DETAILED ACTION
Status of the Application
Claims 2-21 have been examined in this application. This communication is the first action on the merits. The information disclosure statement (IDS) submitted on 10/10/2024; was filed with this application. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner
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 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.
This action is a Non-Final Action on the merits in response to the application filed on 12/17/2024.
Claims 2-21 remain pending in this application.
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 2-10 are directed towards a system, claims 11-19 are directed towards a method. and claims 20, 21 are directed towards a computer-readable medium, all of which are among the statutory categories of invention.
Claims 2-21 are rejected under 35 U.S.C. 101 because the claims are directed to a judicial exception without significantly more.
Step 1: This part of the eligibility analysis evaluates whether the claim falls within any statutory category. See MPEP 2106.03. The claim recites at least one step or act. Thus, the claim is to a process, which is one of the statutory categories of invention. (Step 1: YES).
Step 2A, Prong One: This part of the eligibility analysis evaluates whether the claim recites a judicial exception. As explained in MPEP 2106.04, subsection II, a claim “recites” a judicial exception when the judicial exception is “set forth” or “described” in the claim.
With respect to claims 2-21, the independent claims (claims 2, 11, and 20) are directed to managing sale data, In independent claim 2, the bolded limitations emphasized below correspond to the abstract ideas of the claimed invention:
Claim 2, a computer system for training a neural network for predicting a customer value associated with an entity, comprising:
a processor;
a computer-readable medium having stored thereon instructions that are executable to cause the computer system to perform operations comprising:
receiving input data from past transaction data associated with the entity at the neural network comprising a series of sequentially connected neural network modules;
forwardly passing the input data through the series of neural network modules via a plurality of communication pathways, including forwardly passing a respective intermediate output from each of the series of neural network modules directly to one or more following neural network modules;
training the neural network based on a predicted customer value from an output layer module of the series of neural network modules and an actual customer value from the past transaction data;
generating, by the trained neural network, a customer value associated with a new entity based on an input of profile data associated with the new entity.
these steps fall within and recite an abstract ideas because they are directed to a method of organizing human activity which includes commercial interaction such as sales activities and/or business relations (See MPEP 2106.04(a)(2) II).
If a claim limitation, under its broadest reasonable interpretation, covers commercial interaction, then it falls within the “method of organizing human activity” grouping of abstract ideas. Therefore, If the identified limitation(s) falls within any of the groupings of abstract ideas enumerated in the MPEP 2106, the analysis should proceed to Prong Two. (Step 2A, Prong One: YES).
Step 2A, Prong Two: This part of the eligibility analysis evaluates whether the claim as a whole integrates the recited judicial exception into a practical application of the exception or whether the claim is “directed to” the judicial exception. This evaluation is performed by (1) identifying whether there are any additional elements recited in the claim beyond the judicial exception, and (2) evaluating those additional elements individually and in combination to determine whether the claim as a whole integrates the exception into a practical application. See MPEP 2106.04(d). The claim recites the additional elements of machine learning model, device, processor, computer-readable medium. The claims recite the steps are performed by the machine learning model, device, processor, computer-readable medium.
The limitations of
forwardly passing the input data through the series of neural network modules via a plurality of communication pathways, including forwardly passing a respective intermediate output from each of the series of neural network modules directly to one or more following neural network modules;
training the neural network based on a predicted customer value from an output layer module of the series of neural network modules and an actual customer value from the past transaction data;
are mere data gathering and processing recited at a high level of generality, and thus are insignificant extra-solution activity. See MPEP 2106.05(g) (“whether the limitation is significant”). In addition, all uses of the recited judicial exceptions require such data gathering and output, and, as such, these limitations do not impose any meaningful limits on the claim. These limitations amount to necessary data gathering and outputting. See MPEP 2106.05.
Further, the limitations are recited as being performed by neural network, processor, computer-readable medium, neural network modules, output layer modules. The neural network, processor, computer-readable medium, neural network modules, output layer modules are recited at a high level of generality. In limitation (a), the neural network is used as a tool to perform the generic computer function of receiving data. See MPEP 2106.05(f). The neural network is used to perform an abstract idea, as discussed above in Step 2A, Prong One, such that it amounts to no more than mere instructions to apply the exception using a generic computer. See MPEP 2106.05(f). Additionally, claim 2 recites neural network. The general use of a neural network does not provide a meaningful limitation to transform the abstract idea into a practical application.
Even when viewed in combination, these additional elements do not integrate the recited judicial exception into a practical application (Step 2A, Prong Two: NO), and the claim is directed to the judicial exception. (Step 2A: YES).
Step 2B: This part of the eligibility analysis evaluates whether the claim as a whole amounts to significantly more than the recited exception i.e., whether any additional element, or combination of additional elements, adds an inventive concept to the claim. See MPEP 2106.05. As explained with respect to Step 2A, Prong Two, the additional elements are the neural network, processor, computer-readable medium, neural network modules, output layer modules. The additional elements were found to be insignificant extra-solution activity in Step 2A, Prong Two, because they were determined to be insignificant limitations as necessary data gathering and processing. Then, the machine learning techniques recited in the claim are disclosed at a high-level of generality (see at least Specification [0024] “Densely connected neural networks may be used in some instances to calculate CV or another quantity. A unified model architecture (which may or may not make use of a densely connected neural network) can also be used to predict not only CV, but also sub-components of CV. Additionally, a structured convolutional neural network can be used to predict CV or another quantity using raw data (rather than engineered data) in various embodiments.”]) and does not amount to significantly more than the abstract idea.
However, a conclusion that an additional element is insignificant extra solution activity in Step 2A, Prong Two should be re-evaluated in Step 2B. See MPEP 2106.05, subsection I.A. At Step 2B, the evaluation of the insignificant extra-solution activity consideration takes into account whether or not the extra-solution activity is well understood, routine, and conventional in the field. See MPEP 2106.05(g). As discussed in Step 2A, Prong Two above, the recitations of
forwardly passing the input data through the series of neural network modules via a plurality of communication pathways, including forwardly passing a respective intermediate output from each of the series of neural network modules directly to one or more following neural network modules;
training the neural network based on a predicted customer value from an output layer module of the series of neural network modules and an actual customer value from the past transaction data;
are recited at a high level of generality. These elements amount to transmitting and processing data are well understood, routine, conventional activity. See MPEP 2106.05(d), subsection II. 10 As discussed in Step 2A, Prong Two above, the recitation of a neural network, processor, computer-readable medium, neural network modules, output layer modules to perform limitations amounts to no more than mere instructions to apply the exception using a generic computer component. Even when considered in combination, these additional elements represent mere instructions to implement an abstract idea or other exception on a computer and insignificant extra-solution activity, which do not provide an inventive concept. (Step 2B: NO).
Dependent claims 3-10, 12-19, 21 do not contain any new additional elements. Rather, these claims offer further descriptive limitations of elements found in the independent claims. In this case, the claims are rejected for the same reasons at step 2a, prong one; step 2a, prong 2; and step 2b. Thus, the claim is not patent eligible.
Regarding the dependent claims, dependent claims 3, 10, 12, 19, 21 recite modules; claims 4, 5, 8, 9, 13, 14, 17, 18 recite neural network. The dependent claims 3-10, 12-19, 21 recite limitations that are not technological in nature and merely limits the abstract idea to a particular environment. Claims 3-10, 12-19, 21 recites neural network, processor, computer-readable medium, neural network modules, output layer modules which are considered an insignificant extra-solution activities of collecting and analyzing data; see MPEP 2106.05(g). Claims 3-10, 12-19, 21 recites neural network, processor, computer-readable medium, neural network modules, output layer modules, which merely recites an instruction to apply the abstract idea using a generic computer component; MPEP 2106.05(f). Additionally, claims 3-10, 12-19, 21 recite steps that further narrow the abstract idea. No additional elements are disclosed in the dependent claims that were not considered in independent claims 1, 11, and 20. Therefore claims 3-10, 12-19, 21 do not provide meaningful limitations to transform the abstract idea into a patent eligible application of the abstract idea such that the claims amount to significantly more than the abstract idea itself.
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 of this title, 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.
Claims 2-21 are rejected under 35 U.S.C. 103 as being unpatentable over United States Patent Publication US 20160247173, Manoharan, et al. to hereinafter Manoharan in view of United States Patent Publication US 20160342891, Ross, et al.
Referring to Claim 2, Manoharan teaches a computer system for training a neural network for predicting a customer value associated with an entity (
Manoharan teaches prediction of customer lifetime value, which is a customer value associated with a customer/entity. Manoharan teaches that “Customer Lifetime Value (CLV), in general, refers to an estimation of an overall value that an organization may derive from customers during their association with the organization.” Manoharan further teaches that the customer value “may be in terms of overall profitability or revenue generated by the association of the customers with the organization,” see paragraphs [0003–0005].
Manoharan further teaches a Data Analysis System (“DAS”) for predicting CLV. Manoharan teaches that the DAS “may predict CLV using a segment level churn,” see paragraphs [0015–0018].),
comprising: a processor (
Manoharan teaches that “The DAS 104 may include a processors 202, an interfaces 204, and a memory 206,” see paragraph [0029]. Manoharan additionally teaches that “The processor 202, amongst other capabilities, may be configured to fetch and execute computer-readable instructions stored in the memory 206,” see paragraph [0030].);
a computer-readable medium having stored thereon instructions that are executable to cause the computer system to perform operations comprising (
Manoharan teaches that its modules may be machine-readable software instructions executed by a processing unit. In particular, Manoharan teaches that “the modules 208 may be machine-readable instructions software which, when executed by a processor/processing unit, perform any of the described functionalities,” see paragraph [0034].
Manoharan further teaches that “The machine-readable instructions may be stored on an electronic memory device, hard disk, optical disk or other machine-readable storage medium or non-transitory medium,” see paragraph [0034].):
receiving input data from past transaction data associated with the entity (
Manoharan teaches receiving customer purchasing/transaction data: “The DAS may receive datasets including data representative of purchasing behavior of customers over a predefined time period.” see paragraph [0016]. It also states that its data collection module collects “data related to transactions conducted by customers.” See paragraphs [0021–0024; 0043–0049].)
generating, by the trained neural network, a customer value associated with a new entity based on an input of profile data associated with the new entity (
Manoharan teaches that the DAS analyzes transaction data and predicts customer lifetime value. Manoharan teaches that after the customer data is analyzed, “the DAS 104 may predict expected lifetime of the customers based on their purchasing behavior,” see paragraph [0028].
Manoharan further teaches that the prediction module “may predict the CLV for each customer of each segment” based on the predicted expected lifetime, see paragraph [0046]. Manoharan teaches that the resulting CLV can be stored in a database for later use, see paragraphs [0047; 0054–0055].).
training the neural network (See Ross) based on a predicted customer value from an output layer module of the series of neural network modules (See Ross) and an actual customer value from the past transaction data (
Manoharan predicts CLV from historical purchasing behavior but uses weighted RFM segmentation and churn calculations, not disclosed neural-network supervised training against actual labels. It states CLV is predicted based on expected lifetime computed from churn/transaction characteristics. See paragraphs [0021–0029; 0050–0058].);
Manoharan does not explicitly teach at the neural network comprising a series of sequentially connected neural network modules; forwardly passing the input data through the series of neural network modules via a plurality of communication pathways; including forwardly passing a respective intermediate output from each of the series of neural network modules directly to one or more following neural network modules; training the neural network based on a predicted customer value from an output layer module of the series of neural network modules and an actual customer value from the past transaction data
However, Ross teaches these limitations.
at the neural network comprising a series of sequentially connected neural network modules (
Ross teaches that “Neural networks are machine learning models that employ one or more layers of models to generate an output, e.g., a classification, for a received input,” see paragraph [0003].
Ross teaches that some neural networks include hidden layers and an output layer, and that “The output of each hidden layer is used as input to the next layer in the network, i.e., the next hidden layer or the output layer of the network,” see paragraph [0003].
Ross further teaches that “The layers of the neural network are arranged in a sequence, each with a respective set of weights,” see paragraph [0016].);
forwardly passing the input data through the series of neural network modules via a plurality of communication pathways (
Ross teaches that “the neural network receives the input and processes it through each of the neural network layers in the sequence to generate the inference, with the output from one neural network layer being provided as input to the next neural network layer,” see paragraph [0017].
Ross further teaches that an input to a network layer may be an original neural-network input or the output of an earlier layer. Ross teaches that “Data inputs to a neural network layer, e.g., either the input to the neural network or the outputs of the layer below the layer in the sequence, to a neural network layer can be referred to as activation inputs to the layer,” see paragraph [0017].),
including forwardly passing a respective intermediate output from each of the series of neural network modules directly to one or more following neural network modules (
Ross teaches that layer output is retained and used as input to a subsequent layer. Ross teaches that “[t]he output of the layer can be stored in the unified buffer for use as an input to a subsequent layer in the neural network,” see paragraph [0022].
Ross additionally teaches non-strictly-sequential network arrangements. Ross teaches that “the layers of the neural network are arranged in a directed graph,” such that “any particular layer can receive multiple inputs, multiple outputs, or both,” see paragraph [0018].);
training the neural network based on a predicted customer value (See Manoharan) from an output layer module of the series of neural network modules and an actual customer value from the past transaction data (See Manoharan) (
Ross teaches that layer output is retained and used as input to a subsequent layer. Ross teaches that “[t]he output of the layer can be stored in the unified buffer for use as an input to a subsequent layer in the neural network,” see paragraph [0022].
Ross additionally teaches non-strictly-sequential network arrangements. Ross teaches that “the layers of the neural network are arranged in a directed graph,” such that “any particular layer can receive multiple inputs, multiple outputs, or both,” see paragraph [0018].
Ross teaches that each neural-network layer operates in accordance with layer parameters. Ross teaches that “Each layer of the network generates an output from a received input in accordance with current values of a respective set of parameters,” see paragraph [0003].
Ross teaches a matrix computation unit that receives “a plurality of weight inputs and a plurality of activation inputs” and generates accumulated values, and a vector computation unit that applies “an activation function to each accumulated value” to generate activated layer values, see paragraphs [0005–0006].
Ross teaches that although the disclosed hardware is described for neural-network inference, “the hardware can be used for one or more of the following convolutional or fully-connected neural network training,” see paragraph [0059].);
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to implement Manoharan’s customer-value/CLV prediction system using the multi-layer neural-network computation techniques taught by Ross, with the motivation of applying a known machine-learning architecture to process Manoharan’s large-volume, multi-parameter customer and transaction data and generate a customer-value prediction more efficiently.
Manoharan teaches that conventional customer-value prediction techniques “consider fewer number of parameters and do not analyze a large volume of data,” resulting in lower accuracy and inefficient handling of large volumes of customer data, see paragraphs [0014–0015].
Ross teaches that neural networks generate outputs from received inputs using multiple layers, each having parameters, see paragraphs [0003; 0016–0017]. Ross further teaches that the disclosed neural-network processor improves efficiency by increasing speed and throughput and reducing power and cost, and can process neural-network layers containing a large number of inputs per neuron, see paragraph [0007].
Accordingly, a person of ordinary skill would have had reason to apply Ross’s known multi-layer neural-network processing architecture to the transaction-data/CLV-prediction system of Manoharan to improve processing of complex, high-volume, multi-parameter customer data and to obtain a corresponding predictive output.
Referring to Claim 3, Manoharan teaches the computer system of claim 2, Manoharan does not explicitly teach wherein the operation of forwardly passing the input data comprises:
, through a first set of data communication pathways leading from outputs of the first layer module to an input of a second layer module of the series of neural network modules, an input of a third layer module of the series of neural network modules, and an input of a fourth layer module of the series of neural network modules; forwardly passing first intermediate data from the second layer module, through a second set of data communication pathways leading from outputs of the second layer module to the input of the third layer module and the input of the fourth layer module; forwardly passing second intermediate data from the third layer module, through a third set of data communication pathways leading from outputs of the third layer module to inputs of the fourth layer module; forwardly passing third intermediate data from only outputs of the fourth layer module to an output layer module.
However, Ross teaches wherein the operation of forwardly passing the input data comprises:
, through a first set of data communication pathways leading from outputs of the first layer module to an input of a second layer module of the series of neural network modules, an input of a third layer module of the series of neural network modules, and an input of a fourth layer module of the series of neural network modules (
Ross discloses directed-graph arrangements where a layer may have multiple outputs and subsequent layers may receive outputs, as this can be interpreting as first-to-second/third/fourth topology. “Any particular layer can receive multiple inputs, multiple outputs, or both.” See paragraphs [0018–0021].);
forwardly passing first intermediate data from the second layer module, through a second set of data communication pathways leading from outputs of the second layer module to the input of the third layer module and the input of the fourth layer module (
Ross teaches that “the neural network receives the input and processes it through each of the neural network layers in the sequence to generate the inference, with the output from one neural network layer being provided as input to the next neural network layer,” see paragraph [0017].
Ross further teaches that an input to a network layer may be an original neural-network input or the output of an earlier layer. Ross teaches that “Data inputs to a neural network layer, e.g., either the input to the neural network or the outputs of the layer below the layer in the sequence, to a neural network layer can be referred to as activation inputs to the layer,” see paragraph [0017].
Ross discloses directed-graph arrangements where a layer may have multiple outputs and subsequent layers may receive outputs, as this can be interpreting as first-to-second/third/fourth topology. “Any particular layer can receive multiple inputs, multiple outputs, or both.” See paragraphs [0018–0021].);
forwardly passing second intermediate data from the third layer module, through a third set of data communication pathways leading from outputs of the third layer module to inputs of the fourth layer module (
Conventional next-layer disclosure only. “The output of each hidden layer is used as input to the next layer in the network.” See paragraphs [0003–0011].);
forwardly passing third intermediate data from only outputs of the fourth layer module to an output layer module (
Ross describes hidden layers and output layer, with hidden-layer output used as input to the next or output layer. See paragraphs [0003–0011].);
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to implement Manoharan’s customer-value/CLV prediction system using the multi-layer neural-network computation techniques taught by Ross, with the motivation of applying a known machine-learning architecture to process Manoharan’s large-volume, multi-parameter customer and transaction data and generate a customer-value prediction more efficiently.
Manoharan teaches that conventional customer-value prediction techniques “consider fewer number of parameters and do not analyze a large volume of data,” resulting in lower accuracy and inefficient handling of large volumes of customer data, see paragraphs [0014–0015].
Ross teaches that neural networks generate outputs from received inputs using multiple layers, each having parameters, see paragraphs [0003; 0016–0017]. Ross further teaches that the disclosed neural-network processor improves efficiency by increasing speed and throughput and reducing power and cost, and can process neural-network layers containing a large number of inputs per neuron, see paragraph [0007].
Accordingly, a person of ordinary skill would have had reason to apply Ross’s known multi-layer neural-network processing architecture to the transaction-data/CLV-prediction system of Manoharan to improve processing of complex, high-volume, multi-parameter customer data and to obtain a corresponding predictive output.
Referring to Claim 4, Manoharan teaches the computer system of claim 2, Manoharan does not explicitly teach wherein the operation of training the neural network further comprise:
repeatedly adjusting one or more components of the neural network based on a comparison of the predicted customer value and the actual customer value; wherein the adjusting comprises updating weighting and/or functions associated with neurons of the one or more components of the neural network.
However, Ross teaches these limitations.
wherein the operation of training the neural network further comprise:
repeatedly adjusting one or more components of the neural network based on a comparison of the predicted customer value and the actual customer value (
Ross discloses weight inputs and activation functions, and refers broadly to fully connected neural-network training, but retrieved text does not detail iterative backpropagation or label-based parameter updates. “Each layer … generate[s] an output … in accordance with current values of a respective set of parameters.” See paragraphs [0003, 0008–0011; 0068–0071].),
wherein the adjusting comprises updating weighting and/or functions associated with neurons of the one or more components of the neural network (
Ross expressly uses “weight inputs” and applies “an activation function” to accumulated values. “The vector computation unit applies an activation function to the accumulated values.” See paragraphs [0009–0011; 0022–0029]).
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to implement Manoharan’s customer-value/CLV prediction system using the multi-layer neural-network computation techniques taught by Ross, with the motivation of applying a known machine-learning architecture to process Manoharan’s large-volume, multi-parameter customer and transaction data and generate a customer-value prediction more efficiently.
Manoharan teaches that conventional customer-value prediction techniques “consider fewer number of parameters and do not analyze a large volume of data,” resulting in lower accuracy and inefficient handling of large volumes of customer data, see paragraphs [0014–0015].
Ross teaches that neural networks generate outputs from received inputs using multiple layers, each having parameters, see paragraphs [0003; 0016–0017]. Ross further teaches that the disclosed neural-network processor improves efficiency by increasing speed and throughput and reducing power and cost, and can process neural-network layers containing a large number of inputs per neuron, see paragraph [0007].
Accordingly, a person of ordinary skill would have had reason to apply Ross’s known multi-layer neural-network processing architecture to the transaction-data/CLV-prediction system of Manoharan to improve processing of complex, high-volume, multi-parameter customer data and to obtain a corresponding predictive output.
Referring to Claim 5, Manoharan teaches the computer system of claim 2, wherein the trained neural network (See Ross) is configured to predict the customer value associated with the new entity without past transaction information associated with the new entity (
Manoharan teaches prediction from historical customer purchasing behavior and transaction data. Manoharan teaches that the DAS receives data representative of “purchasing behavior of customers over a predefined time period,” see paragraph [0016].
Manoharan further teaches that customer transaction characteristics can include frequency, recency, propensity to purchase, and transaction value, see paragraphs [0017; 0028; 0037–0044].).
Referring to Claim 6, Manoharan teaches the computer system of claim 2, wherein the input data comprises one or more of:
mailing address, country of residence, linked funding sources, email address, device information or network information (
Manoharan teaches that its stored data may include “customer data 218, transaction data 220 and other data 222 which includes margin data and derived data like recency, frequency and monetary values,” see paragraph [0036].
Manoharan further teaches that its dataset may be stored in relation to customer data, see paragraph [0038].
Additionally, Manoharan discloses cash flow in which is equivalent to linked funding sources, see paragraph [0013, 0046]).
Referring to Claim 7, Manoharan teaches the computer system of claim 2, wherein the operations further comprise:
Manoharan does not explicitly teach performing an input selection operation on inputs received at the third layer, , wherein the input selection operation includes weighting an output from one previous layer differently than an output from another previous layer.
However, Ross teaches performing an input selection operation on inputs received at the third layer, , wherein the input selection operation includes weighting an output from one previous layer differently than an output from another previous layer (
Ross teaches multiple inputs/multiple outputs in directed graph networks and weights used in layer computations, but no disclosed third-layer input-selection operation that differently weights outputs from separate preceding layers. See paragraphs [0018–0021; 0022–0029].).
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to implement Manoharan’s customer-value/CLV prediction system using the multi-layer neural-network computation techniques taught by Ross, with the motivation of applying a known machine-learning architecture to process Manoharan’s large-volume, multi-parameter customer and transaction data and generate a customer-value prediction more efficiently.
Manoharan teaches that conventional customer-value prediction techniques “consider fewer number of parameters and do not analyze a large volume of data,” resulting in lower accuracy and inefficient handling of large volumes of customer data, see paragraphs [0014–0015].
Ross teaches that neural networks generate outputs from received inputs using multiple layers, each having parameters, see paragraphs [0003; 0016–0017]. Ross further teaches that the disclosed neural-network processor improves efficiency by increasing speed and throughput and reducing power and cost, and can process neural-network layers containing a large number of inputs per neuron, see paragraph [0007].
Accordingly, a person of ordinary skill would have had reason to apply Ross’s known multi-layer neural-network processing architecture to the transaction-data/CLV-prediction system of Manoharan to improve processing of complex, high-volume, multi-parameter customer data and to obtain a corresponding predictive output.
Referring to Claim 8, Manoharan teaches the computer system of claim 2, wherein the operations further comprise:
approving or denying the user request based on the output (
Manoharan says stored CLV may be used “to provide promotional offers and discounts to the customers,” not to approve or deny a user request. See paragraphs [0028].).
Manoharan does not explicitly teach receiving a user request from a user; based on inputting user data corresponding to the user to the optimized neural network, receiving an output from the optimized neural network.
However, Ross teaches
receiving a user request from a user (
Ross provides generic classification/inference from input, See paragraphs [0006–0018] and user-request approval/denial; “a computer can interact with a user by sending documents to and receiving documents from a device that is used by the user; for example, by sending web pages to a web browser on a user's client device in response to requests received from the web browser.”. See paragraphs [0066].);
based on inputting user data corresponding to the user to the optimized neural network, receiving an output from the optimized neural network (
Ross teaches generating an inference from an input. Ross teaches that “given an input, the neural network can compute an inference for the input,” see paragraph [0016].
Ross further teaches that the network processes the received input through sequential layers to generate the inference, see paragraph [0017].);
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to implement Manoharan’s customer-value/CLV prediction system using the multi-layer neural-network computation techniques taught by Ross, with the motivation of applying a known machine-learning architecture to process Manoharan’s large-volume, multi-parameter customer and transaction data and generate a customer-value prediction more efficiently.
Manoharan teaches that conventional customer-value prediction techniques “consider fewer number of parameters and do not analyze a large volume of data,” resulting in lower accuracy and inefficient handling of large volumes of customer data, see paragraphs [0014–0015].
Ross teaches that neural networks generate outputs from received inputs using multiple layers, each having parameters, see paragraphs [0003; 0016–0017]. Ross further teaches that the disclosed neural-network processor improves efficiency by increasing speed and throughput and reducing power and cost, and can process neural-network layers containing a large number of inputs per neuron, see paragraph [0007].
Accordingly, a person of ordinary skill would have had reason to apply Ross’s known multi-layer neural-network processing architecture to the transaction-data/CLV-prediction system of Manoharan to improve processing of complex, high-volume, multi-parameter customer data and to obtain a corresponding predictive output.
Referring to Claim 9, Manoharan teaches the computer system of claim 8, wherein the trained neural network is used to predict a quantifiable number indicative of a financial risk of approving or denying the user request (
Manoharan teaches a quantifiable customer value relating to profitability or revenue. Manoharan teaches that CLV is an estimation of an overall value to the organization and may be “in terms of overall profitability or revenue,” see paragraph [0003].
Manoharan also teaches a churn value, described as “a probability that a customer may not be in association with the organization,” see paragraph [0043].
Manoharan says stored CLV may be used “to provide promotional offers and discounts to the customers,” not to approve or deny a user request. See paragraphs [0028].).
Referring to Claim 10, Manoharan teaches the computer system of claim 2, Manoharan does not explicitly teach wherein each of the series of neural network modules including a dense layer that has a plurality of neurons connected to all neurons in an immediately preceding neural network module.
However, Ross teaches wherein each of the series of neural network modules including a dense layer that has a plurality of neurons connected to all neurons in an immediately preceding neural network module (
Ross expressly teaches fully connected neural-network layers. Ross teaches that host-interface parameters can specify “a type of layer to be processed, e.g., a convolutional layer or a fully connected layer,” see paragraph [0023].
Ross further teaches that its hardware can be used for “fully-connected neural network training,” see paragraph [0059].).
Therefore, it would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to implement Manoharan’s customer-value/CLV prediction system using the multi-layer neural-network computation techniques taught by Ross, with the motivation of applying a known machine-learning architecture to process Manoharan’s large-volume, multi-parameter customer and transaction data and generate a customer-value prediction more efficiently.
Manoharan teaches that conventional customer-value prediction techniques “consider fewer number of parameters and do not analyze a large volume of data,” resulting in lower accuracy and inefficient handling of large volumes of customer data, see paragraphs [0014–0015].
Ross teaches that neural networks generate outputs from received inputs using multiple layers, each having parameters, see paragraphs [0003; 0016–0017]. Ross further teaches that the disclosed neural-network processor improves efficiency by increasing speed and throughput and reducing power and cost, and can process neural-network layers containing a large number of inputs per neuron, see paragraph [0007].
Accordingly, a person of ordinary skill would have had reason to apply Ross’s known multi-layer neural-network processing architecture to the transaction-data/CLV-prediction system of Manoharan to improve processing of complex, high-volume, multi-parameter customer data and to obtain a corresponding predictive output.
Claims 1-19- recite limitations that stand rejected via the art citations and rationale applied to claims 2-10.
Claims 20, 21 recite limitations that stand rejected via the art citations and rationale applied to claims 2, 3. Regarding, a non-transitory computer-readable medium having stored thereon instructions that are executable by a computer system to cause the computer system to perform operations (
Manoharan expressly teaches non-transitory computer-readable media and instructions. Manoharan teaches: “The machine-readable instructions may be stored on an electronic memory device, hard disk, optical disk or other machine-readable storage medium or non-transitory medium,” see paragraph [0034].):
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Gong, (US10417650), teaches a distributed and automated system for predicting a customer lifetime value.
Zhu et al., (US20180253637), teach “Churn Prediction Using Static And Dynamic Features.”
Young., (WO2018005433), teach “Dynamically managing artificial neural networks”
Chamberlain et al. teach: “Customer Lifetime Value Prediction Using Embeddings.”; 2017, Proceedings of the 23rd ACM SIGKDD international conference on knowledge discovery and data mining.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to UCHE BYRD whose telephone number is (571)272-3113. The examiner can normally be reached Mon.-Fri..
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, Patricia Munson can be reached at (571) 270-5396. 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.
/UCHE BYRD/Examiner, Art Unit 3624
Prosecution Insights