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 .
DETAILED ACTION
2. This office action is in response to the original filing of 02/29/2024. Claims 1-20 are pending and have been considered below.
Claim Rejections - 35 USC § 101
3. 35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to abstract ideas without significantly more.
Claim 1:
Step 1: The claim is directed to a method, falling under one of the four statutory categories of invention.
Step 2A Prong 1: The claim recites following abstract ideas:
The limitations “utilizing, by the device, a counterfactual model to generate counterfactuals for the trained model based on the training data and the query”; “determining, by the device, contextual parameters associated with ranking the counterfactuals”; “utilizing, by the device, a counterfactual ranking model to calculate a weighted metric based on the query, the counterfactuals, and the contextual parameters; “utilizing, by the device, the counterfactual ranking model to apply the weighted metric to the counterfactuals to generate ranked counterfactuals for the trained model ” under broadest reasonable interpretation covers a mental process including an observation, evaluation, judgment or opinion that could be performed in the human mind or with the aid of pencil and paper.
2A – Prong 2: This judicial exception is not integrated into a practical application. In particular, claim 1 recites the additional elements:
“receiving, by a device, a trained model, training data utilized to train the trained model”, and “a query associated with generating counterfactuals amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g)).; and performing, by the device, one or more actions based on the ranked counterfactuals” amounts to no more than mere instructions to apply the exception using a generic computer component. The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). As explained by the Supreme Court, a claim directed to a judicial exception cannot be made eligible "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use." Diamond v. Diehr, 450 U.S. 175, 192 n.14, 209 USPQ 1, 10 n. 14 (1981). Thus, limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea.
Claim 8:
Step 1: The claim is directed to a method, falling under one of the four statutory categories of invention.
Step 2A Prong 1: The claim recites following abstract ideas:
The limitations “generate counterfactuals for the trained model based on the training data and the query”; “determine, by the device, contextual parameters associated with ranking the counterfactuals”; under broadest reasonable interpretation covers a mental process including an observation, evaluation, judgment or opinion that could be performed in the human mind or with the aid of pencil and paper.
“calculate a weighted metric based on the query, the counterfactuals, and the contextual parameters“ 2106.04(a)(2)(I)(C) “Mathematical Calculations A claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. There is no particular word or set of words that indicates a claim recites a mathematical calculation. That is, a claim does not have to recite the word "calculating" in order to be considered a mathematical calculation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation.”
2A – Prong 2: This judicial exception is not integrated into a practical application. In particular, claim 8 recites the additional elements:
”apply the weighted metric to the counterfactuals to generate ranked counterfactuals for the trained model ”; and performing, by the device, one or more actions based on the ranked counterfactuals” amounts to no more than mere instructions to apply the exception using a generic computer component. The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). As explained by the Supreme Court, a claim directed to a judicial exception cannot be made eligible "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use." Diamond v. Diehr, 450 U.S. 175, 192 n.14, 209 USPQ 1, 10 n. 14 (1981). Thus, limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea.
“one or more processors configured to”, amount to elements that have been recognized as well-understood, routine, and conventional activity in particular fields, as demonstrate by: Relevant court decision: the followings are examples of court decisions demonstrating well-understood, routine and conventional activities, see e.g., MPEP 2106.05(d)(II) and MPEP 2106.05(f)(2): Computer readable storage media comprising instructions to implement a method, e.g., see Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015).
The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). As explained by the Supreme Court, a claim directed to a judicial exception cannot be made eligible "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use." Diamond v. Diehr, 450 U.S. 175, 192 n.14, 209 USPQ 1, 10 n. 14 (1981). Thus, limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea.
2B Analysis: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
”apply the weighted metric to the counterfactuals to generate ranked counterfactuals for the trained model ”; and performing, by the device, one or more actions based on the ranked counterfactuals” amounts to no more than mere instructions to apply the exception using a generic computer component. The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). As explained by the Supreme Court, a claim directed to a judicial exception cannot be made eligible "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use." Diamond v. Diehr, 450 U.S. 175, 192 n.14, 209 USPQ 1, 10 n. 14 (1981). Thus, limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea.
“one or more processors configured to”, amount to elements that have been recognized as well-understood, routine, and conventional activity in particular fields, as demonstrate by: Relevant court decision: the followings are examples of court decisions demonstrating well-understood, routine and conventional activities, see e.g., MPEP 2106.05(d)(II) and MPEP 2106.05(f)(2): Computer readable storage media comprising instructions to implement a method, e.g., see Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015).
The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). As explained by the Supreme Court, a claim directed to a judicial exception cannot be made eligible "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use." Diamond v. Diehr, 450 U.S. 175, 192 n.14, 209 USPQ 1, 10 n. 14 (1981). Thus, limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea.
Claim 15:
Step 1: The claim is directed to a method, falling under one of the four statutory categories of invention.
Step 2A Prong 1: The claim recites following abstract ideas:
The limitations “utilizing, by the device, a counterfactual model to generate counterfactuals for the trained model based on the training data and the query”; “determining, by the device, contextual parameters associated with ranking the counterfactuals”; “utilizing, by the device, a counterfactual ranking model to calculate a weighted metric based on the query, the counterfactuals, and the contextual parameters; “utilizing, by the device, the counterfactual ranking model to apply the weighted metric to the counterfactuals to generate ranked counterfactuals for the trained model ” under broadest reasonable interpretation covers a mental process including an observation, evaluation, judgment or opinion that could be performed in the human mind or with the aid of pencil and paper.
2A – Prong 2: This judicial exception is not integrated into a practical application. In particular, claim 15 recites the additional elements:
“receiving, by a device, a trained model, training data utilized to train the trained model”, and “a query associated with generating counterfactuals amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g)).; and performing, by the device, one or more actions based on the ranked counterfactuals” amounts to no more than mere instructions to apply the exception using a generic computer component. The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). As explained by the Supreme Court, a claim directed to a judicial exception cannot be made eligible "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use." Diamond v. Diehr, 450 U.S. 175, 192 n.14, 209 USPQ 1, 10 n. 14 (1981). Thus, limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea.
“one or more processors” and “..readable medium” amount to elements that have been recognized as well-understood, routine, and conventional activity in particular fields, as demonstrate by: Relevant court decision: the followings are examples of court decisions demonstrating well-understood, routine and conventional activities, see e.g., MPEP 2106.05(d)(II) and MPEP 2106.05(f)(2): Computer readable storage media comprising instructions to implement a method, e.g., see Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015).
The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). As explained by the Supreme Court, a claim directed to a judicial exception cannot be made eligible "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use." Diamond v. Diehr, 450 U.S. 175, 192 n.14, 209 USPQ 1, 10 n. 14 (1981). Thus, limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea.
2B Analysis: The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception.
“receiving, by a device, a trained model, training data utilized to train the trained model”, and “a query associated with generating counterfactuals amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g)).; and performing, by the device, one or more actions based on the ranked counterfactuals” amounts to no more than mere instructions to apply the exception using a generic computer component. The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). As explained by the Supreme Court, a claim directed to a judicial exception cannot be made eligible "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use." Diamond v. Diehr, 450 U.S. 175, 192 n.14, 209 USPQ 1, 10 n. 14 (1981). Thus, limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea.
“one or more processors” and “..readable medium” amount to elements that have been recognized as well-understood, routine, and conventional activity in particular fields, as demonstrate by: Relevant court decision: the followings are examples of court decisions demonstrating well-understood, routine and conventional activities, see e.g., MPEP 2106.05(d)(II) and MPEP 2106.05(f)(2): Computer readable storage media comprising instructions to implement a method, e.g., see Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015).
The courts have found limitations directed to obtaining information electronically, recited at a high level of generality, to be well-understood routine, and conventional (see MPEP 2106.05(d)(II), “receiving or transmitting data over a network”, "electronic record keeping," and "storing and retrieving information in memory"). As explained by the Supreme Court, a claim directed to a judicial exception cannot be made eligible "simply by having the applicant acquiesce to limiting the reach of the patent for the formula to a particular technological use." Diamond v. Diehr, 450 U.S. 175, 192 n.14, 209 USPQ 1, 10 n. 14 (1981). Thus, limitations that amount to merely indicating a field of use or technological environment in which to apply a judicial exception do not amount to significantly more than the exception itself, and cannot integrate a judicial exception into a practical application. Accordingly, these additional elements do not integrate the abstract idea into a practical application because it does not impose any meaningful limits on practicing the abstract idea.
Claim 2 recites “receiving user-defined weightings associated with ranking the counterfactuals; and adjusting the weighted metric based on the user-defined weightings” amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g))..
Claim 3 recites “filtering the ranked counterfactuals based on a threshold and to generate a subset of top ranked counterfactuals” under broadest reasonable interpretation covers a mental process including an observation, evaluation, judgment or opinion that could be performed in the human mind or with the aid of pencil and paper.
Claim 4 recites “wherein the weighted metric is based on costs associated with implementing the counterfactuals, similarities of the counterfactuals to the query, and effects of the counterfactuals on a predicted outcome” amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g)).
Claim 5 recites “wherein utilizing the counterfactual ranking model to calculate the weighted metric comprises: calculating a cost metric for the counterfactuals; calculating a similarity metric for the counterfactuals; and calculating an outcome effect metric for the counterfactuals, wherein the weighted metric corresponds to a combination of the cost metric, the similarity metric, and the outcome effect metric” 2106.04(a)(2)(I)(C) “Mathematical Calculations A claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. There is no particular word or set of words that indicates a claim recites a mathematical calculation. That is, a claim does not have to recite the word "calculating" in order to be considered a mathematical calculation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation.
Claim 6 recites “wherein utilizing the counterfactual ranking model to calculate the weighted metric comprises: calculating a first metric that represents a quantity of perturbed features between the query and each of the counterfactuals; calculating a second metric that represents a spatial distance between the query and each of the counterfactuals; calculating a third metric that represents a difference in a probability value between the query and each of the counterfactuals; normalizing the first metric, the second metric, and the third metric to generate a normalized first metric, a normalized second metric, and a normalized third metric, respectively; and calculating the weighted metric based on the normalized first metric, the normalized second metric, and the normalized third metric” 2106.04(a)(2)(I)(C) “Mathematical Calculations A claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. There is no particular word or set of words that indicates a claim recites a mathematical calculation. That is, a claim does not have to recite the word "calculating" in order to be considered a mathematical calculation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation.
Claim 7 recites “wherein utilizing the counterfactual ranking model to calculate the weighted metric comprises: calculating metrics based on the query, the counterfactuals, and the contextual parameters; and applying ranking weights to the metrics to calculate the weighted metric” 2106.04(a)(2)(I)(C) “Mathematical Calculations A claim that recites a mathematical calculation, when the claim is given its broadest reasonable interpretation in light of the specification, will be considered as falling within the "mathematical concepts" grouping. A mathematical calculation is a mathematical operation (such as multiplication) or an act of calculating using mathematical methods to determine a variable or number, e.g., performing an arithmetic operation such as exponentiation. There is no particular word or set of words that indicates a claim recites a mathematical calculation. That is, a claim does not have to recite the word "calculating" in order to be considered a mathematical calculation. For example, a step of "determining" a variable or number using mathematical methods or "performing" a mathematical operation may also be considered mathematical calculations when the broadest reasonable interpretation of the claim in light of the specification encompasses a mathematical calculation.
Claim 9 recites “wherein each of the counterfactuals represents a minimal change in input data for the trained model to alter an output of the trained model” amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g)).
Claim 10 recites “wherein the one or more processors are further configured to: normalize the contextual parameters prior to utilizing the contextual parameters to calculate the weighted metric” amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g))..
Claim 11 recites “wherein the one or more processors, to perform the one or more actions, are configured to one or more of: provide the ranked counterfactuals for display; or integrate the ranked counterfactuals in a real-time decision support system” amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g)).
Claim 12 recites “wherein the one or more processors, to perform the one or more actions, are configured to one or more of: utilize one of the ranked counterfactuals to generate a strategy for customer retention; or utilize one of the ranked counterfactuals for a real-time customer service system” amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g)).
Claim 13 recites “wherein the one or more processors, to perform the one or more actions, are configured to: receive feedback about the ranked counterfactuals; and modify a counterfactual ranking model based on the feedback”. amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g)).
Claim 14 recites “wherein the one or more processors, to perform the one or more actions, are configured to: utilize one of the ranked counterfactuals for predictive maintenance of a network” amount to insignificant extra solution activity like mere data gathering, MPEP 2106.05(g)).
Claims 16-20 contain subject matter similar to that of claims 2-3, 5-7, respectively and are rejected on the same grounds.
Claim Rejections - 35 USC § 103
4. 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.
Claim(s) 1, 4-6, 8, 10, 15, and 18-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Dandl et al. (Multi-Objective Counterfactual Explanations) in view of Argawal et al. (A General Framework for Counterfactual Learning-to-Rank).
Claim 1. Dandle discloses a method, comprising:
receiving, by a device, a trained model, training data utilized to train the trained model, and a query associated with generating counterfactuals (..an already fitted/model-agnostic prediction function and observed training data §4.1 expressly identifies (\hat f) as the prediction function and (X^{obs}) as the observed, i.e., training, data.§4.1, pp. 452–454, particularly Eq. (1) and accompanying text…explains that model-agnostic counterfactual techniques operate using an already fitted model §2, p. 451);
utilizing, by the device, a counterfactual model to generate counterfactuals for the trained model based on the training data and the query (..generate counterfactual data points through its MOC methodology §4.1, pp. 452–454, and §4.2, pp. 454–456. Specifically, evaluate candidate feature vectors according to the four counterfactual objectives and uses NSGA-II to search the candidate space §4.2, pp. 454–456.);
determining, by the device, contextual parameters associated with ranking the counterfactuals (determining multiple quantitative objectives for each counterfactual (In particular: (o_1) measures distance between the model prediction and desired outcome; (o_2) measures distance between the original and counterfactual feature vectors; (o_3) counts changed features; and (o_4) measures distance from observed training data §4.1, Eq. (1), pp. 452–454.);
utilizing, by the device, a counterfactual ranking model to calculate a weighted metric based on the query, the counterfactuals, and the contextual parameters (Dandl acknowledges that prior counterfactual approaches optimize a collapsed, weighted sum of multiple objectives §3, pp. 450–452, §4.2, 4.3, 4.5); and
performing, by the device, one or more actions based on the ranked counterfactuals (Dandl expressly states that counterfactuals provide actionable options for changing a predicted outcome §1, pp. 449–450, and §3, pp. 450–452, 4.3).
Dandl does not explicitly disclose utilizing, by the device, the counterfactual ranking model to apply the weighted metric to the counterfactuals to generate ranked counterfactuals for the trained model.
However, Agarwal teaches utilizing, by the device, the counterfactual ranking model to apply the weighted metric to the counterfactuals to generate ranked counterfactuals for the trained model (weighted ranking metrics, including metrics expressed as individual relevance values multiplied by a rank-weighting function §1-4.1, pp. 5–8 and the additive-metric formulation). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Dandl further in view of Agarwal to incorporate the above cited feature. One would have been motivated to do so to provide a quantitative ranking of multiple candidate counterfactual explanations according to both counterfactual quality and ranking performance.
Claim 4. Dandl and Agarwal disclose the method of claim 1, Dandl further discloses wherein the weighted metric is based on costs associated with implementing the counterfactuals, similarities of the counterfactuals to the query, and effects of the counterfactuals on a predicted outcome (First, (o_3) measures the number of changed features, which directly represents the amount of modification required to implement a counterfactual. §4.1, p. 453, Eq. (1). Second, (o_2) measures the distance between the original and counterfactual inputs using Gower distance. §4.1, pp. 452–453. Dandl expressly gives: (o_2(x,x^)) quantifies the distance between (x^) and (x). §4.1, pp. 452–453. Third, (o_1) measures the distance between the counterfactual prediction and the desired outcome. §4.1, pp. 452–453).
Claim 5. Dandl and Agarwal disclose the method of claim 1, Dandl further discloses wherein utilizing the counterfactual ranking model to calculate the weighted metric comprises: calculating a cost metric for the counterfactuals; calculating a similarity metric for the counterfactuals; and calculating an outcome effect metric for the counterfactuals, wherein the weighted metric corresponds to a combination of the cost metric, the similarity metric, and the outcome effect metric ( outcome-distance objective (o_1), §4.1, pp. 452–453; input-distance objective (o_2), §4.1, pp. 452–453; changed-feature objective (o_3), §4.1, p. 453; and training-data proximity objective (o_4), §4.1, pp. 453–454. Dandl further expressly recognizes the use of weighted combinations of multiple objectives in the counterfactual field §3, pp. 450–452.) and Agarwal teaches a ranking metric formed from weighted components, pp. 5–14. One would have been motivated to do so to yield a ranking score incorporating the respective costs, distances/similarity, and predicted-outcome effects
Claim 6. Dandl and Agarwal disclose the method of claim 1, Dandl further discloses wherein utilizing the counterfactual ranking model to calculate the weighted metric comprises: calculating a first metric that represents a quantity of perturbed features between the query and each of the counterfactuals; calculating a second metric that represents a spatial distance between the query and each of the counterfactuals; calculating a third metric that represents a difference in a probability value between the query and each of the counterfactuals; normalizing the first metric, the second metric, and the third metric to generate a normalized first metric, a normalized second metric, and a normalized third metric, respectively (Quantity of perturbed features (o_3) is exactly the number of changed features and is defined using the (L_0) norm: (o_3(x,x^) = ||x-x^||_0) §4.1, p. 453, Eq. (1). Distance (o_2) is a Gower-distance measure between the original and counterfactual inputs, §4.1, pp. 452–453. Probability/outcome difference For classification models, Dandl expressly assumes that the prediction function returns the probability for a selected class §4.1, p. 452. The first objective (o_1) then measures the distance between that predicted probability and the desired probability range, §4.1, pp. 452–453. Normalization Dandl additionally teaches explicit normalization in §4.3. It transforms feature-importance standard deviations into probabilities using a min/max normalization equation: (P(value\ differs)) is calculated from the minimum and maximum ICE standard deviations and mapped into a prescribed probability range, §4.3, p. 456); and Agarwal discloses calculating the weighted metric based on the normalized first metric, the normalized second metric, and the normalized third metric (weighted ranking metrics, pp. 5–14 while Dandl expressly recognizes weighted combinations of multiple objectives §3, pp. 450–452;). One would have been motivated to do so that the different metrics could be meaningfully aggregated.
Claim 10. Dandl and Agarwal disclose the device of claim 8, wherein the one or more processors are further configured to: normalize the contextual parameters prior to utilizing the contextual parameters (§4.3, “Further Modifications—Initialization,” Dandl explains that feature importance for an individual prediction can be measured using the standard deviation of an ICE curve. Dandl then states that the standard deviation for each feature is transformed into probabilities within a specified range, §4.3, p. 456, lines corresponding to the discussion of initialization; Eq. following the discussion of σjICE\sigma_j^{ICE}) …. normalization/transformation equation for P(value differs)P(\text{value differs}) Eq. immediately following the discussion of σjICE\sigma_j^{ICE} §4.3, p. 456)
Agarwal further discloses to calculate the weighted metric (Agarwal supplies the ranking side. The paper describes a counterfactual learning-to-rank framework for unbiased training using propensity-weighted rank-based metrics, including DCG, and applies the framework to ranking functions including deep networks, Abstract; §§3–5). One would have been motivated to do so to normalize places heterogeneous parameters on a common scale and facilitate meaningful weighted aggregation.
Claims 8, 15 and 18-19 represent the device and medium of claims 1 and 5-6 and are rejected along the same rationale.
5. Claim(s) 2-3, 9, 12-14 and 16-17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Dandl et al. (Multi-Objective Counterfactual Explanations) in view of Argawal et al. (A General Framework for Counterfactual Learning-to-Rank) and further in view of Sharpe et al. (US20240112072).
Claim 2. Dandl and Agarwal disclose the method of claim 1, but fail to explicitly disclose comprising: receiving user-defined weightings associated with ranking the counterfactuals; and adjusting the weighted metric based on the user-defined weightings.
However, Sharpe discloses obtaining a user preference and applying an adjustment parameter in response to that preference... In particular, obtaining a ranked ordering of features representing a user preference and applying an adjustment parameter to features based on that preference ([0049],[0095])… further teaches proportional adjustment parameters and parameters for an objective function based on the user preference ([0050]). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Dandl further in view of Sharpe to incorporate the above cited feature. One would have been motivated to do so to permit the selection according to user-specific or application-specific.
Claim 3. Dandl and Agarwal disclose the method of claim 1, but fail to explicitly disclose further comprising: filtering the ranked counterfactuals based on a threshold and to generate a subset of top ranked counterfactuals.
However, Sharpe discloses Sharpe similarly expressly compares a model probability with a threshold and determines whether the threshold is satisfied ([0024], [0040], [0052]). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Dandl further in view of Sharpe to incorporate the above cited feature. One would have been motivated to do so to satisfy a desired rank change.
Claim 9. Dandl and Agarwal disclose the device of claim 8, but fail to explicitly disclose wherein each of the counterfactuals represents a minimal change in input data for the trained model to alter an output of the trained model.
However, Sharpe discloses a counterfactual sample may minimize the amount of change to the original feature values while still changing the machine-learning output ([0002], [0025], [0042]). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Dandl further in view of Sharpe to incorporate the above cited feature. One would have been motivated to do so to satisfy a desired rank change.
Claim 12. Dandl and Agarwal disclose the device of claim 8, but fail to explicitly disclose wherein the one or more processors, to perform the one or more actions, are configured to one or more of: utilize one of the ranked counterfactuals to generate a strategy for customer retention; or utilize one of the ranked counterfactuals for a real-time customer service system.
However, Sharpe discloses utilize one of the ranked counterfactuals to generate a strategy for customer retention (Sharpe describes generating a counterfactual for a user whose banking request was rejected and providing the user with a recommendation identifying changes that could result in approval… further teaches that the recommendation may identify a specific action the user should take..)([0035][0077],[0085]). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Dandl further in view of Sharpe to incorporate the above cited feature. One would have been motivated to do so to establish trust in the output generated by machine learning models.
Claim 13. Dandl and Agarwal disclose the device of claim 8, but fail to explicitly disclose wherein the one or more processors, to perform the one or more actions, are configured to: receive feedback about the ranked counterfactuals; and modify a counterfactual ranking model based on the feedback.
However, Sharpe discloses receive feedback about the ranked counterfactuals (model outputs can be fed back into the model as training input, including together with user indications concerning output accuracy or other reference feedback information..) ([0061); and modify a counterfactual ranking model based on the feedback (training a second model using a loss function associated with counterfactual samples and adjusting model weights) ([0041],[0085]).. ])(update its configurations, including weights, biases, or other parameters, based on prediction assessment and reference feedback information.) ([0062]). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Dandl further in view of Sharpe to incorporate the above cited feature. One would have been motivated to do so to establish trust in the output generated by machine learning models.
Claim 14. Dandl and Agarwal disclose the device of claim 8, but fail to explicitly disclose wherein the one or more processors, to perform the one or more actions, are configured to: utilize one of the ranked counterfactuals for predictive maintenance of a network.
However, Sharpe discloses wherein the one or more processors, to perform the one or more actions, are configured to: utilize one of the ranked counterfactuals for predictive maintenance of a network ([0006],[0010]). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Dandl further in view of Sharpe to incorporate the above cited feature. One would have been motivated to do so to establish trust in the output generated by machine learning models.
Claims 16, 17 represent the device and medium of claims 2-3 and are rejected along the same
rationale.
6. Claim(s) 7, 11 and 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Dandl et al. (Multi-Objective Counterfactual Explanations) in view of Argawal et al. (A General Framework for Counterfactual Learning-to-Rank) and further in view of Salimiparsa (Counterfactual Explanations for Rankings, 2023).
Claim 7. Dandl and Agarwal disclose the method of claim 1, wherein utilizing the counterfactual ranking model to calculate the weighted metric comprises: but fail to explicitly disclose calculating metrics based on the query, the counterfactuals, and the contextual parameters; and applying ranking weights to the metrics to calculate the weighted metric.
However, Salimiparsa provides the ranking-specific application by calculating the rank of each generated counterfactual and comparing it against the original rank, pp. 2-3, §2-3. Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Dandl further in view of Salimiparsa to incorporate the above cited feature. One would have been motivated to do so to determine metrics from the relevant counterfactual/model information and applying ranking weights to produce an ordered set of candidate counterfactuals.
Claim 11. Dandl and Agarwal disclose the device of claim 8, but fail to explicitly disclose wherein the one or more processors, to perform the one or more actions, are configured to one or more of: provide the ranked counterfactuals for display; or integrate the ranked counterfactuals in a real-time decision support system.
However, Salimiparsa provide the ranked counterfactuals for display (method explains relative rankings; uses counterfactual examples; changes feature values to affect an entity's position within a ranking; and is directed toward decision-support applications (p. 1, Abstract,)…receives a machine-learning model, a data instance, and a desired rank change (p. 2, §3)…( The paper then expressly reports that users can request that an item be ranked higher or lower and that:
“the system will display the required changes to feature values.”) (p. 4, §5). Therefore, it would have been obvious to someone of ordinary skill in the art before the effective filing date of the claimed invention to have modified Dandl further in view of Salimiparsa to incorporate the above cited feature. One would have been motivated to do so to communicate counterfactual explanations of rankings to users and display the feature changes required to obtain a desired ranking.
Claim 20 represents the medium of claim 7 and is rejected along the same rationale.
Conclusion
7. The prior art made of record and not relied upon is considered pertinent to applicant's disclosure (See PTO-892).
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Phenuel S. Salomon whose telephone number is (571) 270-1699. The examiner can normally be reached on Mon-Fri 7:00 A.M. to 4:00 P.M. (Alternate Friday Off) EST.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Usmaan Saeed can be reached on (571) 272-4046. The fax phone number for the organization where this application or proceeding is assigned is 571-273-3800.
Information regarding the status of an application may be obtained from the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/PHENUEL S SALOMON/Primary Examiner, Art Unit 2146
Prosecution Insights