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 .
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 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more. The claims recite a mental process of observation, judgement and evaluation. This judicial exception is not integrated into a practical application nor does it amount to significantly more because the additional elements of the claims amounts to mere extra-solution activity in combination with generic computer hardware to implement the abstract idea. See the analysis below.
Claims 1, 10 and 19
Step 1: The claims recites a non-transitory computer readable medium, method, and system therefore, it falls into the statutory categories.
Step 2A Prong 1: The claim recites, inter alia:
Generating training data sets at least by: comparing a plurality of set of time-series data to a set of feature-based rules, wherein each feature-based rules in the set of feature-based rules is mapped to a respective classification; based on detecting a match between a particular feature-based rule and one or more life cycle-based features of a particular set of life-cycle-based features extracted from a particular set of time-series data: labelling the particular set of time-series data with a particular life-cycle classification mapped to the particular feature-based rule; (This is a mental process of observation, evaluation and judgment wherein a user compares timer-series data and features to feature based rules to see if the features of the time-series matches the rules. If the time-series features matches the rules then time-series is labelled with the lifecycle class assigned to a rule. This can be done with the aid of paper. For example, a user reviews times-series data, such as historical transaction data for a product. The user determines features of the data such as average price, trends, etc. Then compares that features to see if any rules have those features. For example, a rule might say if sales of a product are steady for over 3 months then product is mature. The user would compare any sales trends in the sales data to the rule, and if a time-series has a trend that matches it is labelled as mature.)
Step 2A Prong 2:
This judicial exception is no integrated into a practical application. Aside from the limitations above, the claim recites:
A non-transitory computer readable medium comprising instructions which, when executed by one or more hardware processors, causes performance of operations comprising; a processor and memory (claim 19); (These limitations amount to using generic computer hardware to execute the abstract idea, see MPEP 2106.05(f).)
training a machine learning model to generate life-cycle classifications for product-store pairs and training the machine learning model based on the training data sets; applying the machine learning model to the target set of time-series data to generate a particular life-cycle classification for the target product-store pair. (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) – Examiner’s note: high level recitation of training a machine learning model with previously determined data and applying a machine learning model to data);
obtaining the training data sets, each training data set comprising: historical time-series data for a respective product-store pair and a life-cycle classification, corresponding to a life-cycle type, for the historical time-series data; receiving a target set of time-series data for a target product-store pair; and (This amount to receiving data, which is data collection, and as such extra-solution activity, see MPEP 2106.05(g).)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are mere insignificant extra solution activity in combination of generic computer hardware performing generic functions that are implemented to perform the disclosed abstract idea above.
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional elements of “obtaining the training data sets, each training data set comprising: historical time-series data for a respective product-store pair and a life-cycle classification, corresponding to a life-cycle type, for the historical time-series data; receiving a target set of time-series data for a target product-store pair;” amount to transmitting data and is well-understood, routine and conventional and does not amount to significantly more. See MPEP 2106.06(d)(II) wherein it cites “The courts have recognized the following computer functions as well‐understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity. i. Receiving or transmitting data over a network, e.g., using the Internet to gather data”. The limitations of “training a machine learning model to generate life-cycle classifications for product-store pairs and training the machine learning model based on the training data sets; applying the machine learning model to the target set of time-series data to generate a particular life-cycle classification for the target product-store pair.” amounts to using machine learning as tool to apply an abstract idea, see MPEP 2106.05(f), and linking the abstract idea to technological field, see MPEP 2106.05(h). The additional limitations of “A non-transitory computer readable medium comprising instructions which, when executed by one or more hardware processors, causes performance of operations comprising; a processor and memory;” amount to using generic computer hardware to execute the abstract idea, see MPEP 2106.05(f).)
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are well, understood, routine and conventional activity as disclosed in combination of generic computer hardware performing generic functions that are implemented to perform the disclosed abstract idea above.
Claims 2, 11 and 20
Step 2A Prong 1: The claim recites, inter alia:
extracting a set of life-cycle-based features from each set of time-series data of the plurality of sets of time-series data by comparing characteristics of the time-series data to particular features to determine whether the features are present in the time-series data. (This is a mental step of observation, evaluation and judgment wherein a user compares time-series data to particular features to see if they are present in the data, can be done with the aid of pen and paper.)
Step 2A Prong 2:
This judicial exception is no integrated into a practical application. Aside from the limitations above, the claim recites:
accessing the plurality of sets of time-series data corresponding to sales of one or more products from a plurality of stores; (This is data collection which is extra-solution activity see MPEP 2106.05(g)
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are mere insignificant extra solution activity that are implemented to perform the disclosed abstract idea above.
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional elements of “accessing the plurality of sets of time-series data corresponding to sales of one or more products from a plurality of stores;” amount to data collection as it is accessing data from memory and is well-understood, routine and conventional and does not amount to significantly more. See MPEP 2106.06(d)(II) wherein it cites “The courts have recognized the following computer functions as well‐understood, routine, and conventional functions when they are claimed in a merely generic manner (e.g., at a high level of generality) or as insignificant extra-solution activity. iv. Storing and retrieving information in memory, Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015); OIP Techs., 788 F.3d at 1363, 115 USPQ2d at 1092-93;”.
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are well, understood, routine and conventional activity as disclosed that are implemented to perform the disclosed abstract idea above.
Claims 3 and 12
Step 2A Prong 1: The claim recites, inter alia:
comparing, in a predetermined sequence, the particular set of time-series data to a respective feature-based rule among the set of feature-based rules; and based on detecting the match between the particular feature-based rule and the particular set of time-series data: refraining from comparing the particular set of time-series data to any additional feature-based rules among the set of feature-based rules. (This is a mental step of observation, evaluation and judgment wherein a user compares time-series data to feature-base rules and when the time-series is matched to a rule, the use stop looking for matches.)
Step 2A Prong 2:
This judicial exception is no integrated into a practical application. Aside from the limitations above, the claim recites: There are not additional elements in the claim.
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. There no additional elements.
Claims 4 and 13
Step 2A Prong 1: The claim recites, inter alia:
Claims 4 and 13 inherit the mental processes in claims 1 and 10.
Step 2A Prong 2:
This judicial exception is no integrated into a practical application. Aside from the limitations above, the claim recites: There are additional elements in the claim are: “wherein the machine learning model is a random forest classifier model”. This limitation is cited at high level of generality and result in using the machine learning model as a tool to implement the abstract idea, see MPEP 2106.05(f).
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are generic computer hardware performing generic functions that are implemented to perform the disclosed abstract idea above.
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional elements are “wherein the machine learning model is a random forest classifier model”, this limitation is cited at high level of generality and result in using the machine learning model as a tool to implement the abstract idea, see MPEP 2106.05(f).
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they generic computer hardware used to implemented to perform the disclosed abstract idea above.
Claims 5 and 14
Step 2A Prong 1: The claim recites, inter alia:
Claims 5 and 14 inherit the mental processes in claims 1 and 10.
Step 2A Prong 2:
This judicial exception is no integrated into a practical application. Aside from the limitations above, the claim recites: There are additional elements in the claim are: “wherein the operations further comprise: applying the machine learning model to a second set of time-series data associated with a second product-store pair to generate a second life-cycle classification for the second product-store pair, wherein the second life-cycle classification is a long-life-cycle classification.”. This limitation is cited at high level of generality and result in using the machine learning model as a tool to implement the abstract idea, see MPEP 2106.05(f).
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are generic computer hardware performing generic functions that are implemented to perform the disclosed abstract idea above.
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional elements are “wherein the operations further comprise: applying the machine learning model to a second set of time-series data associated with a second product-store pair to generate a second life-cycle classification for the second product-store pair, wherein the second life-cycle classification is a long-life-cycle classification.”, this limitation is cited at high level of generality and result in using the machine learning model as a tool to implement the abstract idea, see MPEP 2106.05(f).
Claims 6 and 15
Step 2A Prong 1: The claim recites, inter alia:
Claims 6 and 15 inherit the mental processes in claims 1 and 10.
Step 2A Prong 2:
This judicial exception is no integrated into a practical application. Aside from the limitations above, the claim recites: There are additional elements in the claim are: ““wherein the target product-store pair corresponds to a first product sold from a first store, and wherein the second product-store pair corresponds to the first product sold from a second store.”. This limitation amounts to use particular type of data to be used and manipulate and is extra-solution activity, see MPEP 2106.05(g) and linking to a particular field of use, sales data, see MPEP 2106.05(h).
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are extra-solution activity in combination with the disclosed abstract idea above.
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional elements are “wherein the target product-store pair corresponds to a first product sold from a first store, and wherein the second product-store pair corresponds to the first product sold from a second store.”, this limitation does not amount to significantly more than the judicial exception as it simply linking the abstract idea to particular field of use, see MPEP 2106.05(h).
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they are extra-solution activity that links the abstract to particular field of use.
Claims 7 and 16
Step 2A Prong 1: The claim recites, inter alia:
determining the target product-store pair corresponds to the short-life cycle classification; determining the second product-store pair corresponds to the long-life cycle classification; (These element amount to a mental process of observation, judgement and evaluation wherein a user transaction data related to a target product-store pair is a short-life cycle such as seasonal sales and second product store-pair corresponds to long-life cycle which sales for long period of times.)
Step 2A Prong 2:
This judicial exception is no integrated into a practical application. Aside from the limitations above, the claim recites: There are additional elements in the claim are: “applying a first forecasting model to a third set of time-series data associated with the target product-store pair to generate a first forecast for the target product-store pair; applying a second forecasting model to a fourth set of time-series data associated with the second product-store pair to generate a second forecast for the second product-store pair”, these limitation amount to 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). Examiner’s note: high level recitation of using a machine learning model with previously determined data.
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are machine learning as generic tools to implement the abstract idea disclosed above.
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional elements are “applying a first forecasting model to a third set of time-series data associated with the target product-store pair to generate a first forecast for the target product-store pair; applying a second forecasting model to a fourth set of time-series data associated with the second product-store pair to generate a second forecast for the second product-store pair”, these limitation amount to 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). Examiner’s note: high level recitation of using a machine learning model with previously determined data.
The additional elements as disclosed above in combination of the abstract idea are not sufficient to amount to significantly more than the judicial exception as they generic tools to implement the abstract idea disclosed above.
Claims 8 and 17
Step 2A Prong 1: The claim recites, inter alia:
Claims 8 and 17 inherit the mental processes in claims 7 and 16.
Step 2A Prong 2:
This judicial exception is no integrated into a practical application. Aside from the limitations above, the claim recites: There are additional elements in the claim are: “applying the machine learning model to a fifth set of time-series data associated with a third product-store pair to generate a third life-cycle classification for the third product-store pair, wherein the third life-cycle classification is an inconclusive-type classification.”. This limitation is cited at high level of generality and result in using the machine learning model as a tool to implement the abstract idea, see MPEP 2106.05(f).
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are generic computer hardware performing generic functions that are implemented to perform the disclosed abstract idea above.
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional elements are “applying the machine learning model to a fifth set of time-series data associated with a third product-store pair to generate a third life-cycle classification for the third product-store pair, wherein the third life-cycle classification is an inconclusive-type classification.”, this limitation is cited at high level of generality and result in using the machine learning model as a tool to implement the abstract idea, see MPEP 2106.05(f).
The additional elements as disclosed above alone or in combination do not amount to significantly more than the judicial exception as they are generic computer hardware performing generic functions that are implemented to perform the disclosed abstract idea above.
Claims 9 and 18
Step 2A Prong 1: The claim recites, inter alia:
Determining the third product-store pair corresponds to the inconclusive-type classification; (This is a mental process of observation, judgement and evaluation wherein a user evaluate the product-store pair sales data and determine there is not enough information to classify the data.)
comparing attributes of the third product-store pair to attributes of at least one of the target product-store pair and the second product-store pair to assign an interim classification to the third product-store pair, wherein the interim classification corresponds to one of the long-life-cycle classification and the short-life-cycle classification; and (This amounts to a mental process of observation, evaluation and judgement wherein a user compares the inconclusive product-store pair attributes to a target and second product-store pair that have been classified, and based on it being similar to one of the two, assigning the class of the matching product-store pair to the inconclusive product-store pair.)
Step 2A Prong 2:
This judicial exception is no integrated into a practical application. Aside from the limitations above, the claim recites: There are additional elements in the claim are: “applying one of the first forecasting model and the second forecasting model to a sixth set of time-series data associated with the third product-store pair to generate a third forecast for the third product-store pair.”. This limitation is cited at high level of generality and result in using the machine learning model as a tool to implement the abstract idea, see MPEP 2106.05(f).
The additional elements as disclosed above alone or in combination do not integrate the judicial exception into practical application as they are generic computer hardware performing generic functions that are implemented to perform the disclosed abstract idea above.
Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception. The additional elements are “applying one of the first forecasting model and the second forecasting model to a sixth set of time-series data associated with the third product-store pair to generate a third forecast for the third product-store pair.”, this limitation is cited at high level of generality and result in using the machine learning model as a tool to implement the abstract idea, see MPEP 2106.05(f).
The additional elements as disclosed above alone or in combination do not amount to significantly more than the judicial exception as they are generic computer hardware performing generic functions that are implemented to perform the disclosed abstract idea above.
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 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.
Claims 1-2, 5-11 and 14-20 are rejected under 35 U.S.C. 103 as being unpatentable over Ratner et al. (“Data Programming: Creating Large Training Sets, Quickly” – hereinafter Ratner) in view of Li et al. (US2020/0143246 A1 – hereinafter Li) and further in view of Kamarchik et al. (US 2022/0245572 A1 – hereinafter Kamarchik)
In regards to claim 1, Ratner discloses a non-transitory computer readable medium comprising instructions which, when executed by one or more hardware processors, causes performance of operations comprising:
generating training data sets at least by: (Ratner page 1 section 1 second paragraph cites “To help reduce the cost of training set creation, we propose data programming, a paradigm for the programmatic creation and modeling of training datasets.”)
comparing a plurality data to a set of feature-based rules, wherein each feature-based rule in the set of feature-based rules is mapped to a respective classification; (Ratner page 3-4 section 3 example 3.1 teaches using labeling functions (feature based rules) to label data, it also refers to figure 1 wherein label functions have features and are labelled as lamda 1-3 if the data matches the rules.)
based on detecting a match between a particular feature-based rule and one or more features of a training data features extracted from a particular set of data: labeling the particular set of data with a particular classification mapped to the particular feature-based rule; (Ratner page 3-4 section 3 example 3.1 teaches using labeling functions (feature based rules) to label data, it also refers to figure 1 wherein label functions have features and are labelled as lamda 1-3 if the data matches the rules.)
training a machine learning model to generate classifications for data, the training comprising: obtaining the training data sets, each training data set comprising: data; and a classification for the data; training the machine learning model based on the training data sets; (Ratner page 1 last paragraph teaches creating training data with labels using the labeling functions and then training a machine learning model using that data, wherein it cites “To address this, we model the labeling functions as a generative process, which lets us automatically denoise the resulting training set by learning the accuracies of the labeling functions along with their correlation structure. In turn, we use this model of the training set to optimize a stochastic version of the loss function of the discriminative model that we desire to train.” Also see page 4 “Noise-Aware Empirical Loss” first paragraph and equation 3, it teaches training a machine learning model using training data and labels to minimize the loss function. In the equation 3. S is training examples, f(x) is features, Y is labels, w is model parameters. Thus, the model is trained using the time-series data and labels.)
However Ratner does not explicitly disclose wherein in the data is time series data; wherein each feature-based rule is mapped to a life-cycle classification; labelling time-series data with a particular life-cycle classification; generating classifications for product-store pairs; using historical time-series data for a respective product-store pair and classifying it; receiving a target set of time-series data; and applying the machine learning model to the target set of time-series data to generate a particular life-cycle classification for the target product-store pair.
Li discloses using historical time-series data and comparing a plurality of sets of time-series data to a set of feature-based rules, wherein each feature-based rule in the set of feature-based rules is mapped to a respective classification; and time-series data being classified into lifecycle classifications. (Li abstract and para. [0004-0006] cites “a segmentation operation for categorizing the plurality of time series into a plurality of demand classes . Each time series in the plurality of time series is categorized into a particular demand class among the plurality of demand classes based on demand characteristics of the time series.”; para. [0212] teaches using historical time-series data wherein it cites “A time series can be categorized into the short class if the time series has a short record of historical data (e.g., fewer than a predefined number of historical data points).”; also see para. [0231] which cites “Because the demand class for each time series is determined based on the characteristics of the time series, there are demand classes that do not contain any time series.”. This teaches mapping/classifying time-series to a demand class based on characteristics (features) of the time-series. The characteristics of the time-series is the features of the instant claim, the mapped classification is the “demand classes”, and the feature-based rules matching is the characteristics of the time-series being compared to the demand classes to determine that’s where they belong. Also see para. [0212 and 0213] teaches timer-series data being classified in demand classes which include life cycle class, wherein it cites “As one example, an intermittent time series can represent the demand for software that is ending its life cycle and for which there are sporadic periods with no downloads.” and “The demand characteristics that are used to determine the demand class of a time series can include the demand life cycle,…”. Also Li disclose “a non-transitory computer readable medium comprising instructions which, when executed by one or more hardware processors” in claim 23 and a system comprising a processor and memory in claim 1.)
It would have been obvious to one of ordinary skill in the art before the earliest effective filing date of the claimed invention to modify the teachings of the Ratner with that Li in order to allow for using timer-series data, labelling/classifying the data into lifecycle classes, and associated each lifecycle class with a rule as both reference deal with using machine learning models and training data. Also Li para. [0212-0213] teaches what demand patterns are for retired, year-round-seasonal, year-round-non-seasonal, etc. It would obvious to one or ordinary skill to create rules for each lifecycle based on those demand patterns, and labeling/classifying the time-series data based on matching or having features set in the rules. The using of timer-series data in machine allows for capturing temporal dependencies, and that coupled with the training data creation and labeling of Ratner creates a system that capable of taking time-series data, capturing temporal patterns such as product demand trends, and forecasting demand in an accurate and efficient manner.
However, Ratner in view of Li does not explicitly disclose wherein the historical time-series data being product-store pairs; receiving a target set of time-series data; and applying the machine learning model to the target set of time-series data to generate a classification for the target time-series data.
Kamarchik discloses using historical time-series data being product-store pairs; (Kamarchik para. [0006, 0045, and 0060] teaches obtaining historical product sales and para. [0145] teaches sales for a product at each location, thus it teaches product-store pairs, so see para. [0180] that teaches transaction data is sales data of a product including date of purchase, quantity, price paid, location, channel (i.e., internet, in store),) etc.); receiving a target set of time-series data; (Kamarchik figure. 13 element 1310 teaches receiving data, para. [0284] teaches the data is transaction data with locations as it cites “For example, data may be received as a file or aggregated data (e.g., in a spreadsheet or other format) and can include transactions, inventory, product attributes, and/or location attributes.) and applying the machine learning model to the target set of time-series data to generate a classification for the target time-series data. (Kamarchik fig. 1B, shows new time data entered int AI attribute generator, and the AI forecast and analysis generator gives output, this is also shown in figure 13 wherein time-series data in obtained in element 1310, put into machine learning models at elements 1350 and 1370, and getting output from element 1380. Also see fig. 8 and para. [0261-0262] wherein products are matched to lifecycle classes or states by AI model and forecast for demand is also generated.)
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of the Ratner in view of Li with that of Kamarchik in order allow for using a trained machine learning model to further classify new data into a lifecycle state as all the references deal with machine learning and both Li and Kamarchik deal with lifecycle states. The benefit of doing so it allows stores to accurately classify products based on sales trends and to better predict inventory needs relating to restocking items with demand or phasing out products where sales have declined or stopped.
In regards to claim 2, Ratner in view of Li in view of Kamarchik disclose the non-transitory computer readable medium of claim 1, wherein the operations further comprise generating the training data sets at least by:
accessing the plurality of sets of time-series data corresponding to sales of one or more products from a plurality of stores; and (Kamarchik para. [0006, 0045, and 0060] teaches obtaining historical product sales and para. [0145] teaches sales for a product at each location, thus it teaches product-store pairs, so see para. [0180] that teaches transaction data is sales data of a product including date of purchase, quantity, price paid, location, channel (i.e., internet, in store),) etc.)
extracting a set of life-cycle-based features from each set of time-series data of the plurality of sets of time-series data by comparing characteristics of the time-series data to particular features to determine whether the features are present in the time-series data. (Kamarchik para. [0006 and 0008] teaches getting attributes (features) from the product data (historical sales data) about a product wherein the attributes are associated with different product lifecycle phases. Also see figure. 1C wherein transactions (historical sales data) is feed into a Feature extraction unit wherein useable attributes are output. Also see Li para. [0212] teaches using historical time-series data wherein it cites “A time series can be categorized into the short class if the time series has a short record of historical data (e.g., fewer than a predefined number of historical data points).”; also see Li para. [0231] which cites “Because the demand class for each time series is determined based on the characteristics of the time series, there are demand classes that do not contain any time series.”. This teaches mapping/classifying time-series to a demand class based on characteristics (features) of the time-series. Also see Li para. [0212 and 0213] teaches timer-series data being classified in demand classes which include life cycle class, wherein it cites “As one example, an intermittent time series can represent the demand for software that is ending its life cycle and for which there are sporadic periods with no downloads.” and “The demand characteristics that are used to determine the demand class of a time series can include the demand life cycle,…”. This teaches comparing features/characteristics of the time-series to demand classes to see if the features fit in the class or match the class.)
In regards to claim 5, Ratner in view of Li in view of Kamarchik discloses the non-transitory computer readable medium of claim 1, wherein the particular life-cycle classification for the target product-store pair is a short-life-cycle classification, and wherein the operations further comprise: applying the machine learning model to a second set of time-series data associated with a second product-store pair to generate a second life-cycle classification for the second product-store pair, wherein the second life-cycle classification is a long-life-cycle classification. (Kamarchik fig. 4 shows a plurality of sales curves for a plurality of products, and figure 5 shows the various lifecycle states and figure 8 shows matching multiple products and sales curves and sales curves to lifecycle states/profiles. This is also stated Kamarchik in para. [0254-0255 and 261], wherein it teaches matches products to best lifecycle profiles based on attributes (features) of the products. Kamarchik further discloses short lifecycle in paragraph [0065, 0079 and 0080] for short lifecycle and para. [0077-0078] for long lifecycle. Thus, Kamarchik does classify items in lifecycle states and those states include short and long states wherein different items are classified in different groups based on sales history.)
In regards to claim 6, Ratner in view of Li in view of Kamarchik discloses the non-transitory computer readable medium of claim 5, wherein the target product-store pair corresponds to a first product sold from a first store, and wherein the second product-store pair corresponds to the first product sold from a second store. (Kamarchik para. [0121] teaches the system does location information associated with skus were sold in the past, meaning the products sale history include skus and locations associated with a product, also Kamarchik para. [0055 and 0139] teaches locations associated with the sales of products, and Kamarchik para. [0145] teaches forecasting for sales each product at each location, so it would be sales at more than one location for a product.)
In regards to claim 7, Ratner in view of Li in view of Kamarchik disclose the non-transitory computer readable medium of claim 5, wherein the operations further comprise: based on determining the target product-store pair corresponds to the short-life-cycle classification: applying a first forecasting model to a third set of time-series data associated with the target product-store pair to generate a first forecast for the target product-store pair; and based on determining the second product-store pair corresponds to the long-life-cycle classification: applying a second forecasting model to a fourth set of time-series data associated with the second product-store pair to generate a second forecast for the second product-store pair. (Kamarchik fig. 4 shows a plurality of sales curves for a plurality of products, and figure 5 shows the various lifecycle states and figure 8 shows matching multiple products and sales curves and sales curves to lifecycle states/profiles. This is also stated Kamarchik in para. [0254-0255 and 261], wherein it teaches matches products to best lifecycle profiles based on attributes (features) of the products. Kamarchik further discloses short lifecycle in paragraph [0065, 0079 and 0080] for short lifecycle and para. [0077-0078] for long lifecycle. Then in Kamarchik paragraphs [0119-0124] it teaches extracting sales history of product, matching its sales history to a lifecycle state/profile, and then using the profile to forecast future sales of the product. As seen from Kamarchik figures 5 and 8 various products are matched to different curves (lifecycle states) and those are used to predict further sales, thus it would obvious to apply a first forecasting model (lifecycle state) to the time-series data (sales history) and a second forecasting model to a second time-series data, wherein the curves can be any of the identified matching curves including long and short.)
In regards to claim 8, Ratner in view of Li in view of Kamarchik disclose the non-transitory computer readable medium of claim 7, wherein the operations further comprise: applying the machine learning model to a fifth set of time-series data associated with a third product-store pair to generate a third life-cycle classification for the third product-store pair, wherein the third life-cycle classification is an inconclusive-type classification. (Li discloses a timer-series being inconclusive in para. [0212] wherein it cites “Other time series that do not span long time periods and cannot be classified are categorized as an "other" time series.”, wherein the examiner interprets other to be inconclusive as it can not be classified.)
In regards to claim 9, Ratner in view of Li in view of Kamarchik disclose the non-transitory computer readable medium of claim 8, wherein the operations further comprise: based on determining the third product-store pair corresponds to the inconclusive-type classification: comparing attributes of the third product-store pair to attributes of at least one of the target product-store pair and the second product-store pair to assign an interim classification to the third product-store pair, wherein the interim classification corresponds to one of the long-life-cycle classification and the short-life-cycle classification; and based on the interim classification: applying one of the first forecasting model and the second forecasting model to a sixth set of time-series data associated with the third product-store pair to generate a third forecast for the third product-store pair. (Li discloses a timer-series being inconclusive in para. [0212] wherein it cites “Other time series that do not span long time periods and cannot be classified are categorized as "other" time series.”, wherein the examiner interprets other to be inconclusive as it cannot be classified. Kamarchik para. [0050] teaches when a new product is used, there is little or no historical sales data and thus the lifecycle can not be determined. The examiner interprets this as being inclusive or the “other” of Li. Kamarchik para. [0050] further teaches decomposing the product and parameters into attributes. It then compares the attributes of the product in question to other product attributes to find a similar one, and using products lifecycle for forecasting the new product. Li para. [0212] determining if a lifecycle is long or show based on the time-series data and Kamarchik figs. 5 and 8 teaches assigning product and time-series lifecycles. It would have been obvious to combine the reference to compare an inconclusive time-series to other time-series based on attributes and using the closest matching time-series lifecycle to forecast for the inconclusive time-series, whether the time-series, long, short, or any other identified time series as both reference deal with using time-series data and determining demand for a product based on the sales data. Doing so would allow for new products with little to no data be forecasted for as suggest in Eugene reference.)
In regards to claim 10, it is the method embodiment of claim 1 with similar limitations and thus is rejected using the same reasoning as that of claim 1.
In regards to claim 11, it is the method embodiment of claim 2 with similar limitations and thus is rejected using the same reasoning as that of claim 2.
In regards to claim 14, it is the method embodiment of claim 5 with similar limitations and thus is rejected using the same reasoning as that of claim 5.
In regards to claim 15, it is the method embodiment of claim 6 with similar limitations and thus is rejected using the same reasoning as that of claim 6.
In regards to claim 16, it is the method embodiment of claim 7 with similar limitations and thus is rejected using the same reasoning as that of claim 7.
In regards to claim 17, it is the method embodiment of claim 8 with similar limitations and thus is rejected using the same reasoning as that of claim 8.
In regards to claim 18, it is the method embodiment of claim 9 with similar limitations and thus is rejected using the same reasoning as that of claim 9.
In regards to claim 19, it is the system embodiment of claim 1 with similar limitations to claim 1 and thus is rejected using the same reasoning as that of claim 1.
In regards to claim 20, it is the system embodiment of claim 2 with similar limitations to claim 2 and thus is rejected using the same reasoning as that of claim 2.
Claims 3 and 12 are rejected under 35 U.S.C. 103 as being unpatentable over Ratner et al. (“Data Programming: Creating Large Training Sets, Quickly” – hereinafter Ratner) in view of Li et al. (US2020/0143246 A1 – hereinafter Li) in view of Kamarchik et al. (US 2022/0245572 A1 – hereinafter Kamarchik) and further in view of Lavi (US 2012/0204220 A1).
In regards to claim 3, Ratner in view of Li in view of Kamarchik does not explicitly disclose the non-transitory computer readable medium of claim 1, wherein comparing each set of time-series data of the plurality of sets of time-series data to a set of feature-based rules comprises: comparing, in a predetermined sequence, the particular set of time-series data to a respective feature-based rule among the set of feature-based rules; and based on detecting the match between the particular feature-based rule and the particular set of time-series data: refraining from comparing the particular set of time-series data to any additional feature-based rules among the set of feature-based rules. (Examiner note: the examiner interprets this claim to mean that the rules are organized in order list and the timer-series data is compared to the rules one at time in order, and once a first match is found then the comparing stops.)
In regards to claim 12, it is the method embodiment of claim 3 with similar limitations and thus is rejected using the same reasoning as that of claim 3.
Lavi disclose a system that using rules wherein the rules are organized an ordered list of rules and the data is compared to rules in order and when a match is found then it refrains from comparing other rules to the data. (Li para. [0034] cites “The rules are organized in a security rule-set corresponding to a specified security policy. Typically, the security rule-set is configured as an ordered list of rules that is processed from the top to the bottom in sequential order. The rules can be organized based on "first-match-wins" principle in which the first rule that matches a given IP packet will determine that packet's fate.” The IP packet is the time-series data, the rules are the feature-based rules, and this teaches comparing the data to each rule in order and once a match is found, the data is not compared to other rules in the list.)
It would have been obvious to one of ordinary skill in the art before earlies effective filing date of the claimed invention to modify the teachings of the Ratner in view of Li in view of Kamarchik with that of Lavi in order to allow for using first-match-wins principles when matching data to rules as both Ratner and Lavi deal with using rules and the benefit of doing so it allow for fast processing as the system would not have to evaluate every rule and avoids cases where multiple rules match and extra analysis is needed to determine which rule to apply.
Claims 4 and 13 are rejected under 35 U.S.C. 103 as being unpatentable over Ratner et al. (“Data Programming: Creating Large Training Sets, Quickly” – hereinafter Ratner) in view of Li et al. (US2020/0143246 A1 – hereinafter Li) in view of Kamarchik et al. (US 2022/0245572 A1 – hereinafter Kamarchik) and further in view of Jo et al. (US 2019/0220877 A1 – hereinafter Jo).
In regards to claim 4, Ratner in view of Li in view of Kamarchik does not explicitly disclose the non-transitory computer readable medium of claim 1, wherein the machine learning model is a random forest classifier model.
Jo disclose wherein the machine learning model is random forest classifier model. (Jo para. [0062] teaches a machine learning model that forecast a product based on product attributes and prior sales history and determines lifecycle also wherein it states “Specifically, the characteristic calculator 23 calculates forecasting characteristics according to the forecasting requirements (for example, a forecasting period and a model period) that are previously determined (S21). The characteristic calculator 23 calculates static product characteristics (for example, product classification and price) from the sales performance data 11 (S22). The characteristic calculator 23 calculates dynamic product characteristics (for example, a sales interval and average sales figures) from the sales performance data 11 (S23). The characteristic calculator 23 calculates lifecycle characteristics (product phase (introduction stage/growth stage/maturity stage/decline stage)) from launch date information on the product and a sales change in the learning period of the product that are obtained from the sales performance data 11 (S24).”. It further teaches machine learning model being a random forest in para. [0065] wherein it cites “The error forecasting model learning unit 24 learns the error forecasting model 40 by regression analysis using a known machine learning technology, such as random forest, based on the explanatory variable X1 and the objective variable X2.”)
It would have been obvious to one of ordinary skill in the art before earlies effective filing date of the claimed invention to modify the teachings of the Ratner in view of Li in view of Kamarchik with that of the Jo in order to user a random forest model as Li, Kamarchik and Jo deal with lifecycles and forecasting demand or sales. The benefit using a random forest is that random forest handle non-linear data well which is what sales data is, thus it creates an efficient and accurate system for forecasting based on prior sales data.
In regards to claim 13, it is the method embodiment of claim 4 with similar limitations and thus is rejected using the same reasoning as that of claim 4.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PAULINHO E SMITH whose telephone number is (571)270-1358. The examiner can normally be reached Mon-Fri. 10AM-6PM CST.
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, Abdullah Kawsar can be reached at 571-270-3169. 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.
/PAULINHO E SMITH/Primary Examiner, Art Unit 2127
Prosecution Insights