DETAILED ACTION\
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 Objections
Claims 1, 8 and 17 are objected to because of the following informalities: lack of antecedent basis.
Claim 1, 8 and 17 recite “the support cost probability distribution” without preceding “a….”.
Appropriate correction is required.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claim 1, 7 and 18 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claim 1, 7 and 18 recite “wherein the trained model determines the support cost probability distribution based on iterations of applying a splitting criterion to the input to produce two distributions of two resulting split subsets, normalized by the respective sizes of the respective distributions;” But there is no iterations of applying a splitting criterion to the input in the specification. Further the support cost probability distribution can not be determined since there are multiple subsets of probabilities distributions, which one is the support cost probability distribution and how it is determined is described in the specification.
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claim 1, 7 and 18 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, or for pre-AIA the applicant regards as the invention.
Claim 1, 7 and 18 recite “wherein the trained model determines the support cost probability distribution based on iterations of applying a splitting criterion to the input to produce two distributions of two resulting split subsets, normalized by the respective sizes of the respective distributions;” But it is indefinite for failing to particularly point out and distinctly claim the subject matter. Further the support cost probability distribution can not be determined since there are multiple subsets of probabilities distributions, which one is the support cost probability distribution and how it is determined is not described in the specification therefore it is not known. Thus, the scope of the limitation cannot be determined by Examiner. For the purpose of examination, the claims are interpreted with BRI.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1, 3, 8-13, 17 are rejected under 35 U.S.C. 103 as being unpatentable over Riddle et al. (Riddle) US 2021/0004706 in view of Biggs et al. (Biggs) US 2022/0180168, Fukuda et al. (Fukuda) US 2017/0178168 and Dymetman et al. (Dymetman) US 2013/0096877
In regard to claim 1, Riddle disclose A system, ([0010]-[0011] system) comprising:
at least a processor; ([0121][0122] [0125][0131] a processing unit) and
at least one memory that stores executable instructions that, when executed by the at least one processor, ([0121]-[0125][0131] a memory with instructions executed by the processing unit) facilitate performance of operations, ([0011][0029][0073] facilitate actions) comprising:
training an artificial intelligence risk model to produce a trained model, wherein labeled training data for the training ([0011]-[0020] [0024]-[0034] [0043] [0068][0103] training the ML to generate a trained ML model, and labeled data to form a training set)
in response to applying a input to the trained model, wherein the input comprises a feature of a user and a product, producing an output that indicates a first predicted support cost probability distribution that corresponds to the input; ([0012]-[0026] in response to the user input to the trained model, the input has comments associated with a service ticket, such as a product and information about a product user, a assigned service agent, engagement score, etc. and generate an output of change-based probability that product user escalates service for the service ticket (which imply the higher support cost))
based on the first predicted support cost probability distribution, sending offering data to the entity, wherein the offering data is indicative of an offer to modify the product to a modified product, ([0012]-[0029] [0076]-[0095] based on the predicted cost probability distribution change, sending a changed offer to the user, for example, with 95% probability that will escalate service, the new higher level of service is provided, such as extend an agreement to support the product user, etc.) and wherein the modified product is associated with a second predicted support cost with respect to the entity; ([0012]-[0029] [0055]-[0069] [0073]-[0095][0102] based on the predicted cost probability distribution change, sending a changed offer to the product user, with lower probability to escalate the service, (which imply the lower support cost) , such as by extend (upgrade) an agreement to support the product user, etc. and based on user input to agree with the offer from user comments, change to the modified service offer to the user) and
based on receiving acceptance data that is associated with the entity and that is indicative of accepting the offer, switching from the product to the modified product with respect to the entity. ([0012]-[0029] [0055]-[0069] [0073]-[0095][0102] based on the predicted cost probability distribution change, sending a changed offer to the product user, with lower probability to escalate the service, (which imply the lower support cost) , such as by extend (upgrade) an agreement to support the product user, etc. and based on user input to agree with the offer from user comments, change to the modified service offer to the user)
But Riddle fail to explicitly disclose “the labeled training data for the training comprises respective features of users and products, wherein the trained model comprises a causal tree model that is configured to differentiate between first features that are immutable to an entity that utilizes the trained model and second features that are mutable to the entity;”
Biggs disclose the labeled training data for the training comprises respective features of users and products; ([0033]-[0052] groups of customers (with location, age, gender, etc. features), features of products, such as price, etc.)
wherein the trained model comprises a causal tree model that is configured to differentiate between first features that are immutable to an entity that utilizes the trained model and second features that are mutable to the entity; ([0033]-[0047] [0050]-[0052] Fig. 3, using split tree to train the ML model and separate data into two set (features) based on user-defined splitting criterion, such as groups of customers (with same location, age, gender, etc.), one group single male (fixed) with a price vs. another probable houseowner, renter, with a price, etc. Note: please further specify the entity, there are many possibilities and please use functional language to describe the invention.)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Biggs’s method of segmentation and interpretable prescriptive polies generation into Riddle’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Biggs’s using split tree to distinguish the different features would help to provide more ML training method to Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that distinguishing the different features using the split tree would facilitate the finetuning of the ML model and improve the output accuracy of the ML model.
But Riddle and Biggs fail to explicitly disclose “and corresponding labels of respective support costs, wherein the first predicted support cost probability distribution identifies multiple samples that identify respective possible support costs and respective probabilities that each of the respective possible support costs will occur;”
Fukuda disclose and corresponding labels of respective support costs, ([0003][0017][0024]-[0031] corresponding labels of the costs)
wherein the first predicted support cost probability distribution identifies multiple samples that identify respective possible support costs and respective probabilities that each of the respective possible support costs will occur; ([0003][0017][0024]-[0031] cost probability distribution identify test samples that identify each cost for each service and the probabilities the cost estimation for the service.)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Fukuda’s method of training a machine learning model into Biggs and Riddle’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Fukuda’s identify probability distributions for the cost would help to provide a cost estimation method to Biggs and Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing a cost estimation with probability distributions would improve the cost prediction accuracy of the ML model.
But Riddle and Biggs, Fukuda fail to explicitly disclose “and wherein the trained model determines the support cost probability distribution based on iterations of applying a splitting criterion to the input to produce two distributions of two resulting split subsets, normalized by the respective sizes of the respective distributions;”
Dymetman disclose and wherein the trained model determines the support cost probability distribution based on iterations of applying a splitting criterion to the input to produce two distributions of two resulting split subsets, normalized by the respective sizes of the respective distributions; ([0004]-[0008][0031]-[0032] [0042]-[0054] [0064]-[0066] claim 3, the model determine the probability distribution base on repeatedly splitting data set to generate output with two distributions of two resulting split subsets and normalized the size of the distributions)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Dymetman’s of model with split partition method into Fukuda, Biggs and Riddle’s invention as they are related to the same field endeavor of ML model. The motivation to combine these arts, as proposed above, at least because Dymetman’s of model of identifying probability distributions with split partition would help to provide more probability distribution method into to Fukuda, Biggs and Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more probability distribution method with split partition would improve the proposal probability distribution of the ML model.
In regard to claim 3, Riddle, Biggs, Fukuda and Dymetman disclose The system of claim 1, the rejection is incorporated herein.
But Riddle, Fukuda, Dymetman fail to explicitly disclose “wherein the causal tree model is configured to implement decision tree learning to identify a strategy for splitting observed individuals into groups to estimate heterogeneous treatment effects.”
Biggs disclose wherein the causal tree model is configured to implement decision tree learning to identify a strategy for splitting observed individuals into groups to estimate heterogeneous treatment effects. ([0033]-[0047] [0050]-[0052] Fig. 3, using split tree to train the ML model and separate data into two set (features) based on user-defined splitting criterion to segment user such as groups of customers (with location, age, gender, etc.) with various prices to get the maximum revenue.)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Biggs’s method of segmentation and interpretable prescriptive polies generation into Dymetman, Fukuda and Riddle’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Biggs’s using split tree to distinguish the different features would help to provide more ML training method to Dymetman, Fukuda and Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that distinguishing the different features using the split tree would facilitate the finetuning of the ML model and improve the output accuracy of the ML model.
In regard to claim 8, Riddle disclose A method, ([0019]-[0020] method) comprising:
performing learning, by a system comprising a processor, on the explainable artificial intelligence risk model to produce a trained model, ([0011]-[0020] [0024]-[0034] [0043] [0068][0103][0121][0122][0125] [0131] a processing unit, perform learning and on a ML model to generate a trained ML model) wherein labeled training data for the training ([0011]-[0020] [0024]-[0034] [0043] [0068][0103] training the ML to generate a trained ML model, and labeled data to form a training set)
in response to applying a first input to the trained model, wherein the first input comprises a feature of a user and a product, producing, by the trained model, an output that indicates a first predicted support cost probability distribution that corresponds to the first input; ([0012]-[0026] in response to the user input to the trained model, the input has comments associated with a service ticket, such as a product and information about a product user, a assigned service agent, engagement score, etc. and generate an output of change-based probability that product user escalates service for the service ticket (which imply the higher support cost))
based on the first predicted support cost probability distribution, sending, by the system, offering data to the entity, wherein the offering data is indicative of an offer to modify the product to a modified product, ([0012]-[0029] [0076]-[0095] based on the predicted cost probability distribution change, sending a changed offer to the user, for example, with 95% probability that will escalate service, the new higher level of service is provided, such as extend an agreement to support the product user, etc.) and wherein the modified product is associated with a second predicted support cost with respect to the entity; and based on receiving acceptance data that is associated with the entity and that is indicative of accepting the offer, switching, by the system, from the product to the modified product with respect to the entity. ([0012]-[0029] [0055]-[0069] [0073]-[0095][0102] based on the predicted cost probability distribution change, sending a changed offer to the product user, with lower probability to escalate the service, (which imply the lower support cost) , such as by extend (upgrade) an agreement to support the product user, etc. and based on user input to agree with the offer from user comments, change to the modified service offer to the user)
But Riddle fail to explicitly disclose “an explainable artificial intelligence risk model, wherein the trained model comprises a causal tree model;”
Biggs disclose an explainable artificial intelligence risk model, wherein the trained model comprises a causal tree model; ([0006]-[0007] [0020]-[0024] [0033]-[0047] [0050]-[0052] the ML model is a interpretable policy model,8using split tree to train the ML model and separate data into two set based on user-defined splitting criterion)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Biggs’s method of segmentation and interpretable prescriptive polies generation into Sandler’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Biggs’s using split tree to distinguish the different features would help to provide more ML training method to Sandler’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that distinguishing the different features using the split tree would facilitate the finetuning of the ML model and improve the output accuracy of the ML model.
But Riddle and Biggs fail to explicitly disclose “performing supervised learning,
and corresponding labels of respective support costs, wherein the first predicted support cost probability distribution identifies multiple samples that identify respective possible support costs and respective probabilities that each of the respective possible support costs will occur;”
Fukuda disclose performing supervised learning, ([0025] supervised ML)
and corresponding labels of respective support costs ([0003][0017][0024]-[0031] corresponding labels of the costs)
wherein the first predicted support cost probability distribution identifies multiple samples that identify respective possible support costs and respective probabilities that each of the respective possible support costs will occur; ([0003][0017][0024]-[0031] cost probability distribution identify test samples that identify each cost for each service and the probabilities the cost estimation for the service.)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Fukuda’s method of training a machine learning model into Biggs and Riddle’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Fukuda’s identify probability distributions for the cost would help to provide a cost estimation method to Biggs and Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing a cost estimation with probability distributions would improve the cost prediction accuracy of the ML model.
But Riddle and Biggs, Fukuda fail to explicitly disclose “and wherein the trained model determines the support cost probability distribution based on iterations of applying a splitting criterion to the input to produce two distributions of two resulting split subsets, normalized by the respective sizes of the respective distributions;”
Dymetman disclose and wherein the trained model determines the support cost probability distribution based on iterations of applying a splitting criterion to the input to produce two distributions of two resulting split subsets, normalized by the respective sizes of the respective distributions; ([0004]-[0008][0031]-[0032] [0042]-[0054] [0064]-[0066] claim 3, the model determine the probability distribution base on repeatedly splitting data set to generate output with two distributions of two resulting split subsets and normalized the size of the distributions)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Dymetman’s of model with split partition method into Fukuda, Biggs and Riddle’s invention as they are related to the same field endeavor of ML model. The motivation to combine these arts, as proposed above, at least because Dymetman’s of model of identifying probability distributions with split partition would help to provide more probability distribution method into to Fukuda, Biggs and Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more probability distribution method with split partition would improve the proposal probability distribution of the ML model.
In regard to claim 9, Riddle, Biggs, Fukuda and Dymetman disclose The method of claim 8, the rejection is incorporated herein.
But Riddle, Fukuda and Dymetman fail to explicitly disclose “wherein the causal tree model comprises an uplift tree model that is configured to determine an impact of a treatment on a target given a feature.”
Biggs disclose wherein the causal tree model comprises an uplift tree model that is configured to determine an impact of a treatment on a target given a feature. ([0005]- [0009] [0020]-[0024][0034]-[0052] quantify between accuracy and interpretability to provide guidance to a decision-maker, “a difference between the best expected outcome and an expected outcome for the interpretable prescriptive policy is determined, where the difference quantifies a trade-off between the first policy and interpretability of the interpretable prescriptive policy in terms of expected outcome” the interpretability cost is utilized as a loss function to facilitate finetuning/adjusting/updating of the boosted tree model)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Biggs’s method of segmentation and interpretable prescriptive polies generation into Dymetman, Fukuda and Riddle ’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Biggs’s using split tree to distinguish the different features would help to provide more ML training method to Dymetman, Fukuda and Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that distinguishing the different features using the split tree would facilitate the finetuning of the ML model and improve the output accuracy of the ML model.
In regard to claim 10, Riddle, Biggs, Fukuda, Dymetman disclose The method of claim 8, the rejection is incorporated herein.
But Fukuda, Riddle, Dymetman fail to explicitly disclose “wherein the labeled training data omits a feature of the product that is able to be part of a second input to the trained model.”
Biggs disclose wherein the labeled training data omits a feature of the product that is able to be part of a second input to the trained model. ([0034]-[0039] a path from the root node of the tree to a particular leaf node of the tree specifies particular splitting segments, where constructing a decision tree with a user-defined splitting criterion based on the feature of the product, and the optimal path is selected with largest gain, the other segments are ignored)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Biggs’s method of segmentation and interpretable prescriptive polies generation into Dymetman, Fukuda and Riddle ’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Biggs’s using split tree to filter different features would help to provide more feature filtering method to Dymetman, Fukuda and Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that filtering features using the split tree would facilitate the finetuning of the ML model and improve the output accuracy of the ML model.
In regard to claim 11, Riddle, Biggs, Fukuda and Dymetman disclose The method of claim 8, the rejection is incorporated herein.
But Fukuda and Riddle, Dymetman fail to explicitly disclose “wherein the trained model is a first trained model, and further comprising: determining, by the system, a first risk type and a second risk type associated with the features of users and products, wherein the first trained model is associated with the first risk type; and performing supervised learning, by the system, on the explainable artificial intelligence risk model to produce a second trained model, wherein the second trained model is associated with the second risk type.”
Biggs disclose wherein the trained model is a first trained model, and further comprising: determining, by the system, a first risk type and a second risk type associated with the features of users and products, wherein the first trained model is associated with the first risk type; ([0007]-[0009][0020]-[0028] [0034]-[0052] training a first AI model, the first model is associated with a likelihood of a desired outcome for a given action, which is risk minimization problem associated with features of product or a customer, for example, purchase an item) and performing supervised learning, by the system, on the explainable artificial intelligence risk model to produce a second trained model, wherein the second trained model is associated with the second risk type. ([0007]-[0009] [0020]-[0028] [0034]-[0052] the second model with a prescriptive tree and trained using labeled data (disclosed by Sandler) and generate interpretable prescriptive policies to determine the expected outcome for the interpretable prescriptive policy which is a second risk type and this is user defined, for example, expected revenue maximization to produce the trained model)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Biggs’s training multiple ML models into Dymetman, Riddle and Fukuda’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Biggs’s training multiple ML models would help to provide more ML training method to Dymetman, Riddle and Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that training multiple ML models would facilitate model training based on the target.
In regard to claim 12, Riddle, Biggs and Fukuda, Dymetman disclose The method of claim 11, the rejection is incorporated herein.
But Fukuda and Riddle, Dymetman fail to explicitly disclose “wherein the first trained model and the second trained model are part of a group of trained models, and further comprising: training, by the system, a third trained model that is configured to process first inputs of the first risk type using the first trained model and second inputs of the second risk type using the second trained model; and wherein producing, by the first trained model, the output is performed with the third trained model.
Biggs disclose wherein the first trained model and the second trained model are part of a group of trained models, ([0007]-[0009] [0020]-[0028] [0034]-[0052] the first and second trained models are integrated together) and further comprising: training, by the system, a third trained model that is configured to process first inputs of the first risk type using the first trained model and second inputs of the second risk type using the second trained model; and wherein producing, by the first trained model, the output is performed with the third trained model. ([0007]-[0009] [0020]-[0028] [0034]-[0052] training a first AI model, the first model is associated with a likelihood of a desired outcome for a given action, which is risk minimization problem associated with features of product or a customer, for example, purchase an item, the second model with a prescriptive tree and trained using labeled data (disclosed by Sandler) and generate interpretable prescriptive policies (based on a set of rules defined) to determine the expected outcome for the interpretable prescriptive policy which is a second risk type and this is user defined, for example, expected revenue maximization to produce the different trained models)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Biggs’s training multiple ML models into Dymetman, Riddle and Fukuda’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Biggs’s training multiple ML models would help to provide more ML training method to Dymetman, Riddle and Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that training multiple ML models would facilitate model training based on the target.
In regard to claim 13, Riddle, Biggs and Fukuda, Dymetman disclose The method of claim 8, the rejection is incorporated herein.
But Riddle and Fukuda, Dymetman fail to explicitly disclose “wherein a feature of the features of users and products comprises a geographical location associated with the user.”
Biggs disclose wherein a feature of the features of users and products comprises a geographical location associated with the user. ([0033]-[0052] groups of customers (with location, age, gender, etc. features))
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Biggs’s training multiple ML models into Dymetman, Riddle and Fukuda’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Biggs’s training data with user features would help to provide more ML training data into Dymetman, Riddle and Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more ML training data with user features would facilitate model training based on the target user.
In regard to claim 17, claim 17 is a medium claim corresponding to the method claim 8 above and, therefore, is rejected for the same reasons set forth in the rejections of claim 8.
Claims 4-6 are rejected under 35 U.S.C. 103 as being unpatentable over Riddle et al. (Riddle) US 2021/0004706, Biggs et al. (Biggs) US 2022/0180168, Fukuda et al. (Fukuda) US 2017/0178168 and Dymetman et al. (Dymetman) US 2013/0096877
as applied to claim 1, further in view of Sandler et al. (Sandler) US 2020/0380545
In regard to claim 4, Riddle and Biggs, Fukuda, Dymetman disclose The system of claim 1, the rejection is incorporated herein.
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein the operations further comprise: while training the artificial intelligence risk model, determining that a first value of the predicted support cost is less than a predetermined threshold value; and increasing a penalty associated with the first value.”
Sandler disclose wherein the operations further comprise: while training the artificial intelligence risk model, determining that a first value of the predicted support cost is less than a predetermined threshold value; and increasing a penalty associated with the first value. ([0042]-[0048] training the ML model, determine of the quantity of the product available when the user-specific price was determined for purchase is less than a threshold amount, the user-specific price may be increased)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Sandler’s method of ML training into Dymetman, Biggs, Fukuda and Riddle’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Sandler’s method of ML training with a threshold value would help to provide more training criteria into Dymetman, Biggs, Fukuda and Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more training criteria would facilitate the finetuning of the ML model and improve the output accuracy of the ML model.
In regard to claim 5, Riddle and Biggs, Fukuda, Dymetman disclose The system of claim 1, the rejection is incorporated herein.
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein the operations further comprise: while training the artificial intelligence risk model, determining that a first value of the predicted support cost is greater than a predetermined threshold value; and decreasing a penalty associated with the first value.”
Sandler disclose wherein the operations further comprise: while training the artificial intelligence risk model, determining that a first value of the predicted support cost is greater than a predetermined threshold value; and decreasing a penalty associated with the first value. ([0042]-[0048] training the ML model, determine of the quantity of the product available when the user-specific price was determined for purchase is greater than a threshold amount, the user-specific price may be decreased)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Sandler’s method of ML training into Dymetman, Biggs, Fukuda and Riddle’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Sandler’s method of ML training with a threshold value would help to provide more training criteria into Dymetman, Biggs, Fukuda and Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more training criteria would facilitate the finetuning of the ML model and improve the output accuracy of the ML model.
In regard to claim 6, Riddle and Biggs, Fukuda, Dymetman disclose The system of claim 1, the rejection is incorporated herein.
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein a feature of the features of users and products comprises an industry of the user.”
Sandler disclose wherein a feature of the features of users and products comprises an industry of the user. ([0021-[0028] user profile include various user information, such as job, etc. although not mentioned, but this is a very common attribute of the user included in the user profile which is not an invention)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Sandler’s method of ML training into Dymetman, Biggs,Fukuda and Riddle’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Sandler’s user feature would help to provide more training data feature into Dymetman, Biggs,Fukuda and Riddle’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more training data feature would facilitate the finetuning of the ML model and improve the output accuracy of the ML model.
Claim 2 is rejected under 35 U.S.C. 103 as being unpatentable over
Riddle et al. (Riddle) US 2021/0004706, Biggs et al. (Biggs) US 2022/0180168 and Fukuda et al. (Fukuda) US 2017/0178168, and Dymetman et al. (Dymetman) US 2013/0096877 as applied to claim 1, further in view of Chen et al. (Chen) US 2022/0405623
In regard to claim 2, Riddle and Biggs, Fukuda, Dymetman disclose The system of claim 1, the rejection is incorporated herein.
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein the trained model is configured to provide a user-understandable explanation regarding why the trained model produced the output.”
Chen disclose wherein the trained model is configured to provide a user-understandable explanation regarding why the trained model produced the output. ([0003]-[0017] [0027]-[0033] provide the explanation for the output generated by the ML model)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Chen’s explainable AI into Dymetman, Riddle and Biggs, Fukuda’s invention as they are related to the same field endeavor of training ML model. The motivation to combine these arts, as proposed above, at least because Chen’s providing why the output is generated based on the input would help to provide more ML training reasoning method to Dymetman, Riddle and Biggs, Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing the ML explanation to the output generated based on the input would facilitate ML model development and decision-making and improve the output accuracy of the ML model.
Claims 7, 14-16 are rejected under 35 U.S.C. 103 as being unpatentable over
Riddle et al. (Riddle) US 2021/0004706, Biggs et al. (Biggs) US 2022/0180168, Fukuda et al. (Fukuda) US 2017/0178168 and Dymetman et al. (Dymetman) US 2013/0096877 as applied to claim 1, further in view of DeFranks et al. (DeFranks) US 2019/0208918.
In regard to claim 7, Riddle and Biggs, Fukuda, Dymetman disclose The system of claim 1, the rejection is incorporated herein.
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein a feature of the features of users and products comprises a size of an entity associated with the user.”
DeFranks disclose wherein a feature of the features of users and products comprises a size of an entity associated with the user. ([0021]-[0029] [0036] [0042] the bedding system including king size or queen size, etc. bed with size of the mattress, etc. associated with the user)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate DeFranks’s active comfort controlled bedding system into Dymetman, Riddle and Biggs, Fukuda’s invention as they are related to the same field endeavor of using feature data of the products. The motivation to combine these arts, as proposed above, at least because DeFranks’s active comfort controlled bedding system would help to provide more product feature data into Dymetman, Riddle and Biggs, Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more product features would help to train the ML model.
In regard to claim 14, Riddle and Biggs, Fukuda, Dymetman disclose The method of claim 8, the rejection is incorporated herein.
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein a feature of the features of users and products comprises a dust removal capability of a geographical location associated with the user.”
DeFranks disclose wherein a feature of the features of users and products comprises a dust removal capability of a geographical location associated with the user. ([0006]-[0008] [0047] [0048] reverse air flow feature of the air blower assembly can remove dust particles and it is at a location associated with the user)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate DeFranks’s active comfort controlled bedding system into Dymetman, Riddle and Biggs, Fukuda’s invention as they are related to the same field endeavor of using feature data of the products. The motivation to combine these arts, as proposed above, at least because DeFranks’s active comfort controlled bedding system would help to provide more product feature data into Dymetman, Riddle and Biggs, Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more product features would help to train the ML model.
In regard to claim 15, Riddle and Biggs, Fukuda, Dymetman disclose The method of claim 8, the rejection is incorporated herein.
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein a feature of the features of users and products comprises a cooling capability of a geographical location associated with the user.”
DeFranks disclose wherein a feature of the features of users and products comprises a cooling capability of a geographical location associated with the user. ([0006]-[0008] [0047] [0048] air blower assembly can provide cooling and it is at a location associated with the user)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate DeFranks’s active comfort controlled bedding system into Dymetman, Riddle and Biggs, Fukuda’s invention as they are related to the same field endeavor of using feature data of the products. The motivation to combine these arts, as proposed above, at least because DeFranks’s active comfort controlled bedding system would help to provide more product feature data into Dymetman, Riddle and Biggs, Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more product features would help to train the ML model.
In regard to claim 16, Riddle and Biggs, Fukuda, Dymetman disclose The method of claim 8, the rejection is incorporated herein.
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein a feature of the features of users and products comprises an electrical infrastructure of a geographical location associated with the user.”
DeFranks disclose wherein a feature of the features of users and products comprises an electrical infrastructure of a geographical location associated with the user. ([0006]-[0008] [0033] [0034] [0047] [0048][0064] air blower assembly which is a electrical infrastructure and it is at a location associated with the user)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate DeFranks’s active comfort controlled bedding system into Dymetman, Riddle and Biggs, Fukuda’s invention as they are related to the same field endeavor of using feature data of the products. The motivation to combine these arts, as proposed above, at least because DeFranks’s active comfort controlled bedding system would help to provide more product feature data into Dymetman, Riddle and Biggs, Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more product features would help to train the ML model.
Claim 18 is rejected under 35 U.S.C. 103 as being unpatentable over
Riddle et al. (Riddle) US 2021/0004706, Biggs et al. (Biggs) US 2022/0180168, Fukuda et al. (Fukuda) US 2017/0178168, and Dymetman et al. (Dymetman) US 2013/0096877 as applied to claim 1, further in view of Vega et al. (Vega) US 2021/0123771.
In regard to claim 18, Riddle and Biggs, Fukuda, Dymetman disclose The non-transitory computer-readable medium of claim 17, the rejection is incorporated herein.
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein a feature of the features of users and products comprises an electricity usage associated with the product.”
Vega disclose wherein a feature of the features of users and products comprises an electricity usage associated with the product. ([0076]-[0086] [0419] historical usage electricity data associated with the smart device, etc.)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Vega’s optimizing utility consumption into Dymetman, Riddle and Biggs, Fukuda’s invention as they are related to the same field endeavor of using feature data of the products. The motivation to combine these arts, as proposed above, at least because Vega’s optimizing utility consumption with usage data would help to provide more product feature data into Dymetman, Riddle and Biggs, Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more product features would help to train the ML model.
Claim 19 is rejected under 35 U.S.C. 103 as being unpatentable over
Riddle et al. (Riddle) US 2021/0004706, Biggs et al. (Biggs) US 2022/0180168, Fukuda et al. (Fukuda) US 2017/0178168 and Dymetman et al. (Dymetman) US 2013/0096877 as applied to claim 1, further in view of Belchee et al. (Belchee) US 2017/0063865.
In regard to claim 19, Riddle and Biggs, Fukuda, Dymetman disclose The non-transitory computer-readable medium of claim 17, the rejection is incorporated herein.
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein a feature of the features of users and products comprises an indication of whether a program or an operating system associated with the product is current.”
Belchee disclose wherein a feature of the features of users and products comprises an indication of whether a program or an operating system associated with the product is current. ([0013] [0029] [0033] user device may provide indications of characteristics whether user device is running current software version)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Belchee’s characteristics of user devices into Dymetman, Riddle and Biggs, Fukuda’s invention as they are related to the same field endeavor of using feature data of the products. The motivation to combine these arts, as proposed above, at least because Belchee’s software version data would help to provide more product feature data into Dymetman, Riddle and Biggs, Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more product feature data would help to train the ML model.
Claim 21 is rejected under 35 U.S.C. 103 as being unpatentable over
Riddle et al. (Riddle) US 2021/0004706, Biggs et al. (Biggs) US 2022/0180168, Fukuda et al. (Fukuda) US 2017/0178168 and Dymetman et al. (Dymetman) US 2013/0096877as applied to claim 1, further in view of Kolen et al. (Kolen) US 2020/0306632.
In regard to claim 21, Riddle and Biggs, Fukuda, Dymetman disclose The system of claim 1, the rejection is incorporated herein.
But Biggs, Fukuda and Dymetman fail to explicitly disclose “predicting a lower maintenance cost than an observed maintenance cost by an amount; and predicting a higher maintenance cost than the observed maintenance cost by the amount;”
Riddle disclose predicting a maintenance cost with an observed maintenance cost by an amount; and predicting a maintenance cost with the observed maintenance cost by the amount; ([0014]-[0026] [0035]-[0044] [0094][0095] [0106]-[0108] generate expected cost of an escalation for each customer and with observed service escalation by an amount, for example hour of time, etc. )
But Riddle and Biggs, Fukuda, Dymetman fail to explicitly disclose “wherein the training of the artificial intelligence risk model to produce a trained model comprises: implementing a first training penalty in training the artificial intelligence risk model, wherein the first training penalty corresponds to predict a lower value than the observed value by the amount; and implementing a second training penalty in training the artificial intelligence risk model, wherein the second training penalty corresponds to predicting a higher value than the observed value, wherein the first training penalty is greater than the second training penalty.”
Kolen disclose wherein the training of the artificial intelligence risk model to produce a trained model comprises: implementing a first training penalty in training the artificial intelligence risk model, wherein the first training penalty corresponds to predict a lower value than the observed value by the amount; and implementing a second training penalty in training the artificial intelligence risk model, wherein the second training penalty corresponds to predicting a higher value than the observed value, wherein the first training penalty is greater than the second training penalty. ([0078]-[0089] [0097]-[0098] the penalty can be associated with threshold degree of accuracy, above or below the threshold, etc. with lower or higher penalties. Note: higher or lower penalties can be based on the implementation which is a design choice)
It would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made to incorporate Kolen’s predictive execution of game engines into Dymetman, Riddle and Biggs, Fukuda’s invention as they are related to the same field endeavor of using feature data of the products in training ML model. The motivation to combine these arts, as proposed above, at least because Kolen’s ML model with training penalty would help to provide more training criteria into Dymetman, Riddle and Biggs, Fukuda’s system. Therefore it would have been obvious to one having ordinary skill in the art before the effective filing data of the claimed invention was made that providing more training criteria with penalty would help to provide more accurate predictions to the user by the trained ML model.
Response to Arguments
Applicant’s arguments with respect to claims 1-19, 21 filed on 4/28/2026 have been considered but are moot because the arguments do not apply to the current rejection.
Conclusion
The prior art made of record and not relied upon is considered pertinent to Applicant's disclosure.
PATENT PUB. # PUB. DATE INVENTOR(S) TITLE
US 20160180372 A1 2016-06-23 Lu et al.
SYSTEMS AND METHODS FOR ONLINE ADVERTISEMENT REALIZATION PREDICTION
Lu et al. disclose a computer system implementing a method for ad realization prediction may be configured to receive a plurality of target realization factors associated with a target ad display opportunity; determine a reference realization probability score of the target ad display opportunity based on a global reference realization probability distribution associated with an ad display realization probability decision tree; using the reference realization probability score, determine an ad realization probability score of the target ad display opportunity according to a piecewise calibrated realization probability function; and return the ad realization probability score… see abstract.
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to XUYANG XIA whose telephone number is (571)270-3045. The examiner can normally be reached Monday-Friday 8am-4pm.
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, Jennifer Welch can be reached at 571-272-7212. 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.
XUYANG XIA
Primary Examiner
Art Unit 2143
/XUYANG XIA/Primary Examiner, Art Unit 2143
Prosecution Insights