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 .
Response to Arguments
Examiner’s Note: Examiner notes that claims 11 and 12 have a status indicator of “Original” even though the claims are amended. For the sake of compact prosecution, Examiner is interpreting the claims as though they recite “Currently Amended.”
Regarding the 35 USC 112(f) Claim Interpretation, Examiner has fully considered Applicant’s arguments and amendments. The amended claims no longer invoke 35 USC 112(f). Therefore, the claim interpretation has been withdrawn.
Regarding the 35 USC 112(b) rejection, Examiner has fully considered Applicant’s arguments and amendments. The claim amendments are sufficient to overcome the 35 USC 112(b) rejection. Accordingly, the 35 USC 112(b) rejection is withdrawn.
Regarding the 35 USC 101 rejection, Examiner has fully considered Applicant’s arguments and amendments.
Regarding Applicant’s assertion of “The Applicant respectfully submits that one or more features of amended independent claim 1 cannot be performed/executed by human mind. These steps are inextricably tied to a machine which requires computational processing performed interoperably by electronic components such as an apparatus with at least a circuitry, processor, and a memory as described at least at paragraphs [0022], [0033] of originally filed Specification. The Applicant submits that the subject matter of independent claim 1 is not directed towards an abstract idea and is tied to a machine which executes and performs say, the one or more features recited in amended independent claim 1. That is, especially the steps of retrieving, by the circuitry, a demand curve data associated with the second identifier of the second product from a database, wherein the demand curve data is indicative of product consumption quantities with respect to the varying unit product values of the plurality of products; computing, by the circuitry, a slope from the demand curve data of the second product, wherein the slope indicates variation in the product consumption quantities with respect to a product value of the second product; based on the slope, calculating, by the circuitry, an optimal range of a unit product value for the first product; and storing, by the circuitry, the optimal range of the unit product value for the first product in the database performed by the apparatus are not merely directed to abstract mental steps but describe specific computer operations which involve analyzing potentially vast amount of data associated with the plurality of products in real-time.” Examiner respectfully disagrees. The limitations of computing and calculating, cited above, but for the limitation of “by the circuitry,” are part of the abstract limitations for consideration under Step 2A, Prong 1. With respect to the limitation of retrieving, but for the limitations of “by the circuitry” and “from a database,” as drafted, is a part of the abstract limitations for consideration under Step 2A, Prong 1. These limitations, as drafted, is a process that, under its broadest reasonable interpretation, but for the language of “by circuitry,” covers an abstract idea but for the recitation of generic computer components. That is, other than reciting “by circuitry,” nothing in the claim elements preclude the steps from being interpreted as an abstract idea. For example, with the exception of the “by circuitry,” language, the claim steps in the context of the claim encompass an abstract idea directed to “Certain Methods of Organizing Human Activity.” The circuitry and database of the claim, even when considered in combination with the limitation of “storing” and the other additional elements of the claims, are not sufficient to prove integration into a practical application or anything significantly more.
Regarding Applicant’s assertion of “In particular, the use of identifiers comprising numerical vectors generated using natural language processing is not merely "organizing information," but instead requires conversion of text strings into numeric feature-space representations that enable computational similarity evaluation, and such vector representations do not exist in the human mind in any meaningful or practical way and cannot be manipulated mentally in the manner required by the claim.” Examiner respectfully asserts that this limitation is part of the additional elements for consideration under Step 2A, Prong 2 and Step 2B. Therefore, this limitation is not part of the abstract limitations for consideration under Step 2A, Prong 1.
Regarding Applicant’s assertion of “Likewise, the "closest identifier" selection is not a subjective or conceptual comparison but a quantitative computation over numerical identifier space to locate a matching product instance having sufficient data, which is fundamentally different from a human selecting "similar products" by intuition. Further, deriving a slope from demand curve data retrieved from a database is not a generic calculation divorced from implementation. This requires a computing system to retrieve stored multi-point consumption/value datasets and compute a model parameter from those datasets, which again cannot reasonably be performed as a mental step for an arbitrary number of products or datapoints. Finally, requiring that the computed optimal range be stored in the database integrates the output into a computer's data layer as a machine-usable artifact, confirming that the claim is directed to a specific computerized data-processing pipeline that produces a concrete system result rather than an abstract idea. Accordingly, the focus of the claim is on a particular technical procedure for representing product name information as vectors, performing computational matching and model-based range calculation using database-derived demand data, and persistently recording the result in system storage features that are inherently technological and not a mental process. See at least at paragraphs [0031]-[0047, [0052)-[0064] of originally filed Specification.,” Examiner respectfully disagrees with Applicant’s assertion. The identification of the closest identifier, as drafted, under considerations of the broadest reasonable interpretation of the claim, is part of the abstract limitations under consideration of Step 2A, Prong 1. This limitation of the claim, under considerations of the broadest reasonable interpretation, does not comprise or recite any additional elements for consideration under Step 2A, Prong 2 or Step 2B. Therefore, this limitation, as drafted, cannot integrate the judicial exception into a practical application because it is part of the abstract limitations for consideration under Step 2A, Prong 1.
Regarding Applicant’s assertion of “For instance, the subject matter of independent claim 1 can be practically realized in for instance, in network environment where computing hardware like system servers or workstations with processor(s), memory, interfaces, and enterprise networking hosting a cluster generation engine and a computation engine that are coupled to the organization's database over the network. Product names and product-consumption datapoints are read from the database, the cluster generation engine executes the disclosed pipeline lowercasing, stopword removal, and vectorization using tools such as sentence transformers, to convert product names into numerical identifiers. The engine then determines an optimal number of clusters via the elbow method and applies K-means to group products with similar names. Within each cluster, the computation engine identifies sparse-data products fewer than three points and pairs each with the closest sufficient-data product using the distance-based selection. It then retrieves the matched product's demand-curve data from the facility database, computes the slope, and calculates an optimal unit-value range for the sparse-data product. The resulting price ranges are stored back in the database for downstream use by in-facility pricing systems and dashboards, with all read/write and orchestration occurring on the local servers and networked environment (See at least at paragraphs [0022]-[0024], [0031]-[0033], [0043]- [0045], [0062]-[0064], [0066]-[0069]). This accounts for practical implementation.,” Examiner respectfully disagrees. Examiner primarily respectfully asserts that the present claims, under considerations of broadest reasonable interpretation, do not specifically recite a network environment, a cluster generation engine, enterprise networking, an “elbow method,” or other asserted additional elements. Applicant is reminded that it is impermissible to import limitations from the specification into the claims. (See MPEP 2111.01(II)) In order for Applicant to be given the argued interpretation, the claims must positively recite the argued subject matter.
The argued limitations above recite several abstract limitations (i.e. computing) for consideration. These abstract limitations are not sufficient to prove integration into a practical application or anything significantly more because these claims do not recite any additional elements for consideration. The combination of additional elements positively recited in the claim, such as the database, circuitry, and other additional elements, when considered in the context of the claims, are not sufficient to prove integration into a practical application or anything significantly more. These additional elements are mere instructions to implement an abstract idea using a computer in its ordinary capacity, or merely uses the computer as a tool to perform the identified abstract idea. See MPEP 2106.05(f). Use of a computer or other machinery in its ordinary capacity for tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., certain methods of organizing human activity) does not integrate a judicial exception into a practical application or anything significantly more. See MPEP 2106.05(f).
Regarding Applicant’s assertion of “In view of such challenges, the subject matter of amended independent claim 1 proposes a method that applies natural language processing to transform product names into numerical identifiers that capture semantic similarity, clusters these identifiers using an optimal cluster count, and identifies within each cluster a product having sufficient consumption data that is closest by vector distance and optional unit-value variation constraints to a product with insufficient data. The method then retrieves the demand-curve data of the sufficient-data product, computes the slope to characterize how consumption varies with price, and uses this slope to algorithmically derive an optimal range of unit product values for the plurality of products. The computed optimal range is stored in a database for downstream pricing workflows, thereby enabling accurate price optimization even when direct consumption data is inadequate. This technical solution yields practical advantage by enabling accurate computation of an optimal unit product value even when a product lacks sufficient consumption data, a task that traditional demand-curve-based pricing systems cannot perform. By transforming product names into numerical identifiers using NLP and clustering them to identify semantically similar items, the method efficiently reuses demand-curve intelligence from related products with adequate data. This reduces computational overhead, eliminates the need for additional data collection, and ensures consistent and scalable pricing optimization across large product catalogs. Furthermore, by automating distance-based pairing, slope extraction, and optimal-range calculation, the method improves the precision and reliability of pricing recommendations, enhances revenue-generation capability, and supports continuous, data-driven decision-making within enterprise pricing environments. See at least at paragraphs [0017]- [0019], [0022]-[0029], [0052]-[0069] of originally filed Specification.,” Examiner respectfully disagrees. Examiner respectfully asserts that an improvement to calculating a price range, as drafted, would be an improvement to the abstract limitations for consideration under Step 2A, Prong 1. The additional elements of the claim, such as the natural language processing techniques, are not improved by the claim. 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”
Accordingly, the 35 USC 101 rejection is maintained.
Regarding the 35 USC 103 rejection, Examiner has fully considered Applicant’s arguments and amendments.
Regarding Applicant’s assertion of “Further, even if Sawarkar and Wellmann are combined, the combination still does not render the amended claim obvious,” Applicant’s arguments with respect to the previous prior art combination of the record have been considered but are moot because the new grounds of rejection does not rely on any reference applied in the prior art rejection for any teachings or matter specifically challenged in the argument. The claims are rejected under a new grounds of rejection, which was necessitated by amendment. Examiner has introduced the Rovatsos reference to cure the deficiencies of the prior art combination of the record.
Accordingly, the 35 USC 103 rejection is maintained.
Claim Objections
Claim 13 is objected to because of the following informalities: Examiner suggests amending the claim for the sake of antecedence by reciting “wherein to determine the distance, [[the]]a computing engine is to compute a root mean square distance between the first identifier and the second identifier.”
Appropriate correction is required.
Claim Rejections - 35 USC § 112
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 16-20 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.
Regarding claim 16, the metes and bounds of the claim are rendered unclear due to the combination of limitations of “A non-transitory computer-readable medium having instructions stored thereon, the instructions, when executed by a processor, cause the processor to perform operations comprising” and “retrieving, by the circuitry, a demand curve data associated with the second identifier of the second product from a database, wherein the demand curve data is indicative of product consumption quantities with respect to the varying unit product values of the plurality of products; computing, by the circuitry, a slope from the demand curve data of the second product.” The metes and bounds of the claim are rendered unclear because the preamble initially recites a non-transitory computer readable medium that is executed by a processor that performs the operations of the claim. However, the body of the claim recites limitations performed by “the circuitry.” There is no antecedent basis for the claimed “the circuitry.” Furthermore, it is unclear if the claim is performed by the “processor” of the preamble, or if the claim is performed by “the circuitry.” Therefore, the metes and bounds of the claim are rendered unclear because it is unclear what structure is performing the steps of the claim.
For the sake of compact prosecution, Examiner is interpreting the claim as reciting “by the processor.” Examiner suggests amending the claim to clarify what structure is performing the claimed functions.
Dependent claims 17-20 are rejected due to dependency on rejected base claim 16.
Accordingly, claims 16-20 are rejected under 35 USC 112(b).
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 1-20 are rejected under 35 USC 101 because the claimed invention is directed to a judicial exception (i.e. abstract idea) without anything significantly more.
Step 1: Claims 1-7 are directed to a method, claims 8-15 are directed to a system, and claims 16-20 are directed to a non-transitory computer readable medium. Therefore, the claims are directed to patent eligible categories of invention.
Step 2A, Prong 1: Claims 1, 8, and 16 recite computing an optimal range of a unit product value for the first product, constituting an abstract idea based on “Certain Methods of Organizing Human Activity” related to commercial interactions including advertising or marketing sales activities or behaviors. Independent claim 1 recites limitations including “generating, two or more clusters for a plurality of products, each of the two or more clusters having a respective set of products with similar names, the plurality of products having product consumption data with one or more data points indicative of varying product consumption quantities with respect to varying unit product values for a respective product, wherein the generation of the two or more clusters comprises: transforming names of the plurality of products into identifiers with numerical vectors, and clustering the identifiers into two or more clusters based on similarity in names of the plurality of the plurality of products; within each cluster, for a first product having product consumption data with less than three data points, determining, a second product that has product consumption data with more than three data points, wherein the determination of the second product comprises identification of a second identifier associated with the second product that is closest to a first identifier associated with the first product; retrieving, a demand curve data associated with the second identifier of the second product, wherein the demand curve data is indicative of product consumption quantities with respect to the varying unit product values of the plurality of products: computing, a slope from the demand curve data of the second product, wherein the slope indicates variation in the product consumption quantities with respect to a product value of the second product based on the slope, calculating, an optimal range of a unit product value for the first product.” Independent claim 8 recites limitations including “determine a plurality of identifiers corresponding to names of a plurality of products, the plurality of products having product consumption data with one or more data points indicative of varying product consumption quantities with respect to varying unit product values for a respective product; generate two or more clusters for the plurality of identifiers, each of the two or more clusters having a respective set of identifiers from amongst the plurality of identifiers, wherein the generation of the two or more clusters comprises transforming names of the plurality of products into set of identifiers with numerical vectors, the set of identifiers relating to products with similar names; within each of the two or more clusters, for a first identifier associated with product consumption data having less than three data points, determine a second identifier associated with product consumption data having more than three data points, wherein the determination is based at least in part on a closest distance between the first identifier and the second identifier; retrieve a demand curve data associated with the second identifier of a second product, wherein the demand curve data is indicative of product consumption quantities with respect to the varying unit product values; compute a slope from the demand curve data of the second product, wherein the slope indicates variation in the product consumption quantities with respect to a product value of the second product; based on the slope, calculate an optimal range of a unit product value for a product associated with the first identifier.” Independent claim 16 recites limitations including “generating identifiers corresponding to names of a plurality of products, wherein the plurality of products having product consumption data with one or more data points indicative of varying product consumption quantities with respect to varying unit product values for a respective product, wherein the generation of the identifiers comprises transforming the names of the plurality of products into set of identifiers with numerical vectors; clustering the identifiers into an optimal number of clusters, the clustering is based on a similarity in the names of the plurality of products; within each cluster, determining a pair of identifiers closest to each other and having a variation in product values within a pre-defined limit, a first identifier from the pair of identifiers relates to a product having product consumption data with less than three data points and a second identifier from the pair of identifiers relates to a product having product consumption data with more than three data points; retrieving, a demand curve data associated with the second identifier of the second product, wherein the demand curve data is indicative of product consumption quantities with respect to the varying unit product values of the plurality of products; computing, a slope from the demand curve data of the second product, wherein the slope indicates variation in the product consumption quantities with respect to a product value of the second product; based on the slope, calculating, an optimal range of a unit product value for the product associated with the first identifier.” These limitations, as drafted, is a process that, under its broadest reasonable interpretation, but for the language of “by circuitry,” covers an abstract idea but for the recitation of generic computer components. That is, other than reciting “by circuitry,” nothing in the claim elements preclude the steps from being interpreted as an abstract idea. For example, with the exception of the “by circuitry,” language, the claim steps in the context of the claim encompass an abstract idea directed to “Certain Methods of Organizing Human Activity.”
Dependent claims 2, 4-7, 9, 11-15, and 17-20 further narrow the abstract idea identified in the independent claims and do not introduce further additional elements for consideration.
Dependent claims 3 and 10 will be evaluated under Step 2A, Prong 2 below.
Step 2A, Prong 2: Independent claims 1, 8, and 16 do not integrate the judicial exception into a practical application. Independent claim 1 is directed to a method performed “by a circuitry.” Claim 8 is a system comprising “a processor” and “circuitry.” Claim 16 recites “a non-transitory computer-readable medium having instructions stored thereon, the instructions, when executed by a processor, cause the processor to perform operations comprising” within the preamble of the claim, as well as limitations performed “by circuitry.” Claim 1 further recites the additional element, similarly recited in claims 8 and 16, including “retrieving, by the circuitry, a demand curve data associated with the second identifier of the second product from a database.” Claim 1 further recites the additional element, similarly recited in claims 8 and 16, including “storing, by the circuitry, the optimal range of the unit product value for the first product in the database.” These additional elements are mere instructions to implement an abstract idea using a computer in its ordinary capacity, or merely uses the computer as a tool to perform the identified abstract idea. See MPEP 2106.05(f). Use of a computer or other machinery in its ordinary capacity for tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., certain methods of organizing human activity) does not integrate a judicial exception into a practical application. See MPEP 2106.05(f).
Claim 1 further recites the additional element of “transforming names of the plurality of products into identifiers with numerical vectors using natural language processing techniques.” Claim 8 further recites the additional element of “wherein the generation of the two or more clusters comprises transforming names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques.” Claim 16 further recites the additional element of “wherein the generation of the identifiers comprises transforming the names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques.” These limitations, as drafted, under considerations of the broadest reasonable interpretation, are nothing more than generally linking the use of the judicial exception to a particular technical field. The claims employ generic computer functions to execute an abstract idea, even when limiting the use of the idea to one particular environment. This type of generally linking is not sufficient to prove integration into a practical application. See MPEP 2106.05(h).
Therefore, the additional elements of the independent claims, when considered both individually and in combination, are not sufficient to prove integration into a practical application.
Dependent claims 2, 4-7, 9, 11-15, and 17-20 further narrow the abstract idea identified in the independent claims and do not introduce further additional elements for consideration, which does not prove integration into a practical application.
Dependent claim 3 recites the additional element of “wherein the identifiers are determined using the natural language processing techniques.” Dependent claim 10 recites the additional element of “wherein the circuitry uses the natural language processing techniques to generate the plurality of identifiers.” These limitations, as drafted, under considerations of the broadest reasonable interpretation, are nothing more than generally linking the use of the judicial exception to a particular technical field. The claims employ generic computer functions to execute an abstract idea, even when limiting the use of the idea to one particular environment. This type of generally linking is not sufficient to prove integration into a practical application. See MPEP 2106.05(h).
Therefore, the additional elements of the dependent claims, when considered both individually and in combination, are not sufficient to prove integration into a practical application.
Step 2B: Independent claims 1, 8, and 16 do not comprise anything significantly more than the judicial exception. Independent claim 1 is directed to a method performed “by a circuitry.” Claim 8 is a system comprising “a processor” and “circuitry.” Claim 16 recites “a non-transitory computer-readable medium having instructions stored thereon, the instructions, when executed by a processor, cause the processor to perform operations comprising” within the preamble of the claim, as well as limitations performed “by circuitry.” Claim 1 further recites the additional element, similarly recited in claims 8 and 16, including “retrieving, by the circuitry, a demand curve data associated with the second identifier of the second product from a database.” Claim 1 further recites the additional element, similarly recited in claims 8 and 16, including “storing, by the circuitry, the optimal range of the unit product value for the first product in the database.” These additional elements are mere instructions to implement an abstract idea using a computer in its ordinary capacity, or merely uses the computer as a tool to perform the identified abstract idea. See MPEP 2106.05(f). Use of a computer or other machinery in its ordinary capacity for tasks (e.g., to receive, store, or transmit data) or simply adding a general purpose computer or computer components after the fact to an abstract idea (e.g., certain methods of organizing human activity) is not anything significantly more than the judicial exception. See MPEP 2106.05(f).
Claim 1 further recites the additional element of “transforming names of the plurality of products into identifiers with numerical vectors using natural language processing techniques.” Claim 8 further recites the additional element of “wherein the generation of the two or more clusters comprises transforming names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques.” Claim 16 further recites the additional element of “wherein the generation of the identifiers comprises transforming the names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques.” These limitations, as drafted, under considerations of the broadest reasonable interpretation, are nothing more than generally linking the use of the judicial exception to a particular technical field. The claims employ generic computer functions to execute an abstract idea, even when limiting the use of the idea to one particular environment. This type of generally linking is not sufficient to prove anything significantly more than the judicial exception. See MPEP 2106.05(h).
Therefore, the additional elements of the independent claims, when considered both individually and in combination, are not anything significantly more than the judicial exception.
Dependent claims 2, 4-7, 9, 11-15, and 17-20 further narrow the abstract idea identified in the independent claims and do not introduce further additional elements for consideration, which is not anything significantly more than the judicial exception.
Dependent claim 3 recites the additional element of “wherein the identifiers are determined using the natural language processing techniques.” Dependent claim 10 recites the additional element of “wherein the circuitry uses the natural language processing techniques to generate the plurality of identifiers.” These limitations, as drafted, under considerations of the broadest reasonable interpretation, are nothing more than generally linking the use of the judicial exception to a particular technical field. The claims employ generic computer functions to execute an abstract idea, even when limiting the use of the idea to one particular environment. This type of generally linking is not sufficient to prove anything significantly more than the judicial exception. See MPEP 2106.05(h).
Therefore, the additional elements of the dependent claims, when considered both individually and in combination, are not anything significantly more than the judicial exception.
Accordingly, claims 1-20 are rejected under 35 USC 101.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (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.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
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.
Claim(s) 1-5, 7-12, 14-16, and 18-20 are rejected under 35 U.S.C. 103 as being unpatentable over Sawarkar et al. (US 20220138786 A1) in view of Wellmann et al. (US 20240221009 A1) in view of Rovatsos et al. (US 20240354810 A1).
Regarding claim 1, Sawarkar teaches a method comprising (Fig. 4):
generating, by a circuitry (Fig. 12 and [0127-0130] teach a computer system/server comprises a processor, memory, and software including instructions for performing the methodologies of the invention, wherein the system includes a processor coupled to the memory elements, as well as in [0111] teaches the demand prediction system is integrated into a computer system comprising a memory, processors, and more), two or more clusters for a plurality of products ([0074-0075] teach performing macro clustering to identify different product categories by segmenting products into different categories, which can be performed using input features including characteristics of the products, as well as in [0079] teaches the product categories for demand learning include each cluster that is considered as a product category having a distinct demand range and characteristics, wherein the clusters have close characteristics that are combined based on the price sensitivity index, wherein the closeness can be measured by a distance metric; see also: [0073, 0080]),
the plurality of products having product consumption data with one or more data points indicative of varying product consumption quantities with respect to varying unit product values for a respective product ([0067] teaches the historical product data may include data points for the variety of products, such that the macro-clustering segments the data into product categories, wherein the macro-clustering can segment the data into categories according to different discovered demand behavior and a price sensitivity index is determined for every macro-cluster, wherein the macro-clusters are ranked by the price sensitivity index and the space is discretized as different categories, as well as in [0073] teaches there are sets of discrete price and demand values for each product or service are defined, wherein these sets of values are the initial historical product data, wherein the different products may be within a same space and have different demand curves, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price; see also: [0063, 0065, 0101-0104]),
within each cluster, for a first product having product consumption data with less than three data points ([0067] teaches in the case of predicting demand for a new product, the product data does not initially have any data points corresponding to the new product, as well as in [0064] teaches sufficient historical data may not be available for the product due to the new product not having any historical demand data, as well as in [0065] teaches in the case where the service was not previously offered on the market, the historic demand data will not be available; see also: [0002, 0089-0090]), determining, by the circuitry (Fig. 12 and [0127-0130] teach a computer system/server comprises a processor, memory, and software including instructions for performing the methodologies of the invention, wherein the system includes a processor coupled to the memory elements, as well as in [0111] teaches the demand prediction system is integrated into a computer system comprising a memory, processors, and more), a second product that has product consumption data with more than three data points ([0067] teaches a macro-customer segmentation model is created which segments the continuous data space into a number of macro clusters, wherein the segmentation can be by value in terms of demand, revenue/profit, etc., wherein the historical product data may include data points for a variety of products, as well as in [0073] teaches there are at least two sets of discrete price and demand values for each product or service, wherein the different products may be within a same space and have different demand curves, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price; see also: [0063, 0065, 0101-0104]),
wherein the determination of the second product comprises identification of a second identifier associated with the second product that is closest to a first identifier associated with the first product ([0067] teaches a macro-customer segmentation model is created which segments the continuous data space into a number of macro clusters, wherein the segmentation can be by value in terms of demand, revenue/profit, etc., wherein the historical product data may include data points for a variety of products, as well as in [0073] teaches there are at least two sets of discrete price and demand values for each product or service, wherein the different products may be within a same space and have different demand curves, wherein [0079] teaches the product categories for demand learning are defined as follows: each cluster is considered as a product category, such as having distinct demand range or characteristics, and clusters having close characteristics are combined/split based on the price sensitivity index, wherein in the macro-clustering method, the closeness can be measured by various distance metrics, such as Euclidian distance, wherein [0069-0070] teach the clustering method can calculate a threshold epsilon for clusters, which is a value that defines a maximum distance between two points in a cluster is below a threshold epsilon, wherein a value for epsilon can be calculated as a distance to the nearest n points for each point, wherein the determined number of clusters is calculated to maximize a desired metric, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price; see also: [0063, 0065, 0101-0104]);
retrieving, by the circuitry (Fig. 12 and [0127-0130] teach a computer system/server comprises a processor, memory, and software including instructions for performing the methodologies of the invention, wherein the system includes a processor coupled to the memory elements, as well as in [0111] teaches the demand prediction system is integrated into a computer system comprising a memory, processors, and more), a demand curve data associated with the second identifier of the second product from a database (Fig. 3 and [0066] teach the demand prediction method uses a k-means clustering method for macro clustering using k=alpha, wherein the k-means clustering is a method of vector quantization that aims to partition data into k clusters in which each observation belongs to a cluster with a nearest mean, as well as in [0078] teaches the macro clustering utilizes k-means clustering algorithms to generate clusters related to demand, revenue, and profit distribution, wherein [0069] teaches the points are clustered using a database scan clustering method or the like, as well as in [0130] teaches the memory provides program code that may be retrieved; see also: [0079, 0084-0085]),
wherein the demand curve data is indicative of product consumption quantities with respect to the varying unit product values of the plurality of products ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, as well as in Fig. 13 and [0109] teach a graph of the demand curves for each discrete price point, which is a probability density function, wherein the demand curves can be generated for each discrete price points, wherein the mean value for each of these demand curves represents the average demand at that price point, wherein the method selects the price based on the revenue that best increases the product of average demand and corresponding price; see also: [0094, 0110, 0116]);
computing, by the circuitry (Fig. 12 and [0127-0130] teach a computer system/server comprises a processor, memory, and software including instructions for performing the methodologies of the invention, wherein the system includes a processor coupled to the memory elements, as well as in [0111] teaches the demand prediction system is integrated into a computer system comprising a memory, processors, and more), a slope from the demand curve data of the second product ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0084] teaches for micro-clustering, each linear demand function is mapped to a point on a plane such that the y-coordinate is the slope of the demand curve and x-coordinate is the demand function, wherein an initial price can be a price that is used for a future product for which the demand learning is required, wherein [0082] teaches a price and demand pair includes data based on available historical data points, wherein the curve is based on prices that are used in the past; see also: [0085, 0094, 0104]),
wherein the slope indicates variation in the product consumption quantities with respect to a product value of the second product based on the slope ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0084] teaches for micro-clustering, each linear demand function is mapped to a point on a plane such that the y-coordinate is the slope of the demand curve and x-coordinate is the demand function, wherein an initial price can be a price that is used for a future product for which the demand learning is required, wherein [0082] teaches a price and demand pair includes data based on available historical data points, wherein the curve is based on prices that are used in the past; see also: [0085, 0094, 0104]),
calculating, by the circuitry (Fig. 12 and [0127-0130] teach a computer system/server comprises a processor, memory, and software including instructions for performing the methodologies of the invention, wherein the system includes a processor coupled to the memory elements, as well as in [0111] teaches the demand prediction system is integrated into a computer system comprising a memory, processors, and more), an optimal range of a unit product value for the first product ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have; see also: [0094]); and
storing, by the circuitry, the optimal range of the unit product value for the first product in the database ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0007] teaches storing the exemplary method steps in a storage medium; see also: [0094]).
However, Sawarkar does not explicitly teach each of the two or more clusters having a respective set of products with similar names, wherein the generation of the two or more clusters comprises: transforming names of the plurality of products into identifiers with numerical vectors using natural language processing techniques, and clustering the identifiers into two or more clusters based on similarity in names of the plurality of the plurality of products.
From the same or similar field of endeavor, Wellman teaches each of the two or more clusters having a respective set of products with similar names ([0045] teaches a clustering model may be configured to determine numeric space representations for attribute data associated with one or more data entities, such as sellers or stores, wherein the numeric space representations may be formed into clusters output from the clustering model, wherein the one or more clusters may be associated with a particular combination of attribute data, wherein the numeric space representations may be formed into clusters based on a similarity metric reaching a threshold, wherein the similarity metric is associated with a similarity of product names; see also: [0043-0044, 0055]),
wherein the generation of the two or more clusters comprises: clustering the identifiers into two or more clusters based on similarity in names of the plurality of the plurality of products ([0045] teaches a clustering model may be configured to determine numeric space representations for attribute data associated with one or more data entities, such as sellers or stores, wherein the numeric space representations may be formed into clusters output from the clustering model, wherein the one or more clusters may be associated with a particular combination of attribute data, wherein the numeric space representations may be formed into clusters based on a similarity metric reaching a threshold, wherein the similarity metric is associated with a similarity of product names; see also: [0043-0044, 0055]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Sawarkar to incorporate the teachings of Wellman to include each of the two or more clusters having a respective set of products with similar names, wherein the generation of the two or more clusters comprises: and clustering the identifiers into two or more clusters based on similarity in names of the plurality of the plurality of products. One would have been motivated to do so in order to generate clusters of behavior patterns based on the reduced-dimension features in order to fully capture consumer behavior patterns over time (Wellmann, [0003, 0008]). By incorporating the teachings of Wellmann, one would have been able to improve prediction of the categories of the items using transaction data (Wellmann, [0006]).
However, the combination of Sawarkar and Wellman does not explicitly teach wherein the generation of the two or more clusters comprises: transforming names of the plurality of products into identifiers with numerical vectors using natural language processing techniques.
From the same or similar field of endeavor, Rovatsos teaches wherein the generation of the two or more clusters comprises: transforming names of the plurality of products into identifiers with numerical vectors using natural language processing techniques ([0080] teaches relying on bidirectional encoder representations from transformers (BERT) natural language processing model to predict revenue for an item using the item name as input, wherein the classifier uses multiple transactional and auxiliary features from recent months including price, as well as item context feature using title BERT embeddings, wherein [0084] teaches categorizing items into clusters using two sets of features, wherein [0091] teaches identifying similar item features including name, wherein a text embedding may be generated based on item name features, wherein [0117] teaches generating context embeddings of items given each item name or product tile, wherein the model can generate an embedding representing the context of the name, wherein [0120] teaches text embeddings and features are obtained for each item, wherein k-means clustering is performed to generate a plurality of clusters; see also: [0030-0031, 0068, 0101]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar and Wellman to incorporate the teachings of Rovatsos to include wherein the generation of the two or more clusters comprises: transforming names of the plurality of products into identifiers with numerical vectors using natural language processing techniques. One would have been motivated to do so in order to generate accurate predictions using item content based similarity clusters (Rovatsos, [0084]). By incorporating the teachings of Rovatsos, one would have been able to identify high potential items through the use of bidirectional encoder representations from transformers embeddings using the item name as input (Rovatsos, [0030-0031]).
Regarding claim 2, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 1 above.
However, Sawarkar does not explicitly teach wherein the generating comprises determining identifiers based on name of each product from the plurality of products.
From the same or similar field of endeavor, Wellmann further teaches wherein the generating comprises determining identifiers based on name of each product from the plurality of products ([0039] teaches identifying an item name, wherein transaction data can be accessed including an item name, item identification number, and more, wherein the categories of items may be associated with particular respective identifiers, such as an item category identifier, which may be standardized, as well as in [0044] teaches aggregating data including item names, wherein [0045] teaches a clustering model may be configured to determine numeric space representations for attribute data associated with one or more data entities, such as sellers or stores, wherein the numeric space representations may be formed into clusters output from the clustering model, wherein the one or more clusters may be associated with a particular combination of attribute data, wherein the numeric space representations may be formed into clusters based on a similarity metric reaching a threshold, wherein the similarity metric is associated with a similarity of product names; see also: [0043, 0055]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar, Wellmann, and Rovatsos to incorporate the further teachings of Wellmann to include wherein the generating comprises determining identifiers based on name of each product from the plurality of products. One would have been motivated to do so in order to generate clusters of behavior patterns based on the reduced-dimension features in order to fully capture consumer behavior patterns over time (Wellmann, [0003, 0008]). By incorporating the teachings of Wellmann, one would have been able to improve prediction of the categories of the items using transaction data (Wellmann, [0006]).
Regarding claim 3, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 2 above.
However, Sawarkar does not explicitly teach wherein the identifiers are determined using the natural language processing techniques.
From the same or similar field of endeavor, Rovatsos further teaches wherein the identifiers are determined using the natural language processing techniques ([0080] teaches relying on bidirectional encoder representations from transformers (BERT) natural language processing model to predict revenue for an item using the item name as input, wherein the classifier uses multiple transactional and auxiliary features from recent months including price, as well as item context feature using title BERT embeddings, wherein [0084] teaches categorizing items into clusters using two sets of features, wherein [0091] teaches identifying similar item features including name, wherein a text embedding may be generated based on item name features, wherein [0117] teaches generating context embeddings of items given each item name or product tile, wherein the model can generate an embedding representing the context of the name, wherein [0120] teaches text embeddings and features are obtained for each item, wherein k-means clustering is performed to generate a plurality of clusters; see also: [0030-0031, 0068, 0101]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar, Wellmann, and Rovatsos to incorporate the further teachings of Rovatsos to include wherein the identifiers are determined using the natural language processing techniques. One would have been motivated to do so in order to generate accurate predictions using item content based similarity clusters (Rovatsos, [0084]). By incorporating the teachings of Rovatsos, one would have been able to identify high potential items through the use of bidirectional encoder representations from transformers embeddings using the item name as input (Rovatsos, [0030-0031]).
Regarding claim 4, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 2 above.
However, Sawarkar does not explicitly teach wherein the method comprises employing a k-means clustering technique for generating the two or more clusters.
From the same or similar field of endeavor, Rovatsos further teaches wherein the method comprises employing a k-means clustering technique for generating the two or more clusters ([0080] teaches relying on bidirectional encoder representations from transformers (BERT) natural language processing model to predict revenue for an item using the item name as input, wherein the classifier uses multiple transactional and auxiliary features from recent months including price, as well as item context feature using title BERT embeddings, wherein [0084] teaches categorizing items into clusters using two sets of features, wherein [0091] teaches identifying similar item features including name, wherein a text embedding may be generated based on item name features, wherein [0117] teaches generating context embeddings of items given each item name or product tile, wherein the model can generate an embedding representing the context of the name, wherein [0120] teaches text embeddings and features are obtained for each item, wherein k-means clustering is performed to generate a plurality of clusters; see also: [0030-0031, 0068, 0101]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar, Wellmann, and Rovatsos to incorporate the further teachings of Rovatsos to include wherein the method comprises employing a k-means clustering technique for generating the two or more clusters. One would have been motivated to do so in order to generate accurate predictions using item content based similarity clusters (Rovatsos, [0084]). By incorporating the teachings of Rovatsos, one would have been able to identify high potential items through the use of bidirectional encoder representations from transformers embeddings using the item name as input (Rovatsos, [0030-0031]).
Regarding claim 5, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 2 above.
Sawarkar further teaches wherein the determining the second product with respect to the first product is based on: a minimum distance between the identifier of the first product and the identifier of the second product (Fig. 3 and [0066] teach the demand prediction method uses a k-means clustering method for macro clustering using k=alpha, wherein the k-means clustering is a method of vector quantization that aims to partition data into k clusters in which each observation belongs to a cluster with a nearest mean, as well as in [0078] teaches the macro clustering utilizes k-means clustering algorithms to generate clusters related to demand, revenue, and profit distribution, wherein [0079] teaches the product categories for demand learning are defined as follows: each cluster is considered as a product category, such as having distinct demand range or characteristics, and clusters having close characteristics are combined/split based on the price sensitivity index, wherein in the macro-clustering method, the closeness can be measured by various distance metrics, such as Euclidian distance, wherein [0070] teaches the density clustering also finds a minimum value of epsilon, ensuring a correct number of clusters is determined, wherein a value of epsilon can be calculated as a distance to the nearest n points; see also: [0084-0085]);
and a pre-defined limit for variation in a product value of the first product with respect to the product value of the second product (Fig. 3 and [0066] teach the demand prediction method uses a k-means clustering method for macro clustering using k=alpha, wherein the k-means clustering is a method of vector quantization that aims to partition data into k clusters in which each observation belongs to a cluster with a nearest mean, as well as in [0078] teaches the macro clustering utilizes k-means clustering algorithms to generate clusters related to demand, revenue, and profit distribution, wherein [0079] teaches the product categories for demand learning are defined as follows: each cluster is considered as a product category, such as having distinct demand range or characteristics, and clusters having close characteristics are combined/split based on the price sensitivity index, wherein in the macro-clustering method, the closeness can be measured by various distance metrics, such as Euclidian distance, wherein [0069-0070] teach the clustering method can calculate a threshold epsilon for clusters, which is a value that defines a maximum distance between two points in a cluster is below a threshold epsilon, wherein a value for epsilon can be calculated as a distance to the nearest n points for each point, wherein the determined number of clusters is calculated to maximize a desired metric; see also: [084-0085]).
Regarding claim 7, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 1 above.
Sawarkar further teaches wherein computing the optimal range of the unit product value for the first product comprises computing a slope from demand curve data of the second product ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0084] teaches for micro-clustering, each linear demand function is mapped to a point on a plane such that the y-coordinate is the slope of the demand curve and x-coordinate is the demand function, wherein an initial price can be a price that is used for a future product for which the demand learning is required, wherein [0082] teaches a price and demand pair includes data based on available historical data points, wherein the curve is based on prices that are used in the past; see also: [0085, 0094, 0104]).
Regarding claim 8, Sawarkar teaches a system comprising (Fig. 12 and [0127-0130] teach a computer system/server comprises a processor, memory, and software including instructions for performing the methodologies of the invention, wherein the system includes a processor coupled to the memory elements, as well as in [0111] teaches the demand prediction system is integrated into a computer system comprising a memory, processors, and more):
a processor (Fig. 12 and [0127-0130] teach a computer system/server comprises a processor, memory, and software including instructions for performing the methodologies of the invention, wherein the system includes a processor coupled to the memory elements, as well as in [0111] teaches the demand prediction system is integrated into a computer system comprising a memory, processors, and more);
a circuitry (Fig. 12 and [0127-0130] teach a computer system/server comprises a processor, memory, and software including instructions for performing the methodologies of the invention, wherein the system includes a processor coupled to the memory elements, as well as in [0111] teaches the demand prediction system is integrated into a computer system comprising a memory, processors, and more), coupled to the processor, to:
determine a plurality of identifiers corresponding to names of a plurality of products ([0074-0075] teach performing macro clustering to identify different product categories by segmenting products into different categories, which can be performed using input features including characteristics of the products, as well as in [0079] teaches the product categories for demand learning include each cluster that is considered as a product category having a distinct demand range and characteristics, wherein the clusters have close characteristics that are combined based on the price sensitivity index, wherein the closeness can be measured by a distance metric; see also: [0073, 0080]; Examiner’s Note: See the 35 USC 103 combination below for teachings pertaining to the entire claim limitation.),
the plurality of products having product consumption data with one or more data points indicative of varying product consumption quantities with respect to varying unit product values for a respective product ([0067] teaches the historical product data may include data points for the variety of products, such that the macro-clustering segments the data into product categories, wherein the macro-clustering can segment the data into categories according to different discovered demand behavior and a price sensitivity index is determined for every macro-cluster, wherein the macro-clusters are ranked by the price sensitivity index and the space is discretized as different categories, as well as in [0073] teaches there are sets of discrete price and demand values for each product or service are defined, wherein these sets of values are the initial historical product data, wherein the different products may be within a same space and have different demand curves, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price; see also: [0063, 0065, 0101-0104]);
generate two or more clusters for the plurality of identifiers ([0067] teaches the historical product data may include data points for the variety of products, such that the macro-clustering segments the data into product categories, wherein the macro-clustering can segment the data into categories according to different discovered demand behavior and a price sensitivity index is determined for every macro-cluster, wherein the macro-clusters are ranked by the price sensitivity index and the space is discretized as different categories, as well as in [0073] teaches there are sets of discrete price and demand values for each product or service are defined, wherein these sets of values are the initial historical product data, wherein the different products may be within a same space and have different demand curves, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price, wherein [0078] teaches the number of clusters is identified based on the inter-cluster vs intra-cluster separation distances, wherein the demand, revenue, and profit distribution for each cluster is visualized, as well as in [0085] teaches choosing an ideal number of k centroid points or clusters for the algorithm, wherein the k-fold cross validation technique can be used to find the ideal K value; see also: [0063, 0065, 0101-0104]),
each of the two or more clusters having a respective set of identifiers from amongst the plurality of identifiers ([0067] teaches the historical product data may include data points for the variety of products, such that the macro-clustering segments the data into product categories, wherein the macro-clustering can segment the data into categories according to different discovered demand behavior and a price sensitivity index is determined for every macro-cluster, wherein the macro-clusters are ranked by the price sensitivity index and the space is discretized as different categories, as well as in [0073] teaches there are sets of discrete price and demand values for each product or service are defined, wherein these sets of values are the initial historical product data, wherein the different products may be within a same space and have different demand curves, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price; see also: [0063, 0065, 0101-0104]),
within each of the two or more clusters, for a first identifier associated with product consumption data having less than three data points ([0067] teaches in the case of predicting demand for a new product, the product data does not initially have any data points corresponding to the new product, as well as in [0064] teaches sufficient historical data may not be available for the product due to the new product not having any historical demand data, as well as in [0065] teaches in the case where the service was not previously offered on the market, the historic demand data will not be available; see also: [0002, 0089-0090]),
determine a second identifier associated with product consumption data having more than three data points ([0067] teaches a macro-customer segmentation model is created which segments the continuous data space into a number of macro clusters, wherein the segmentation can be by value in terms of demand, revenue/profit, etc., wherein the historical product data may include data points for a variety of products, as well as in [0073] teaches there are at least two sets of discrete price and demand values for each product or service, wherein the different products may be within a same space and have different demand curves, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price; see also: [0063, 0065, 0101-0104]),
wherein the determination is based at least in part on a closest distance between the first identifier and the second identifier ([0067] teaches a macro-customer segmentation model is created which segments the continuous data space into a number of macro clusters, wherein the segmentation can be by value in terms of demand, revenue/profit, etc., wherein the historical product data may include data points for a variety of products, as well as in [0073] teaches there are at least two sets of discrete price and demand values for each product or service, wherein the different products may be within a same space and have different demand curves, wherein [0079] teaches the product categories for demand learning are defined as follows: each cluster is considered as a product category, such as having distinct demand range or characteristics, and clusters having close characteristics are combined/split based on the price sensitivity index, wherein in the macro-clustering method, the closeness can be measured by various distance metrics, such as Euclidian distance, wherein [0069-0070] teach the clustering method can calculate a threshold epsilon for clusters, which is a value that defines a maximum distance between two points in a cluster is below a threshold epsilon, wherein a value for epsilon can be calculated as a distance to the nearest n points for each point, wherein the determined number of clusters is calculated to maximize a desired metric, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price; see also: [0063, 0065, 0101-0104]);
retrieve a demand curve data associated with the second identifier of a second product from a database (Fig. 3 and [0066] teach the demand prediction method uses a k-means clustering method for macro clustering using k=alpha, wherein the k-means clustering is a method of vector quantization that aims to partition data into k clusters in which each observation belongs to a cluster with a nearest mean, as well as in [0078] teaches the macro clustering utilizes k-means clustering algorithms to generate clusters related to demand, revenue, and profit distribution, wherein [0069] teaches the points are clustered using a database scan clustering method or the like, as well as in [0130] teaches the memory provides program code that may be retrieved; see also: [0079, 0084-0085]),
wherein the demand curve data is indicative of product consumption quantities with respect to the varying unit product values ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, as well as in Fig. 13 and [0109] teach a graph of the demand curves for each discrete price point, which is a probability density function, wherein the demand curves can be generated for each discrete price points, wherein the mean value for each of these demand curves represents the average demand at that price point, wherein the method selects the price based on the revenue that best increases the product of average demand and corresponding price; see also: [0094, 0110, 0116]);
compute a slope from the demand curve data of the second product ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0084] teaches for micro-clustering, each linear demand function is mapped to a point on a plane such that the y-coordinate is the slope of the demand curve and x-coordinate is the demand function, wherein an initial price can be a price that is used for a future product for which the demand learning is required, wherein [0082] teaches a price and demand pair includes data based on available historical data points, wherein the curve is based on prices that are used in the past; see also: [0085, 0094, 0104]),
wherein the slope indicates variation in the product consumption quantities with respect to a product value of the second product ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0084] teaches for micro-clustering, each linear demand function is mapped to a point on a plane such that the y-coordinate is the slope of the demand curve and x-coordinate is the demand function, wherein an initial price can be a price that is used for a future product for which the demand learning is required, wherein [0082] teaches a price and demand pair includes data based on available historical data points, wherein the curve is based on prices that are used in the past; see also: [0085, 0094, 0104]);
based on the slope, calculate an optimal range of a unit product value for a product associated with the first identifier ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, and wherein [0067] teaches demand curves are mapped to a plane and micro-clustering is created with centroids of k, wherein a target, or optimal, price is calculated for the given combination of k and its maximum revenue at each segment (e.g. each combination of demand and price, segmented by product/service category), wherein the system can identify a final target price; see also: [0094]); and
store the optimal range of the unit product value for the first product in the database ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0007] teaches storing the exemplary method steps in a storage medium; see also: [0094]).
However, Sawarkar does not explicitly teach determine a plurality of identifiers corresponding to names of a plurality of products, wherein the generation of the two or more clusters comprises transforming names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques, the set of identifiers relating to products with similar names.
From the same or similar field of endeavor, Wellman teaches determine a plurality of identifiers corresponding to names of a plurality of products ([0039] teaches identifying an item name, as well as in [0044] teaches aggregating data including item names, wherein [0045] teaches a clustering model may be configured to determine numeric space representations for attribute data associated with one or more data entities, such as sellers or stores, wherein the numeric space representations may be formed into clusters output from the clustering model, wherein the one or more clusters may be associated with a particular combination of attribute data, wherein the numeric space representations may be formed into clusters based on a similarity metric reaching a threshold, wherein the similarity metric is associated with a similarity of product names; see also: [0043, 0055]),
the set of identifiers relating to products with similar names ([0045] teaches a clustering model may be configured to determine numeric space representations for attribute data associated with one or more data entities, such as sellers or stores, wherein the numeric space representations may be formed into clusters output from the clustering model, wherein the one or more clusters may be associated with a particular combination of attribute data, wherein the numeric space representations may be formed into clusters based on a similarity metric reaching a threshold, wherein the similarity metric is associated with a similarity of product names; see also: [0043-0044, 0055]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Sawarkar to incorporate the teachings of determine a plurality of identifiers corresponding to names of a plurality of products, the set of identifiers relating to products with similar names. One would have been motivated to do so in order to generate clusters of behavior patterns based on the reduced-dimension features in order to fully capture consumer behavior patterns over time (Wellmann, [0003, 0008]). By incorporating the teachings of Wellmann, one would have been able to improve prediction of the categories of the items using transaction data (Wellmann, [0006]).
However, the combination of Sawarkar and Wellman does not explicitly teach wherein the generation of the two or more clusters comprises transforming names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques.
From the same or similar field of endeavor, Rovatsos teaches wherein the generation of the two or more clusters comprises transforming names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques ([0080] teaches relying on bidirectional encoder representations from transformers (BERT) natural language processing model to predict revenue for an item using the item name as input, wherein the classifier uses multiple transactional and auxiliary features from recent months including price, as well as item context feature using title BERT embeddings, wherein [0084] teaches categorizing items into clusters using two sets of features, wherein [0091] teaches identifying similar item features including name, wherein a text embedding may be generated based on item name features, wherein [0117] teaches generating context embeddings of items given each item name or product tile, wherein the model can generate an embedding representing the context of the name, wherein [0120] teaches text embeddings and features are obtained for each item, wherein k-means clustering is performed to generate a plurality of clusters; see also: [0030-0031, 0068, 0101]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar and Wellman to incorporate the teachings of Rovatsos to include wherein the generation of the two or more clusters comprises transforming names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques. One would have been motivated to do so in order to generate accurate predictions using item content based similarity clusters (Rovatsos, [0084]). By incorporating the teachings of Rovatsos, one would have been able to identify high potential items through the use of bidirectional encoder representations from transformers embeddings using the item name as input (Rovatsos, [0030-0031]).
Regarding claim 9, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 8 above.
However, Sawarkar does not explicitly teach wherein the plurality of identifiers is indicative of numerical values associated with the names of the plurality of products.
From the same or similar field of endeavor, Wellmann further teaches wherein the plurality of identifiers is indicative of numerical values associated with the names of the plurality of products ([0039] teaches identifying an item name, wherein transaction data can be accessed including an item name, item identification number, and more, wherein the categories of items may be associated with particular respective identifiers, such as an item category identifier, which may be standardized, as well as in [0044] teaches aggregating data including item names, wherein [0045] teaches a clustering model may be configured to determine numeric space representations for attribute data associated with one or more data entities, such as sellers or stores, wherein the numeric space representations may be formed into clusters output from the clustering model, wherein the one or more clusters may be associated with a particular combination of attribute data, wherein the numeric space representations may be formed into clusters based on a similarity metric reaching a threshold, wherein the similarity metric is associated with a similarity of product names; see also: [0043, 0055]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar, Wellmann, and Rovatsos to incorporate the further teachings of Wellmann to include wherein the plurality of identifiers is indicative of numerical values associated with the names of the plurality of products. One would have been motivated to do so in order to generate clusters of behavior patterns based on the reduced-dimension features in order to fully capture consumer behavior patterns over time (Wellmann, [0003, 0008]). By incorporating the teachings of Wellmann, one would have been able to improve prediction of the categories of the items using transaction data (Wellmann, [0006]).
Regarding claim 10, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 8 above.
However, Sawarkar does not explicitly teach wherein the circuitry uses the natural language processing techniques to generate the plurality of identifiers.
From the same or similar field of endeavor, Rovatsos further teaches wherein the circuitry uses the natural language processing techniques to generate the plurality of identifiers ([0080] teaches relying on bidirectional encoder representations from transformers (BERT) natural language processing model to predict revenue for an item using the item name as input, wherein the classifier uses multiple transactional and auxiliary features from recent months including price, as well as item context feature using title BERT embeddings, wherein [0084] teaches categorizing items into clusters using two sets of features, wherein [0091] teaches identifying similar item features including name, wherein a text embedding may be generated based on item name features, wherein [0117] teaches generating context embeddings of items given each item name or product tile, wherein the model can generate an embedding representing the context of the name, wherein [0120] teaches text embeddings and features are obtained for each item, wherein k-means clustering is performed to generate a plurality of clusters; see also: [0030-0031, 0068, 0101]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar, Wellmann, and Rovatsos to incorporate the further teachings of Rovatsos to include wherein the circuitry uses the natural language processing techniques to generate the plurality of identifiers. One would have been motivated to do so in order to generate accurate predictions using item content based similarity clusters (Rovatsos, [0084]). By incorporating the teachings of Rovatsos, one would have been able to identify high potential items through the use of bidirectional encoder representations from transformers embeddings using the item name as input (Rovatsos, [0030-0031]).
Regarding claim 11, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 8 above.
Sawarkar further teaches wherein to generate the two or more clusters, the circuitry is to determine an optimal number of clusters in which the plurality of identifiers is to be clustered ([0067] teaches the historical product data may include data points for the variety of products, such that the macro-clustering segments the data into product categories, wherein the macro-clustering can segment the data into categories according to different discovered demand behavior and a price sensitivity index is determined for every macro-cluster, wherein the macro-clusters are ranked by the price sensitivity index and the space is discretized as different categories, as well as in [0073] teaches there are sets of discrete price and demand values for each product or service are defined, wherein these sets of values are the initial historical product data, wherein the different products may be within a same space and have different demand curves, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price, wherein [0078] teaches the number of clusters is identified based on the inter-cluster vs intra-cluster separation distances, wherein the demand, revenue, and profit distribution for each cluster is visualized, as well as in [0085] teaches choosing an ideal number of k centroid points or clusters for the algorithm, wherein the k-fold cross validation technique can be used to find the ideal K value; see also: [0063, 0065, 0101-0104]).
Regarding claim 12, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 11 above.
Sawarkar further teaches wherein to determine the optimal number of clusters, the circuitry is to define different numbers of clusters and calculate sum of a squared distance between an identifier and a centroid in each cluster ([0067] teaches the historical product data may include data points for the variety of products, such that the macro-clustering segments the data into product categories, wherein the macro-clustering can segment the data into categories according to different discovered demand behavior and a price sensitivity index is determined for every macro-cluster, wherein the macro-clusters are ranked by the price sensitivity index and the space is discretized as different categories, as well as in [0073] teaches there are sets of discrete price and demand values for each product or service are defined, wherein these sets of values are the initial historical product data, wherein the different products may be within a same space and have different demand curves, as well as in [0080] teaches the historical product data includes price and demand, wherein [0081] teaches determining a linear demand function being fit using a least squares method, as well as in [0107] teaches using a least squares method, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price, wherein [0078] teaches the number of clusters is identified based on the inter-cluster vs intra-cluster separation distances, wherein the demand, revenue, and profit distribution for each cluster is visualized, as well as in [0085] teaches choosing an ideal number of k centroid points or clusters for the algorithm, wherein the k-fold cross validation technique can be used to find the ideal K value; see also: [0063, 0065, 0101-0104]).
Regarding claim 14, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 8 above.
Sawarkar further teaches wherein to determine the second identifier, the circuitry is to determine a pre-defined variation in the product value of a first product associated with the first identifier and the product value of the second product associated with the second identifier (Fig. 3 and [0066] teach the demand prediction method uses a k-means clustering method for macro clustering using k=alpha, wherein the k-means clustering is a method of vector quantization that aims to partition data into k clusters in which each observation belongs to a cluster with a nearest mean, as well as in [0078] teaches the macro clustering utilizes k-means clustering algorithms to generate clusters related to demand, revenue, and profit distribution, wherein [0079] teaches the product categories for demand learning are defined as follows: each cluster is considered as a product category, such as having distinct demand range or characteristics, and clusters having close characteristics are combined/split based on the price sensitivity index, wherein in the macro-clustering method, the closeness can be measured by various distance metrics, such as Euclidian distance, wherein [0069-0070] teach the clustering method can calculate a threshold epsilon for clusters, which is a value that defines a maximum distance between two points in a cluster is below a threshold epsilon, wherein a value for epsilon can be calculated as a distance to the nearest n points for each point, wherein the determined number of clusters is calculated to maximize a desired metric; see also: [084-0085]).
Regarding claim 15, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 8 above.
Sawarkar further teaches wherein to compute the optimal range of the unit product value for the first product, the circuitry is to compute the slope from the demand curve data of the second product associated with the second identifier ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0084] teaches for micro-clustering, each linear demand function is mapped to a point on a plane such that the y-coordinate is the slope of the demand curve and x-coordinate is the demand function, wherein an initial price can be a price that is used for a future product for which the demand learning is required, wherein [0082] teaches a price and demand pair includes data based on available historical data points, wherein the curve is based on prices that are used in the past; see also: [0085, 0094, 0104]).
Regarding claim 16, Sawarkar teaches a non-transitory computer-readable medium having instructions stored thereon, the instructions ([0007] teaches a computer program product including a computer readable storage medium with computer usable program code for performing the method, as well as in [0135] teaches the system comprising computer readable storage medium that are executed using one or more hardware processors; see also: [0118, 0124-0125]), when executed by a processor, cause the processor to perform operations comprising:
generating identifiers corresponding to names of a plurality of products ([0074-0075] teach performing macro clustering to identify different product categories by segmenting products into different categories, which can be performed using input features including characteristics of the products, as well as in [0079] teaches the product categories for demand learning include each cluster that is considered as a product category having a distinct demand range and characteristics, wherein the clusters have close characteristics that are combined based on the price sensitivity index, wherein the closeness can be measured by a distance metric; see also: [0073, 0080]; Examiner’s Note: See the 35 USC 103 combination below for teachings pertaining to the entire claim limitation.),
wherein the plurality of products having product consumption data with one or more data points indicative of varying product consumption quantities with respect to varying unit product values for a respective product ([0067] teaches the historical product data may include data points for the variety of products, such that the macro-clustering segments the data into product categories, wherein the macro-clustering can segment the data into categories according to different discovered demand behavior and a price sensitivity index is determined for every macro-cluster, wherein the macro-clusters are ranked by the price sensitivity index and the space is discretized as different categories, as well as in [0073] teaches there are sets of discrete price and demand values for each product or service are defined, wherein these sets of values are the initial historical product data, wherein the different products may be within a same space and have different demand curves, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price; see also: [0063, 0065, 0101-0104]),
clustering the identifiers into an optimal number of clusters ([0067] teaches the historical product data may include data points for the variety of products, such that the macro-clustering segments the data into product categories, wherein the macro-clustering can segment the data into categories according to different discovered demand behavior and a price sensitivity index is determined for every macro-cluster, wherein the macro-clusters are ranked by the price sensitivity index and the space is discretized as different categories, as well as in [0073] teaches there are sets of discrete price and demand values for each product or service are defined, wherein these sets of values are the initial historical product data, wherein the different products may be within a same space and have different demand curves, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price, wherein [0078] teaches the number of clusters is identified based on the inter-cluster vs intra-cluster separation distances, wherein the demand, revenue, and profit distribution for each cluster is visualized, as well as in [0085] teaches choosing an ideal number of k centroid points or clusters for the algorithm, wherein the k-fold cross validation technique can be used to find the ideal K value; see also: [0063, 0065, 0101-0104]),
within each cluster, determining a pair of identifiers closest to each other and having a variation in product values within a pre-defined limit ([0067] teaches a macro-customer segmentation model is created which segments the continuous data space into a number of macro clusters, wherein the segmentation can be by value in terms of demand, revenue/profit, etc., wherein the historical product data may include data points for a variety of products, as well as in [0073] teaches there are at least two sets of discrete price and demand values for each product or service, wherein the different products may be within a same space and have different demand curves, wherein [0079] teaches the product categories for demand learning are defined as follows: each cluster is considered as a product category, such as having distinct demand range or characteristics, and clusters having close characteristics are combined/split based on the price sensitivity index, wherein in the macro-clustering method, the closeness can be measured by various distance metrics, such as Euclidian distance, wherein [0069-0070] teach the clustering method can calculate a threshold epsilon for clusters, which is a value that defines a maximum distance between two points in a cluster is below a threshold epsilon, wherein a value for epsilon can be calculated as a distance to the nearest n points for each point, wherein the determined number of clusters is calculated to maximize a desired metric, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price; see also: [0063, 0065, 0101-0104]),
a first identifier from the pair of identifiers relates to a product having product consumption data with less than three data points and a second identifier from the pair of identifiers relates to a product having product consumption data with more than three data points ([0067] teaches a macro-customer segmentation model is created which segments the continuous data space into a number of macro clusters, wherein the segmentation can be by value in terms of demand, revenue/profit, etc., wherein the historical product data may include data points for a variety of products, as well as in [0073] teaches there are at least two sets of discrete price and demand values for each product or service, wherein the different products may be within a same space and have different demand curves, wherein [0079] teaches the product categories for demand learning are defined as follows: each cluster is considered as a product category, such as having distinct demand range or characteristics, and clusters having close characteristics are combined/split based on the price sensitivity index, wherein in the macro-clustering method, the closeness can be measured by various distance metrics, such as Euclidian distance, wherein [0069-0070] teach the clustering method can calculate a threshold epsilon for clusters, which is a value that defines a maximum distance between two points in a cluster is below a threshold epsilon, wherein a value for epsilon can be calculated as a distance to the nearest n points for each point, wherein the determined number of clusters is calculated to maximize a desired metric, as well as in [0080] teaches the historical product data includes price and demand, as well as in [0074] teaches the clustering can be performed in order to place products into different categories, wherein the input features can include the demographics of customers using the products, characteristics of the product, time-based aggregate features, etc., wherein [0100] teaches utilizing macro-clustering to find the best optimal price; see also: [0063, 0065, 0101-0104]);
retrieving, by the circuitry, a demand curve data associated with the second identifier of the second product from a database (Fig. 3 and [0066] teach the demand prediction method uses a k-means clustering method for macro clustering using k=alpha, wherein the k-means clustering is a method of vector quantization that aims to partition data into k clusters in which each observation belongs to a cluster with a nearest mean, as well as in [0078] teaches the macro clustering utilizes k-means clustering algorithms to generate clusters related to demand, revenue, and profit distribution, wherein [0069] teaches the points are clustered using a database scan clustering method or the like, as well as in [0130] teaches the memory provides program code that may be retrieved; see also: [0079, 0084-0085]),
wherein the demand curve data is indicative of product consumption quantities with respect to the varying unit product values of the plurality of products ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0084] teaches for micro-clustering, each linear demand function is mapped to a point on a plane such that the y-coordinate is the slope of the demand curve and x-coordinate is the demand function, wherein an initial price can be a price that is used for a future product for which the demand learning is required, wherein [0082] teaches a price and demand pair includes data based on available historical data points, wherein the curve is based on prices that are used in the past; see also: [0085, 0094, 0104]);
computing, by the circuitry, a slope from the demand curve data of the second product ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0084] teaches for micro-clustering, each linear demand function is mapped to a point on a plane such that the y-coordinate is the slope of the demand curve and x-coordinate is the demand function, wherein an initial price can be a price that is used for a future product for which the demand learning is required, wherein [0082] teaches a price and demand pair includes data based on available historical data points, wherein the curve is based on prices that are used in the past; see also: [0085, 0094, 0104]),
wherein the slope indicates variation in the product consumption quantities with respect to a product value of the second product ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0084] teaches for micro-clustering, each linear demand function is mapped to a point on a plane such that the y-coordinate is the slope of the demand curve and x-coordinate is the demand function, wherein an initial price can be a price that is used for a future product for which the demand learning is required, wherein [0082] teaches a price and demand pair includes data based on available historical data points, wherein the curve is based on prices that are used in the past; see also: [0085, 0094, 0104]);
based on the slope, calculating, an optimal range of a unit product value for the product associated with the first identifier ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, and wherein [0067] teaches demand curves are mapped to a plane and micro-clustering is created with centroids of k, wherein a target, or optimal, price is calculated for the given combination of k and its maximum revenue at each segment (e.g. each combination of demand and price, segmented by product/service category), wherein the system can identify a final target price; see also: [0094]);
and storing the optimal range of the unit product value for the first product in the database ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0007] teaches storing the exemplary method steps in a storage medium; see also: [0094]).
However, Sawarkar does not explicitly teach generating identifiers corresponding to names of a plurality of products, wherein the generation of the identifiers comprises transforming the names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques; the clustering is based on a similarity in the names of the plurality of products.
From the same or similar field of endeavor, Wellman teaches generating identifiers corresponding to names of a plurality of products ([0039] teaches identifying an item name, as well as in [0044] teaches aggregating data including item names, wherein [0045] teaches a clustering model may be configured to determine numeric space representations for attribute data associated with one or more data entities, such as sellers or stores, wherein the numeric space representations may be formed into clusters output from the clustering model, wherein the one or more clusters may be associated with a particular combination of attribute data, wherein the numeric space representations may be formed into clusters based on a similarity metric reaching a threshold, wherein the similarity metric is associated with a similarity of product names; see also: [0043, 0055]),
the clustering is based on a similarity in the names of the plurality of products ([0045] teaches a clustering model may be configured to determine numeric space representations for attribute data associated with one or more data entities, such as sellers or stores, wherein the numeric space representations may be formed into clusters output from the clustering model, wherein the one or more clusters may be associated with a particular combination of attribute data, wherein the numeric space representations may be formed into clusters based on a similarity metric reaching a threshold, wherein the similarity metric is associated with a similarity of product names; see also: [0043-0044, 0055]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Sawarkar to incorporate the teachings of Wellman to include generating identifiers corresponding to names of a plurality of products, the clustering is based on a similarity in the names of the plurality of products. One would have been motivated to do so in order to generate clusters of behavior patterns based on the reduced-dimension features in order to fully capture consumer behavior patterns over time (Wellmann, [0003, 0008]). By incorporating the teachings of Wellmann, one would have been able to improve prediction of the categories of the items using transaction data (Wellmann, [0006]).
However, the combination of Sawarkar and Wellman does not explicitly teach wherein the generation of the identifiers comprises transforming the names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques.
From the same or similar field of endeavor, Rovatsos teaches wherein the generation of the identifiers comprises transforming the names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques ([0080] teaches relying on bidirectional encoder representations from transformers (BERT) natural language processing model to predict revenue for an item using the item name as input, wherein the classifier uses multiple transactional and auxiliary features from recent months including price, as well as item context feature using title BERT embeddings, wherein [0084] teaches categorizing items into clusters using two sets of features, wherein [0091] teaches identifying similar item features including name, wherein a text embedding may be generated based on item name features, wherein [0117] teaches generating context embeddings of items given each item name or product tile, wherein the model can generate an embedding representing the context of the name, wherein [0120] teaches text embeddings and features are obtained for each item, wherein k-means clustering is performed to generate a plurality of clusters; see also: [0030-0031, 0068, 0101]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar and Wellman to incorporate the teachings of Rovatsos to include wherein the generation of the identifiers comprises transforming the names of the plurality of products into set of identifiers with numerical vectors using natural language processing techniques. One would have been motivated to do so in order to generate accurate predictions using item content based similarity clusters (Rovatsos, [0084]). By incorporating the teachings of Rovatsos, one would have been able to identify high potential items through the use of bidirectional encoder representations from transformers embeddings using the item name as input (Rovatsos, [0030-0031]).
Regarding claim 18, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 16 above.
Sawarkar further teaches wherein the optimal number of clusters is determined using an elbow method (Fig. 3 and [0066] teach the demand prediction method uses a k-means clustering method for macro clustering using k=alpha, wherein the k-means clustering is a method of vector quantization that aims to partition data into k clusters in which each observation belongs to a cluster with a nearest mean, wherein [0087] teaches the selected k value is the one that has a minimum average error value, wherein the lowest k at a first knee or elbow point is selected; see also: [0088-0089]).
Regarding claim 19, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 16 above.
Sawarkar further teaches wherein the clustering is performed using a k-means technique (Fig. 3 and [0066] teach the demand prediction method uses a k-means clustering method for macro clustering using k=alpha, wherein the k-means clustering is a method of vector quantization that aims to partition data into k clusters in which each observation belongs to a cluster with a nearest mean, as well as in [0078] teaches the macro clustering utilizes k-means clustering algorithms to generate clusters related to demand, revenue, and profit distribution; see also: [0079, 0084-0085]).
Regarding claim 20, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 16 above.
Sawarkar further teaches wherein computing the optimal range of the unit product value for the product associated with the first identifier comprises computing the slope from the demand curve data pertaining to the product associated with the second identifier ([0089] teaches optimal prices for each product are generated, wherein given the final set of hypothesis demand function/curves for each product category, the method applies dynamic pricing for a new product that may belong to any of the product category, wherein the method seeks to generate a price for each learning period consecutively such that the demand with theoretically maximize the revenue, wherein [0091-0093] teach determining the predicted optimal price for the product to generate a maximum revenue, wherein the price can be determined based on determining a range of values for the variable that are prices that the product can have, wherein [0084] teaches for micro-clustering, each linear demand function is mapped to a point on a plane such that the y-coordinate is the slope of the demand curve and x-coordinate is the demand function, wherein an initial price can be a price that is used for a future product for which the demand learning is required, wherein [0082] teaches a price and demand pair includes data based on available historical data points, wherein the curve is based on prices that are used in the past; see also: [0085, 0094, 0104]).
Claim(s) 6 is rejected under 35 U.S.C. 103 as being unpatentable over Sawarkar et al. (US 20220138786 A1) in view of Wellmann et al. (US 20240221009 A1) in view of Rovatsos et al. (US 20240354810 A1) in view of Irshad et al. (US 20230394424 A1).
Regarding claim 6, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 5 above.
However, Sawarkar does not explicitly teach wherein the pre-defined limit for variation in the product value of the first product with respect to the product value of the second product is about 25%.
From the same or similar field of endeavor, Irshad teaches wherein the pre-defined limit for variation in the product value of the first product with respect to the product value of the second product is 25% ([0042] teaches each of the clusters can be categorized into demand bands based on their rate of change, wherein the demand bands may be defined using discrete thresholds to separate them, wherein the discrete thresholds may be used to subdivide clusters into groupings that are sufficiently close in terms of types and quantity of demand, wherein the demand may be within a predetermined range or percentage, such as 25%; see also: [0045, 0060]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar, Wellmann, and Rovatsos t to incorporate the teachings of Irshad to include wherein the pre-defined limit for variation in the product value of the first product with respect to the product value of the second product is 25%. One would have been motivated to do so in order to more efficiently anticipate increases in demand (Irshad, [0002]). By incorporating the teachings of Irshad, one would have been able to subdivide clusters into groupings that are sufficiently close (Irshad, [0042]).
Claim(s) 13 is rejected under 35 U.S.C. 103 as being unpatentable over Sawarkar et al. (US 20220138786 A1) in view of Wellmann et al. (US 20240221009 A1) in view of Rovatsos et al. (US 20240354810 A1) in view of Ichikawa et al. (US 20220148064 A1).
Regarding claim 13, the combination of Sawarkar, Wellmann, and Rovatsos t teaches all the limitations of claim 8 above.
However, Sawarkar does not explicitly teach wherein to determine the distance, the computing engine is to compute a root mean square distance between the first identifier and the second identifier.
From the same or similar field of endeavor, Ichikawa teaches wherein to determine the distance, the computing engine is to compute a root mean square distance between the first identifier and the second identifier ([0095] teaches determining the root mean square of the distances in the cluster, wherein the cluster is used for deciding the closeness of meaning between the words, wherein the cluster is an aggregated subset, wherein [0087] teaches aggregating product attributes into one or more subsets based on the closeness of meaning between the words indicating the product attributes from the input inter-word relationship information; see also: [0061, 0118, 0122, 0124]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar, Wellmann, and Rovatsos to incorporate the teachings of Ichikawa to include wherein to determine the distance, the computing engine is to compute a root mean square distance between the first identifier and the second identifier. One would have been motivated to do so in order to identifying the closeness in the meaning between words representing product attributes estimated by optimization through the use of root mean square distance (Ichikawa, [0093-0095]).
Claim(s) 17 is rejected under 35 U.S.C. 103 as being unpatentable over Sawarkar et al. (US 20220138786 A1) in view of Wellmann et al. (US 20240221009 A1) in view of Rovatsos et al. (US 20240354810 A1) in view of Deshpande et al. (US 20160092960 A1).
Regarding claim 17, the combination of Sawarkar, Wellmann, and Rovatsos teaches all the limitations of claim 16 above.
However, Sawarkar does not explicitly teach wherein generating identifiers comprises: converting names of each of the plurality of products in lowercase; removing punctuation and stopwords from the names of each of the plurality of products; and transforming the names of each of the plurality of products into the identifiers.
From the same or similar field of endeavor, Deshpande teaches wherein generating identifiers comprises: converting names of each of the plurality of products in lowercase ([0043] teaches the indexing module may index textual data by performing a text analysis where the analysis may convert the textual data into fundamental units of searching, which may be called text facets, wherein during analysis, the textual data may undergo multiple operations including extracting the words, removing common words, ignoring punctuation, reducing words to root form, and changing words to lower case, wherein after the text analysis is complete, the indexing module may add text facets to a product index, wherein the indexing module may store the indexed product textual data on the server; see also: [0022, 0039, 0064]);
removing punctuation and stopwords from the names of each of the plurality of products ([0043] teaches the indexing module may index textual data by performing a text analysis where the analysis may convert the textual data into fundamental units of searching, which may be called text facets, wherein during analysis, the textual data may undergo multiple operations including extracting the words, removing common words, ignoring punctuation, reducing words to root form, and changing words to lower case, wherein after the text analysis is complete, the indexing module may add text facets to a product index, wherein the indexing module may store the indexed product textual data on the server; see also: [0022, 0039, 0064]);
and transforming the names of each of the plurality of products into the identifiers ([0043] teaches the indexing module may index textual data by performing a text analysis where the analysis may convert the textual data into fundamental units of searching, which may be called text facets, wherein during analysis, the textual data may undergo multiple operations including extracting the words, removing common words, ignoring punctuation, reducing words to root form, and changing words to lower case, wherein after the text analysis is complete, the indexing module may add text facets to a product index, wherein the indexing module may store the indexed product textual data on the server; see also: [0022, 0039, 0064]).
It would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify the combination of Sawarkar, Wellmann, and Rovatsos to incorporate the teachings of Deshpande to include wherein generating identifiers comprises: converting names of each of the plurality of products in lowercase; removing punctuation and stopwords from the names of each of the plurality of products; and transforming the names of each of the plurality of products into the identifiers. One would have been motivated to do so in order to aid in the effectiveness of product search through the use of fuzzy searching in the textual data (Deshpande, [0049-0050]). By incorporating the teachings of Deshpande, one would have been able to improve the accuracy of semantic matching methods by filtering through the matching terms associated with products (Deshpande, [0005-0006]).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure:
Vasudevan et al. (US 12340408 B2) discloses obtaining similar grocery products based on the product name or title of products
Sridhar et al. (US 20240257216 A1) discloses determining a respective product type name similarity score between an anchor product name vector and complementary product name vector
Roy et al. ("Attribute value generation from product title using language models." 2021) discloses generating word embeddings from pretrained BERT models from data including the product title
Akritidis et al. ("A self-verifying clustering approach to unsupervised matching of product titles." 2020) discloses matching products by their titles using clustering
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 Sara G Brown whose telephone number is (469)295-9145. The examiner can normally be reached M-F 8:00 am- 5:00 pm.
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 at (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.
/SARA GRACE BROWN/Primary Examiner, Art Unit 3625