The second wherein clause merely describes the nature of output parameters of the estimate step, which becomes an element of the abstract idea.
35 USC 103: The arguments and amendments have been considered and are persuasive. The previous rejection has been withdrawn, in light of the new amendments, and a new rejection, necessitated by the amendments, has been issued. The new amendments required remapping of the elements to the existing references and a new Channer reference. The rejection, while different still relies on a combination of Wang and Antipov. Contrary to the assertions in the office action, the motivation to combine Wang with Antipov is clearly stated. Wang attempts to determine how much uniqueness of features contributes to sales volume. Antipov provides a metric by which to quantify the importance of each product feature to sales (volume). Please see the complete motivation statement with the rejections below.
Claim Rejections - 35 USC § 112
35 USC 112(a)
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
Written Description
Claims 1-20 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claims 1, 10, and 19 recite, “quantify a contribution of each product attribute on the product sales, based on the established relationship and a game theoretic framework, wherein the contribution of each product attribute on the product sales is quantified to determine weights of the respective product attributes; and estimate demand transferability among the set of products based on the determined weights of the respective product attributes.”
The Applicant’s specification explains in paragraphs [0052]-[0053] and Equations 1-5 output a quantification of relative values of input product attributes. This appears to be an explanation of how to conduct the quantify step. Paragraph [0036] states,
The ML modeling metadata 234 may be an automated model hyper parameter tuning log of automated experiment iterations champion model validation and performance metrics, the SHAP metadata 236 may be an attribute importance for all attribute arrays across all products, and the demand transferability 238 may be a product uniqueness index, a product transferable volume, and a product lost volume.
Product uniqueness index is a metric demonstrated in a 2010 paper by Gerard Hoberg and Gordon Philips (Hoberg et al. on the record). The methods of Hoberg et al. use HHI and C4 indices to determine similarities. The Applicant’s specification provides no guidance for using the Shapley values in this calculation. For at least this reason, a person of ordinary skill in the art would interpret the specification as failing to sufficiently conceive the inventive concept, and, therefore, would determine that the Applicant lacks possession of the invention.
The dependent claims dependent from the rejected claims are rejected based at least on their dependence from the rejected claims.
Enablement
Claims 1-20 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the enablement requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to enable one skilled in the art to which it pertains, or with which it is most nearly connected, to make and/or use the invention.
MPEP 2164.04(a): In order to determine compliance with the enablement requirement of 35 U.S.C. 112(a), the Federal Circuit developed a framework of factors in In re Wands, 858 F.2d 731, 737, 8 USPQ2d 1400, 1404 (Fed. Cir. 1988), referred to as the Wands factors to assess whether any necessary experimentation required by the specification is "reasonable" or is "undue." Consistent with Amgen Inc. et al. v. Sanofi et al., 598 U.S. 594, 2023 USPQ2d 602 (2023), the Wands factors continue to provide a framework for assessing enablement in a utility application or patent, regardless of technology area. See Guidelines for Assessing Enablement in Utility Applications and Patents in View of the Supreme Court Decision in Amgen Inc. et al. v. Sanofi et al., 89 FR 1563 (January 10, 2024). These factors include, but are not limited to: (A) The breadth of the claims; (B) The nature of the invention; (C) The state of the prior art; (D) The level of one of ordinary skill; (E) The level of predictability in the art; (F) The amount of direction provided by the inventor; (G) The existence of working examples; and (H) The quantity of experimentation needed to make or use the invention based on the content of the disclosure.
Claims 1, 10, and 19 recite, “quantify a contribution of each product attribute on the product sales, based on the established relationship and a game theoretic framework, wherein the contribution of each product attribute on the product sales is quantified to determine weights of the respective product attributes; and estimate demand transferability among the set of products based on the determined weights of the respective product attributes.”
The Applicant’s specification explains in paragraphs [0052]-[0053] and Equations 1-5 output a quantification of relative values of input product attributes. This appears to be an explanation of how to conduct the quantify step. Paragraph [0036] states,
The ML modeling metadata 234 may be an automated model hyper parameter tuning log of automated experiment iterations champion model validation and performance metrics, the SHAP metadata 236 may be an attribute importance for all attribute arrays across all products, and the demand transferability 238 may be a product uniqueness index, a product transferable volume, and a product lost volume.
Product uniqueness index is a metric demonstrated in a 2010 paper by Gerard Hoberg and Gordon Philips (Hoberg et al. on the record). The methods of Hoberg et al. use HHI and C4 indices to determine similarities. The Applicant’s specification provides no guidance for using the Shapley values in this calculation. Product transferable volume and product lost volume are recognized terms of art and have no explanation in the specification.
With respect to the most relevant Wands factors: […] (C) The state of the prior art – The product uniqueness index is a seldom-used metric that is not particularly well-known. Few applications have been developed for the metric. […] (E) The level of predictability in the art - The product uniqueness index is a seldom-used metric that is not particularly well-known. Few applications have been developed for the metric. (F) The amount of direction provided by the inventor – The inventor has provided no direction as to how to use the outputs of the Linear SHAP/ random forest as inputs to or otherwise guidance for the determination of the product uniqueness index or any other metric of uniqueness. (G) The existence of working examples – The specification provides no working examples. (H) The quantity of experimentation needed to make or use the invention based on the content of the disclosure – There is a significant gap in the instruction of how outputs from the Linear SHAP/random forest methods would be used to determine any of a product uniqueness index, a product transferable volume, and a product lost volume. Determining how to bridge this gap would require a great deal of experimentation, especially given how little guidance there is in the art and specification of how to use the product uniqueness index, and how there is no guidance as to what a product transferable volume and a product lost volume even are.
For at least these reasons, a person of ordinary skill in the art would interpret the specification as failing to enable the person to make and use the claimed limitations without undue experimentation.
The dependent claims dependent from the rejected claims are rejected based at least on their dependence from the rejected claims.
35 USC 112(b)
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1-20 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
An importance of the one or more product attributes
Independent claims 1, 10, and 19 recite “an importance of the one or more product attributes being determined based on product sales data, product data, product parameters, and financial data associated with the product.” However, the determination of the importance of the product attributes is not positively claimed (e.g., claimed as being something executed by the engine). Further, this importance is never recited again in the independent or respective dependent claims. Further, based on the specification, it is unclear whether this importance is merely an explanation of the later “quantify a contribution of each product attribute on the product sales” or is something different, making the relationship between the elements entirely unclear. Accordingly, a person of ordinary skill in the art would not be able to determine the metes and bounds of the claims based on this free-floating recitation of a non-positively recited importance element.
In the interest of compact prosecution, for purposes of examination, the recitation of the importance will be considered a preliminary representation of a result to be achieved with respect to the quantify step. Despite this, the Applicant is required to bring the claim into compliance with 35 USC 112(b) without adding new matter to overcome this rejection.
Sales data, product data, product parameters, and financial data
Independent claims 1, 10, and 19 recite “an importance of the one or more product attributes being determined based on product sales data, product data, product parameters, and financial data associated with the product.” The terms of the claim require that all of “sales data,” “product data,” “product parameters,” and “financial data” be the basis for the claimed determination of the importance of the “product attributes.” However, it is unclear what the distinction is between the terms “sales data,” “product data,” “product parameters,” and “financial data” associated with the product. For example, the broadest reasonable interpretation of sales data can include product data, product parameters, and financial data. Clearly, sales data is financial data. Further, it is unclear what the relationship between the claimed “product attributes” and the claimed “product sales data, product data, product parameters, and financial data.” The specification does not help distinguish between these elements. The specification provides no guidance with examples of each of the claimed “product attributes,” “sales data,” “product data,” “product parameters,” and “financial data.” Because the specification fails to provide examples of, fails to distinguished between, and fails to show the relationships between the claimed “product attributes,” “sales data,” “product data,” “product parameters,” and “financial data,” a person of ordinary skill in the art would not understand the metes and bounds of the claims and would find the claims to be indefinite.
In the interest of compact prosecution, the claim will be interpreted to recite that sales data is the basis for determining importance of product attributes in a purchase decision. Despite this, the Applicant is required to bring the claim into compliance with 35 USC 112(b) without adding new matter to overcome this rejection.
Potential and potentially
Claims 3 and 12 recite “transfer a quantification of potential volume to other products or potentially lost from a given product category in instances of reduction, change, or increase in product portfolio.” It is unclear what this clause means. The specification provides no clarification. Further, this transfer process is not related to any other operation of the claims from which it depends. Also, the use of the words “potential” or “potentially” renders the scope of the claim unclear.
In the interest of compact prosecution, for the purpose of examination, the clause will be interpreted to mean handling information regarding changing inventory of a product. Despite this, the Applicant is required to bring the claim into compliance with 35 USC 112(b) without adding new matter to overcome this rejection.
One or more product attributes
Claims 9 and 18 recite,
wherein the one or more product attributes are based on at least one of product sale/price data, product hierarchy, promotion and cost, target consumer, outlet location, product availability, outlet location attributes, competition, product distribution, target audience/market, and product parameters.
Again, it is unclear what the relationship is between the claimed “an importance of the one or more product attributes being determined based on product sales data, product data, product parameters, and financial data associated with the product” and the claimed “one or more product attributes are based on at least