DETAILED ACTION
Status of the Application
This Final Office Action is in response to Application Serial 17/977,690. In response to the Examiner’s action mail dated June 04, 2024, Applicant submitted arguments and amendments, mail dated September 04, 2024. Applicant amended claims 1, 13, and 19. Claims 1-20 are examined in light of 35 U.S.C. 101 and 35 U.S.C 103, see below.
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 . 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.
Information Disclosure Statement
An information disclosure statement (IDS) was not submitted for consideration.
Response to Arguments
Applicant's arguments filed September 04, 2025 have been fully considered but they are not persuasive. Applicant’s arguments are moot in view of the revised rejections. Applicant’s argument will be address herein below.
Response to Claim Rejections- 35 U.S.C. 101
On pages 8-12 of the Applicant’s 35 U.S.C. 101 arguments, the Applicant submits,
Technical Problem Being Solved. Applicant discuss current approaches for retail inventory management and computer implements optimization system is not flexible. Current approaches rely almost exclusively on predefined markdown policies. Solutions fail to capture sufficient data and process it optimally. Convenient systems fail to collect, assemble, and/or process the data in an optimal manner.
Examiner acknowledges Applicant’s statement of Technical Problem Being Solved.
Technical Solution Provided by the Claims. Applicant submits the claims recite a specific technical solution that integrates machine learning technology into retail markdown optimization in an unconventional manner. Applicant identifies amended language in claim 1, claim 13, and claim 19.
Examiner acknowledges Applicant’s amendments. The limitations of the claims are analyzed under 35 U.S.C. 101 and 35 U.S.C. 103, see below.
Alignment with PTO July 2025 AI Guidance. Applicant completed an analysis referring to USPTO’s July 2024 Guidance. Applicant identified Step 2A Prong 1 – Not Directed to Abstract Ideas. Applicant pointed to Example 39 “training a neural network” and bullet list similarities of specific operations. The claims recite specific rules for training and operating the machine learning model that go beyond implementing an abstract idea on a computer, similar to McRO. Step2A Prong 2- Practical Application: Applicant traverses even if judicial exceptions are present, the claims integrate them into a practical application by: Improving Computer Technology, Solving Technology-Specific Problems, and Specific Technical Improvements.
Examiner acknowledges Applicant’s statements and amendments. The limitations of the claims are analyzed under 35 U.S.C. 101, see below.
Step 2A Analysis – Claims Are Not Directed to Abstract Ideas
Applicant traverses at Step 2A Prong 1. The claims do not recite abstract ideas but rather specific technological process for retail markdown optimization. The claims recite: Specific data collection from retail systems; Specific machine learning model training and operation with explicit improvements; Specific feedback mechanisms and retraining triggers; Specific integration into retail workflows with concrete technological enhancements. At Step 2A Prong 2, Applicant submits: The additional elements integrate any abstract concepts into practical applications that improve retail technology, as demonstrated by the specific technological improvements recited in the amended claims.
Examiner respectfully disagrees with Applicant’s Step 2A prong one analysis. As discussed above, Applicant asserts the application acknowledges current approache(s) for retail inventory management and computer implements optimization system is not flexible. Current approaches rely almost exclusively on predefined markdown policies. Solutions fail to capture sufficient data and process it optimally. Convenient systems fail to collect, assemble, and/or process the data in an optimal manner. Examiner agrees the claims recite approaches for retail inventory management. Particularly, the claim 1 recite “identifying an item identifier for an item; deriving input features associated with the item identifier from item data obtained from retail systems of a retailer; providing the input features as input to a machine learning model (MLM); receiving a markdown prediction as output from the MLM, …. ; providing the markdown prediction for the item identifier …; maintaining metrics for each item based on its current spoilage, sales, and margin; initiating a feedback …”. The claims are identifying items, providing markdown prediction for the items, and maintaining metrics, which is commercial activities, and thus, the claims are certain methods of organizing human activity. Regarding the item identifier, as presented, the item identifier could be interpreted as a human identifying a tag on the products and making an assessment, and thus, the claims recite a mental concept – evaluation, observation, judgement. Regarding a machine learning model, as presented a machine learning model could be interpreted as a mathematical algorithm that is conducting calculations of metrics such as spoilage, sales, and markdown. A feedback loop points back to the interpretation of a human identifying products and making an assessment.
The amended claims are identifying an item identifier for an item; deriving input features associated with … item data obtained … a retailer; providing the input features as input …; receiving a markdown prediction as output …. ; providing the markdown prediction for the item identifier …; maintaining metrics for each item based on its current spoilage, sales, and margin; initiating a feedback…”, and thus, the claims recite certain methods of organizing human activity – commercial activities. The limitations reciting
initiating a feedback re-training session through a trainer when spoilage is increasing or sales and margins are decreasing; …; creating a feedback loop based on the markdown prediction and other markdown predictions for other items … to minimize corresponding item spoilage, to minimize corresponding item shrink, to maximize corresponding item sales, and to maximize corresponding item margins through modified predictions provided … following the retraining; and replacing static store markdown policies and discretionary managerial oversight with objective and data-driven information generated by the method to item markdowns on a micro level and to sales and margins of a particular store on a macro level: … is optimized and continuously retrained to improve its predicted values based on actual outcomes associated with marked-down are mental concept - evaluation and observation. The claims recite certain methods of organizing human activity and mental concepts, and thus, the claims are directed to an abstract concept at step 2A prong one.
At Step 2A Prong two the claims are not integrated into a practical application. The claims are improving retail inventory management – markdowns. Applicant argues the elements retail systems, machine learning model integrate the any abstract concepts into a practical application. However, as recited, the additional elements are applied to conduct the retail improvement, and thus, these additional elements are not integrated into the abstract concepts. See MPEP 2106.05 (f).
Step 2B Analysis – Inventive Concept Present. Applicant traverses at Step 2B the claims provide significantly more than any abstract idea through: 1. Unconventional Technology Arrangement: The specific combination of real-time metric monitoring, automated retraining triggers, and feedback loops represents an unconventional arrangement not found in prior art.; 2. Specific Technical Improvements: As recited in the amended claims, the system provides concrete improvements, including "reducing item shrink, increasing item margin, and increasing item sales" and "replacing static store markdown policies and discretionary managerial oversight with objective and data-driven information."; 3. Non-Routine Implementation. The claimed system goes beyond routine computer implementation by providing specific technical solutions that enhance existing retailer systems through cloud-based integration.
Examiner respectfully disagrees. Applicant argues of real-time metric monitoring; however, the real-time is merely stated. A person reading a tag is reading data real-time. See Applicant specification [021] “… the input features may be based on actual observed information associated with previous markdowns on previous items within a given store …”. Applicant has not established an improvement rooted in technology at Step 2B.
Rebuttal to Examiner’s Remarks. The Examiner's characterization of the claims as merely "sales activities" and "mental concepts" fails to recognize the specific technological improvements and unconventional arrangements recited in the amended claims. The claims are not directed to fundamental business practices but to specific technological solutions that improve computer functionality in retail environments. The amended claims explicitly recite "wherein the MLM is optimized and continuously retrained to improve its predicted values based on actual outcomes associated with marked- down items," which demonstrates continuous technological improvement rather than static abstract concepts.
The Examiner's reliance on Example 47 is misplaced, as the present claims recite specific improvements to retail technology similar to the eligible claims in Examples 39 and other precedential cases. Accordingly, Applicant respectfully requests that these rejections be withdrawn.
Examiner respectfully disagrees with the Applicant’s 35 U.S.C. 101 Rebuttal. In light of 35 U.S.C. 101, the claims are directed to an abstract concept – improving retail inventory management. As discussed above, the claims use additional elements to conduct the abstract concept. See MPEP 2106.05 (f). The claims recite an improvement in the abstract concept. The additional elements identified in the claims are not integrated into the abstract concept. At step 2B, when considered as a whole the claims do not amount to significantly more. The claims do not recite an improvement that is rooted in technology. Instead, the claims are using technology to conduct the abstract concept.
Regarding the Applicant’s arguments of Subject Matter Eligibility Example 39 and Example 47. Because the Applicant uses machine learning models in the claims and machine learning models are disclosed in the specification, Applicant is pointed to the USPTO’s July 2024 Guidance Example 47.
The Rejection of Claims Under 112
Claims 1-20 were 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 pre-AIA the Applicant, regards as the invention. Applicant believes the above-noted amendments to claims 1, 13, and 19 now obviate these rejections.
Examiner acknowledges the Applicant’s amendments in view of the 112 Rejection. Applicant amended the claims to remove the “F1 predicted values” language. Therefore, the 35 U.S.C. 112 is withdrawn. Applicant is warned of 112(f) objection and 112(b) rejection particularly, “… identifying an item identifier for an item”.
Response to Claim Rejections- 35 U.S.C. 102
On pages 12-14 of the Applicant’s 35 U.S.C. 102 arguments the Applicant submits, Smith does not anticipate elements that are included in the amendments to the claims. Smith’s system focuses on customer specific discounts pricing based on expiration data and customer preferences, not on the comprehensive reduction and margin optimization systems.
Smith Claim 1 does not disclose: maintaining metrics for each item based on spoilage, sales, and margin; initiating feedback re-training session through a trainer; maintaining past features… with the MLM; static policies with the specific micro level approach; the MLM is optimized and continuously trained to improve its predicted values.
Smith Claim 13 does not anticipate: training specifically on “item spoilage, item shrink…”; “monitoring actual observed current item spoilage, current reshrink… for the current perishable item following the current predictions”; “reducing item shrink, increasing item margin, and increasing item sales for each item through predictive capabilities of the MLM”: the MLM is continuously retrained to improve its predicted values based on actual outcomes associated with marked-down-items”
Smith Claim 19 does not anticipate: then the system is provided as cloud-based service that enhances the retailer workflows, and services; comprehensive approach of training perishable item in the store; the MLM is optimized and continuously retrained to improve its predicted values; specific feedback mechanisms and retraining protocols ... enhance the retailer system.
The Applicant’s amendments necessitate grounds for a new rejection. See Prior Art rejection below. Smith teaches a computer system that is connected over network.
Examiner submits Smith is relevant art. Smith teaches a system that determines an expiration date or a target product available for purchase from a merchant. Additionally, in one or more embodiments, the disclosed system utilizes a machine-learning model to dynamically generate discount prices for the target product over time based on the expiration date., Smith [abstract]. Smith disclosed systems utilize a machine-learning model to generate discount prices for a particular customer based on customer data for the customer. Smith [007] The price management system can then determine an expiration date for a target product and utilize a machine-learning model trained using the historical data to generate a prediction of a sale of the target product at a discount price based on the expiration date. Smith [020]. Examiner submits Smith teaches the concept of inventory management, metrics including sales, product expiration, and price management. Smith [Figure 2]. Applicant’s arguments are not persuasive.
The Rejection of Claims Under 35 U.S.C. 103
Applicant traverses claims 3-4, 9, 11, and 18 were rejected under 35 U.S.C. § 103 over Smith Kevin (U.S. 2019/0272557) in view of Brooks (U.S. 2019/0130346). Based on the amendments and remarks presented above with the corresponding independent claims to these rejected dependent claims, Applicant believes that these rejected dependent claims are now in condition for allowance and respectfully requests an indication of the same from the Examiner.
Examiner respectfully disagrees with the Applicant’s 35 U.S.C. 103 arguments. The Applicant’s amendments necessitate grounds for a new rejection. See Prior Art rejection below.
Examiner submits Smith teaches the concept of inventory management, metrics including sales, product expiration, and price management and applies a machine learning model. Smith [Figure 2]. Applicant’s arguments are not persuasive. Brooks disclosures provide a computer-implemented method for perpetual inventory reconciliation. A perpetual inventory (PI) analysis component analyzes item data associated with a plurality of seasonal items using a set of PI purge criteria to identify a set of inactive inventory items. The plurality of seasonal items includes items associated with an inventory area having a positive PI value for a per-item threshold time after an end-of-season (EOS) date., Brooks [004], [030]. Brooks implements machine learning that generates output based on analysis of historical transaction data (e.g., sales trends and inventory levels)., Brooks [062], [042], [086], [087], [Figure 4]. Therefore, it would be obvious to combine before the effective filing date the concept of inventory management, metrics including sales, product expiration, and price management, as taught by Smith, with analyz[ing] item data associated with a plurality of seasonal items, as taught by Brooks, to analyze sensor data obtained from a set of sensor devices associated with the inventory area to calculate a physical inventory value representing a number of physical instances of the selected item within the inventory area., Brooks [004].
Response to Amendment
Claims 1-20 stand pending in this application. Applicant amended claims 1, 13, 19. The claims 1-20 are examined below and are pending in this application.
Regarding the 35 U.S.C. 112 Rejection, the Applicant amended the claims to remove the “F1 predicted values” language. Therefore, the 35 U.S.C. 112 is withdrawn. Applicant is warned of 112(f) objection and 112(b) rejection particularly, “… identifying an item identifier for an item”.
Regarding the 35 U.S.C. 101 rejections of claims 1-20. The pending claims have been fully considered in light of the 2019 Revised PEG Guidance. The Applicant’s arguments are not persuasive.
Regarding the prior art rejections of claims 1-20. The pending claims have claims are examined. See prior art rejection(s), below.
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 1-12,18 are process.
Claims 13-17 are process.
Claims 19-20 are machine.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
The claim 1 recite, “ identifying … for an item; deriving input features associated with … from item data obtained … a retailer; providing the input features as input to…model; receiving a markdown prediction as output … model, wherein the markdown prediction indicates whether the item should or should not be marked down; providing the markdown prediction for the item identifier… ; maintaining metrics for each item based on its current spoilage, sales, and margin; initiating a feedback re-training session through … when spoilage is increasing or sales and margins are decreasing; wherein … model maintains past input features associated with item data for a given store that were identified subsequent to a last training session …; and creating a feedback loop based on the markdown prediction and other markdown predictions for other items by retraining … model to minimize corresponding item spoilage, to minimize corresponding item shrink, to maximize corresponding item sales, and to maximize corresponding item margins through modified predictions provided … following the retraining; and replacing static store markdown policies and discretionary managerial oversight with objective and data-driven information generated by the method to item markdowns on a micro level and to sales and margins of a particular store on a macro level: wherein … model is optimized and continuously retrained to improve its predicted values based on actual outcomes associated with marked-down items.” Claims 1, in view of the claim limitations, recite the abstract idea of, … deriving input features associated with … item data obtained from … a retailer; providing the input features as input to a … model; receiving a markdown prediction…; providing the markdown prediction for the item identifier …; maintaining metrics for each item based on its current spoilage, sales, and margin; initiating a feedback re-training session … when spoilage is increasing or sales and margins are decreasing, which is certain methods of organizing human activity – commercial or legal interactions. Further, the claims recite initiating a feedback re-training session … when spoilage is increasing or sales and margins are decreasing; …; and creating a feedback loop based on the markdown prediction and other markdown predictions …, to minimize corresponding item shrink, to maximize corresponding item sales, and to maximize corresponding item margins through modified predictions, and thus, the claims are mental concept- evaluation, observation and judgement.
Claim 13 recites, “… training … model on input features associated with item spoilage, item shrink, item markdowns, item markdown levels, item sales, and item margins for perishable items of a store to generate markdown predictions and markdown parameters as output, each markdown prediction indicating whether a given perishable item should or should not be marked down, and the markdown parameters indicating, for each perishable item indicated for markdown, a markdown level and a quantity of the perishable item that is to be marked down; receiving an item identifier for a current perishable item during processing of a markdown workflow on a store device associated with the store; providing current input features for the item identifier as input to … model receiving current predictions as output from the … model for the item identifier; integrating the current predictions into the markdown workflow; maintaining metrics for each item based on its current spoilage, sales, and margin; initiating a feedback re-training session through … when spoilage is increasing or sales and margins are decreasing; wherein … maintains past input features associated with item data for a given store that were identified subsequent to a last training session with the … model; and monitoring actual observed current item spoilage, current item shrink, and current item sales/margins for the current perishable item following the current predictions and re-training the … model based thereon to optimize the … model to provide different predictions optimized to reduce item spoilage rates, reduce item shrink rates, and increase item sales and correspondingly item margins;_ and reducing item shrink, increasing item margin, and increasing item sales for each item through predictive capabilities of the … model; wherein the … model is optimized and continuously retrained to improve its predicted values based on actual outcomes associated with marked-down items.” Claim 13 recites training a model with metric data that is used to optimize sales, and thus, the claims are certain methods or organizing human activity – commercial activities.
Claim 19 recites, “… training, per perishable item of a store, … model on data relevant to item spoilage, item shrink, item margin, and item profit to produce as output markdown predictions as to whether the perishable item should or should not be marked down, and markdown parameters if the perishable item is indicated for markdown, the markdown parameters including a markdown level for the perishable item and a quantity of the perishable item to markdown; updating the data daily from retailer systems of a retailer to maintain current data for each of the perishable items; obtaining current predictions provided by … model for the current data in response to receiving perishable item identifiers for current perishable items from a workflow associated with item markdowns of a store; integrating the current predictions into a user interface associated with the workflow; maintaining metrics for each item based on its current spoilage, sales, and margin; initiating a feedback re-training session through … when spoilage is increasing and/or sales and margins are decreasing; wherein … maintains past input features associated with item data for a given store that were identified subsequent to a last training session with the … model; and identifying actual observed item spoilage rates, item shrink rates, and item sales after providing the current predictions, and using the item spoilage rates, the item shrink rates, and the item sales as feedback to initiate a training session with the … model to optimize the … model to provide different markdown predictions optimized to reduce the item spoilage rates, reduce the item shrink rates, and increase the item sales; wherein the … model is optimized and continuously retrained to improve its predicted values based on actual outcomes associated with marked-down marked down items; wherein … that enhances existing retailer systems, workflows, and services. Claim 19 recites training a model with metric data , integrate the data into workflows to retrain the model that is used to optimize sales and thus, the claims are certain methods or organizing human activity – commercial activities.
Claims 1-20, in view of the claim limitations, are recite the abstract idea of, … training a model with metric data that is used to optimize sales, which is certain methods of organizing human activity – commercial or legal interactions.
Furthermore, the claim 1 the claims recite initiating a feedback re-training session … when spoilage is increasing or sales and margins are decreasing; …; and creating a feedback loop based on the markdown prediction and other markdown predictions …, to minimize corresponding item shrink, to maximize corresponding item sales, and to maximize corresponding item margins through modified predictions, and thus, the claims are mental concept- evaluation, observation and judgement. Such concepts could be performed in the human mind (including observation, evaluation, judgement, opinion), and thus, the claims fall within the mental processes.
Accordingly, the claims are directed to certain methods and organizing human activity and mental concepts, and thus, the claims are directed to an abstract idea under the first prong of Step 2A.
This judicial exception is not integrated into a practical application under the second prong of Step 2A. In particular, the claims recite the additional elements beyond the recited abstract idea of, “an item identifier”, “…from retail systems of a retailer”, “a machine learning model (MLM)”, “a trainer”, in claim 1; “a machine learning model (MLM)”, “a store device associated with the store”, “a trainer”, in claim 13; “A system, comprising: a cloud server comprising at least one processor and a non-transitory computer-readable storage medium, the non-transitory computer-readable storage medium comprising executable instructions, wherein the executable instructions, when executed by the at least one processor cause the at least one processor to perform operations comprising:”, “a machine learning model (MLM), “from retailer systems of a retailer”, “a trainer”, in claim 19; however, when viewed as an ordered combination, and pursuant to the broadest reasonable interpretation, each of the additional elements are computing elements recite adding the words “apply it” (or an equivalent) with the judicial exception, or mere instructions to implement an abstract idea on a computer or merely uses a computer as a tool to perform an abstract idea – see MPEP 2106.05 (f)
The dependent claims further cite the additional elements:
Claim 8: “at least one of a retail inventory system”
Claim 20: “a store inventory system”, “a store transaction system”, “a store forecasting system”.
Accordingly, the additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea. The claims also fail to recite any improvements to another technology or technical field, improvements to the functioning of the computer itself.
Regarding the additional elements of the “an item identifier” & “a trainer”, Applicant is pointed to the specification. Applicant is encouraged to positively recite the element as technology. See Applicant’s specification [032], [034] & [021]-[023]. Otherwise, such elements under the broadest reasonable interpretation could be considered actions.
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception. As discussed above, the additional elements when considered both individually and as an ordered combination do not amount to significantly more than the abstract idea. As discussed above with respect to integration of the abstract idea into a practical application, the additional elements of by the one or more processors, on the graphical user interface, the display is/are insufficient to amount to significantly more. -See MPEP 2106.05 (f).
At step 2B, it is MPEP 2106.05 (d) – Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information).
Dependent claims 2-12, 18 further narrow the abstract idea of independent claim 1. Dependent claims 14-17 further narrow the abstract idea of independent claim 13. Dependent claim 19 further narrow the abstract idea of independent claim 20. The claims 1-20 are not patent eligible.
Since there are no limitations in these claims that transform the exception into a patent eligible application such that these claims amount to significantly more than the exception itself, claims 1-20 are rejected under 35 U.S.C. 101 as being directed to non-statutory subject matter.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
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.
Claim(s) 1,2, 5, 6, 7, 8, 10, 12, 13, 14, 15, 16, 17, 19 & 20 /are rejected under 35 U.S.C. 103 as being unpatentable over Smith (US 2019/0,272,557 A1) in view of Lange (US 2012/0,278,154 A1).
Regarding Claim 1, (Currently Amended)
A method, comprising: identifying an item identifier for an item; deriving input features associated with the item identifier from item data obtained from retail systems of a retailer;
Smith [065] discloses the price management system 102 can perform the act 302 of identifying product data. In particular, the price management system 102 can identify product data for a target product by accessing a repository that includes information about products available for purchase from a merchant. Identifying product data for the target product includes the price management system 102 determining an expiration date of the target product. , Smith [065], [Figure 3].
Smith [020] discloses the price management system can provide a discount price to a customer via one or more other devices, such as a digital price tag, scanner device, checkout device, or other device capable of providing digital notifications to a customer., Smith [020], [021]
providing the input features as input to a machine learning model (MLM); receiving a markdown prediction as output from the MLM, wherein the markdown prediction indicates whether the item should or should not be marked down; providing the markdown prediction for the item identifier to a retail device;
Smith [019] teaches a dynamic price management system that dynamically generates digital product notifications reflecting discounts over time based on product expiration data. ,
Smith [020] discloses the dynamic price management system (or simply “price management system”) analyzes historical data, including historical product expiration data, for products previously available for purchase from a merchant and for customers associated with the merchant to determine correspondences between product details, product expiration data, and customer preferences. The price management system can then determine an expiration date for a target product and utilize a machine-learning model trained using the historical data to generate a prediction of a sale of the target product at a discount price based on the expiration date,;
Smith [021] discloses the price management system can analyze product history data (including historical expiration data) and customer history data to generate and provide discount prices for a product based on an expiration date of the product. Specifically, in one or more embodiments, the price management system trains a machine-learning model using product history data for previously available products and customer history data for a plurality of customers associated with a merchant. The price management system can utilize the machine-learning model (e.g., neural network or regression model) to output sale predictions for a product with a particular expiration date and then train the machine-learning model based on a comparison between the output sale predictions and ground truth sales information.;
Smith [022] … the price management system can utilize the trained machine-learning model to generate probabilities of a customer purchasing a target product with an expiration date at a specific price. Smith [020], [021], [022], [Figure 2]
Smith [087] discloses the price management system 102 can also predict losses associated with the possible discount prices based on the probabilities.
maintaining metrics for each item based on its current spoilage, sales, and margin;
Smith [022] discloses the price management system can utilize the trained machine-learning model to generate probabilities of a customer purchasing a target product with an expiration date at a specific price ... the price management system identifies product data for a target product (e.g., an expiration date of the target product and other features of the target product) and provides the product data as input to the trained machine-learning model. , Smith [022], [Figure 2], [Figure 3]
Smith [084] discloses the price management system 102 can also generate, with each probability, a predicted loss to the merchant for the discount price. The loss can be based on an original cost of the target product to the merchant and the price sold.
initiating a feedback re-training session through a trainer when spoilage is increasing or sales and margins are decreasing; wherein the trainer maintains past input features associated with item data for a given store that were identified subsequent to a last training session with the MLM;
Smith [021] discloses the price management system can utilize the machine-learning model (e.g., neural network or regression model) to output sale predictions for a product with a particular expiration date and then train the machine-learning model based on a comparison between the output sale predictions and ground truth sales information.; Smith [021], [095]
Smith [075] discloses as shown in FIG. 3, the machine-learning model uses the product data and the customer data to generate probabilities that the product will sell at a plurality of different prices. Specifically, a trained machine-learning model can apply the learned relationships between product history data and customer history data to the product data and customer data of a target product. For example, the machine-learning model analyzes and processes the product data for the target product and customer data using algorithms and functions of the model that the price management system 102 tuned in relation to FIG. 2.
Smith [079] discloses the price management system 102 can identify an original price of the target product from the product data and then analyze a plurality of possible discount prices based on the original price. The price management system 102 can select the possible discount prices by selecting common price points (e.g., historical prices of products in the product category), specific increments of price points lower than the original price, or other criteria for minimizing the number of possible discount prices to analyze., [075],[079], [Figure 2],[Figure 3]
Examiner submits Smith teaches retraining. Neural networks are particularly effective for tasks involving complex data or pattern recognition. Neural network functions include forward propagation, loss calculation, back propagation and iteration. Smith teaches a trained machine-learning model can apply the learned relationships between product history data and customer history data to the product data and customer data of a target product., Smith [075].
and creating a feedback loop based on the markdown prediction and other markdown predictions for other items by retraining the MLM to minimize corresponding item spoilage, to minimize corresponding item shrink, to maximize corresponding item sales, and to maximize corresponding item margins through modified predictions provided by the MLM following the retraining;
…..
wherein the MLM is optimized and continuously retrained to improve its predicted values based on actual outcomes associated with marked-down items.
Smith [021], [095] and
Smith [081] discloses the machine-learning model can learn that customers with larger families who have a higher purchasing frequency of milk are more likely to be interested in purchasing a gallon of milk that is closer to expiration than customers with smaller families and lower purchasing frequency of milk. The machine-learning model can use that information, to determine a probability that a gallon of milk with a given expiration date is likely to sell at a possible discount price. Additionally, the machine-learning model can use information about demographics of the merchant (e.g., whether the merchant has a higher number of customers with large families than customers with small families) to determine the probability.
Smith [0086] As shown in FIG. 3, the price management system also performs the act 310 of determining a discount price for the target product. Specifically, the price management system 102 can use the probabilities and loss values to determine a discount price for the target product. The price management system 102 can thus attempt to minimize actual losses incurred by the merchant for listing the target product at discount price according to the probability of selling the target product at the discount price., Smith [086], [Figure 3].
Smith [088] discloses the price management system 102 can then use the probability for each discount price to determine a loss that is based on the cost, probability, and discount price. Based on the losses for the discount prices, the price management system 102 can select a discount price for the target product that minimizes the loss to the merchant (e.g., a discount price of 60% in Table 1 with a loss of $200).
Smith [095] discloses the price management system 102 can update a pricing model for a target product in response to receiving additional product data or customer data. For instance, if a customer's purchase habits change during the duration of the pricing model (e.g., prior to the expiration date of the target product), the price management system 102 can input the new customer data into the machine-learning model to update the probabilities and select new discount price(s), if applicable. Similarly, if the product data changes (e.g., available inventory, trends related to the product category or similar product categories), the price management system 102 can use the new product data to update the probabilities and discount price(s).
Although highly suggested in Smith, Examiner relies on Lange to teach:
And replacing static store markdown policies and discretionary managerial oversight with objective and data-driven information generated by the method to item markdowns on a micro level and to sales and margins of a particular store on a macro level:
Lange teaches historical information regarding sale of a merchant's inventory may be employed to assist in identifying inventory items to market to customers and/or to assist in setting offer prices for marketed inventory items. Analysis of such historical information may provide an indication of likelihood of spoilage for inventory. This analysis could be used in both the manual and automatic approaches discussed above for identifying inventory items to market. With reference to FIG. 6, a flow diagram is provided that illustrates a method 600 for using historical information to identify likelihood of spoilage of inventory and using the information for purposes of marketing inventory in accordance with an embodiment of the present invention. As shown at block 602, historical inventory information for a merchant is accessed. This may include information regarding the sale of inventory, including, for instance, what inventory was sold and when the inventory was sold. In some embodiments, historical inventory information from similar merchants may be accessed in addition to or in lieu of information from the merchant. This information from similar merchants may provide a trend for the kind of goods or services offered by the merchants., Lange [064], [Figure 6]. Lange teaches as shown in FIG. 2, an inventory marketing system 202 communicates with merchants, such as merchant 204, to identify inventory items, such as inventory item 206, that are expected to spoil and markets the inventory items to targeted customers, such as the customer 208. Customers may select to purchase the inventory items, and the inventory marketing system 202 may operate to facilitate the transactions., Lange [031], [Figure 2].
Smith discloses a system that determines an expiration date or a target product available for purchase from a merchant. Lange teaches identifying an inventory item likely to spoil. It would have been obvious to one of ordinary skill in the art to combine before the effective filing date, utilizing a machine-learning model to dynamically generate discount prices for the target product over time based on the expiration date, as taught by Smith, with identify inventory items that are likely to spoil and to market those inventory items to targeted customers, as taught by Lange, to avoid letting inventory go spoiled as the spoiled inventory results in loss of revenue to the merchants., Lange [001].
Regarding Claim 2, (Original)
The method of claim 1, wherein the markdown prediction indicates that the item should be marked down, and wherein the method further comprises: receiving, as further output from the MLM, markdown parameters including a markdown level for the item and an item quantity of the item to which to apply the markdown level; and providing the markdown parameters to the retail device.
See claim 1 Smith [021], [095], [081] disclose MLM., markdown and probability of an item selling.
Regarding Claim 5, (Original)
The method of claim 1, further comprising processing the method as a software- as-a-service to the retail systems or retail services.
Smith [0150] discloses a cloud-computing model can also expose various service models, such as, for example, Software as a Service (“SaaS”), Platform as a Service (“PaaS”), and Infrastructure as a Service (“IaaS”).;
Smith [0118] discloses Although the above embodiments describe providing discount prices to customer client devices, the price management system 102 can also provide discount prices to customers via other methods. In particular, the price management system 102 can implement dynamic price tags that provide dynamic discount prices at in-store locations., Smith [0150], [0118], [abstract], [006]
Examiner submits a checkout scanner or other device is a SAAS., Smith [072]
Regarding Claim 6, (Original)
The method of claim 1, wherein identifying further includes receiving the item identifier from a scan performed on a barcode of the item.
See Claim 1, Smith [020] – scanner and Smith [067] discloses Furthermore, the price management system can use checkout scanners that determine an expiration date of the target product and/or to verify products and pricing of the target product. Checkout scanners can also allow the price management system 102 to track products and understand customer behavior and product inventory practices., Smith [067], [072], [020]
Regarding Claim 7, (Original)
The method of claim 1, wherein identifying further includes receiving the item identifier from input received from a user at a user interface.
Smith [067] discloses the price management system 102 can use image-based analysis (e.g., using in-store cameras, cameras on the customer client devices, UV imaging) to identify the remaining shelf life of the target product based on the appearance of the target product (e.g., by identifying brown spots, fluorescence, or other signs of aging on perishable products such as fruit).
Smith [073] discloses the price management system 102 can determine that the customer is viewing a page or interface (e.g., within a client application associated with the merchant or within a webpage of a web browser) that includes information about the target product or product category. Similarly, the price management system 102 can identify the product in a list of products (e.g., a digital shopping list) of the customer.
Regarding Claim 8, (Original)
The method of claim 1, wherein deriving further includes obtaining the item data from at least one of a retail inventory system, a retail transaction system, or a retail forecasting system/service.
Smith [072] discloses the price management system 102 can detect that a customer has added a product to a shopping cart using a smart cart that reads an RFID tag, code, or other identifier on the product. In another instance, the price management system 102 can detect that a customer is checking out to purchase a product by scanning a tag or code of the product using a checkout scanner or other device. The price management system 102 can then provide, in real-time, information about one or more other products in the product category or related product categories based on the determined interest.
Within claim 8, Smith discloses a customer is checking out to purchase a product by scanning a tag or code of the product using a checkout scanner or other device, and thus, Smith discloses a retail inventory system, a retail transaction system. Claim 8 a "Markush" claim recites a list of alternatively useable members. In re Harnisch, 631 F.2d 716, 719-20, 206 USPQ 300, 303 (CCPA 1980); Ex parte Markush, 1925 Dec. Comm'r Pat. 126, 127 (1924). The listing of specified alternatives within a Markush claim is referred to as a Markush group or a Markush grouping. Abbott Labs v. Baxter Pharmaceutical Products, Inc., 334 F.3d 1274, 1280-81, 67 USPQ2d 1191, 1196 (Fed. Cir. 2003) (citing to several sources that describe Markush groups)- See MPEP 706.03.
Regarding Claim 10, (Original)
The method of claim 1, wherein receiving further includes identifying actual observed item spoilage rates, item shrink rates, and item sales after providing the predictions.
Brooks [093] discloses FIG. 4 illustrates a graph diagram of a pricing model 400 for a target product, indicating a correlation between the change in price (indicated by the dashed line 402) of the target product over time and the incentive to buy (indicated by the solid line 404). As illustrated, by changing a discount price of the target product over time, the price management system 102 can maintain, or even increase, a customer's incentive to purchase the product. In particular, by utilizing a machine-learning model trained on historical information for products and customers, the price management system 102 can determine pricing models that change the price to fit the customers' desire to purchase the target product according to the purchase habits of the customers., Brooks [093], [Figure 4]
And Smith [095], [021] – machine learning , Smith [056] – ground truth.
Regarding Claim 12, (Original)
The method of claim 1, wherein providing further includes providing the markdown prediction for the item identifier to a user within a user interface during a markdown workflow being processed on a retail device.
Smith [073] and
Smith Figure [Fig 5B], [0102] discloses For example, as illustrated in FIG. 5A, the discount Interface 502a includes a plurality of products available for purchase from the merchant. The price management system 102 can select the discount prices to display to the customer in response to determining that the customer is likely to be interested in the target products. The price management system 102 can thus determine which products, and in which order, to present within the discount interface 502a based on the customer's interest score.
Regarding Claim 13, (Currently Amended)
A method, comprising: training a machine learning model (MLM) on input features associated with item spoilage, item shrink, item markdowns, item