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 05/29/2024. Claims 1-20 are pending and have been considered below.
Claim Rejections - 35 USC § 103
3. In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
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.
4. Claims 1, 3-7, 9-10, 12-16 and 18-19 are rejected under 35 U.S.C. 103 as being unpatentable over Amoukou et al. (Local and Regional Counterfactual Rules: Summarized and Robust Recourses, September 29, 2022) in view of Liu et al. (US 7,853,615).
Claim 1. Amoukou discloses a method for computing regional counterfactual rules to summarize recourse options, the method being implemented by at least one processor, the method comprising:
receiving, by the at least one processor, model data via at least one input, the model data including information that relates to a target model and a corresponding data set (deriving counterfactual rules using random forests and data-distribution information)(abstract).. (defines a dataset and expressly discusses generating recourse for a target model's predictions) (§ 2)..(training a random forest to predict either the model output or the data-generating process) (§ 3);
training, by the at least one processor, a surrogate model for the target model based on the model data (a dedicated random forest that learns to predict the output of another model) (§ 3);
identifying, by the at least one processor using the surrogate model, at least one rule for an output of the target model based on a predetermined threshold, each of the at least one rule including a counterfactual rule (local counterfactual rules that change predictions with high probability, selecting small feature subsets and favoring high-density regions)(abstract)..(counterfactual decision probability, a minimum probability threshold and rectangular counterfactual rules) (§ 2, Definitions 2.1–2.3);
Amoukou § 2 describes rectangular local and regional rules and further describes generalizing overlapping local counterfactual rules into regional rules but does not explicitly disclose enumerating, by the at least one processor using the surrogate model, at least one boundary for each of the identified at least one rule to partition an input space into a grid with at least one cell, each of the at least one cell including a hyperrectangular cell;
labeling, by the at least one processor using the surrogate model, each of the at least one cell with one of the identified at least one rule based on predetermined optimality criteria; and
merging, by the at least one processor using the surrogate model, each of the labeled at least one cell based on a matching of the at least one rule to generate at least one region of optimality.
However, Liu discloses enumerating, by the at least one processor using the surrogate model, at least one boundary for each of the identified at least one rule to partition an input space into a grid with at least one cell, each of the at least one cell including a hyperrectangular cell (creating a partition file, the partition file comprising a list of the hyperrectangular cells with corresponding range values, the list comprising unique cell identifications corresponding with the hyperrectangular cells, the range values comprising intervals of numerical values of the numerical attributes and the non-numeric attributes associated with the corresponding partition cells…) (claim 14, fig 2a-b); labeling, by the at least one processor using the surrogate model, each of the at least one cell with one of the identified at least one rule based on predetermined optimality criteria (mapping the application space at any point of time into hierarchical hyperrectangular cells of variable sizes according to a current state of said application space; and exporting the application space into the hierarchical hyperrectangular cells; wherein said hierarchical structure comprises leaf nodes storing lists of objects whose communication interest are included in the associated partition cell and non-leaf nodes associated with domains of the application space; wherein the step of mapping satisfies an application-dependent constraint; and wherein each of the hierarchical hyperrectangular cells is associated with a domain defined by a numerical attribute and a non-numerical attribute of an object at that time)(claim 9); and
merging, by the at least one processor using the surrogate model, each of the labeled at least one cell based on a matching of the at least one rule to generate at least one region of optimality (col. 8, lines 42-49, 55-67 and col. 90). 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 Amoukou further in view of Liu to incorporate the above cited features. One would have been motivated to do so to support the distribution of state updates.
Claim 3. Amoukou and Liu disclose the method of claim 1, Amoukou further discloses wherein each of the at least one surrogate model corresponds to a tree-based machine learning model that includes a plurality of nodes (§ 3.1 describes a random forest comprising randomized decision trees), Liu further discloses the plurality of nodes including at least one end node that represents a hyperrectangular rule (claim 14, fig 2a-b). One would have been motivated to do so to support the distribution of state updates.
Claim 4. Amoukou and Liu disclose the method of claim 1, Amoukou further discloses wherein the identifying of the at least one rule further comprises: determining, by the at least one processor using the surrogate model, at least one end node based on the model data, wherein each of the at least one end node corresponds to a mapping of the counterfactual rule in the input space; and merging, by the at least one processor using the surrogate model, each the at least one end node based on a matching of at least one characteristic (examining tree leaves to identify directions for modifying predictions… overlapping local counterfactual rectangles and their generalization into regional rules) (§ 2).
Claim 5. Amoukou and Liu disclose the method of claim 1, Amoukou further discloses wherein the predetermined threshold includes a threshold value that defines a minimum probability of a different model output for the output of the target model, the threshold value including a percentage value (§ 2, Definitions 2.1–2.2, pp. 3–4, defines the probability of obtaining a desired alternative outcome and requires it to meet a specified minimum.. uses a probability threshold, π. Expressing that probability as a percentage is a mathematical representation of the same quantity, although the cited passage does not require percentage notation).
Claim 6. Amoukou and Liu disclose the method of claim 1, Amoukou further discloses wherein the predetermined optimality criteria include a sparsity criterion that requires a minimal change in a feature value of the target model to alter a predictive outcome (§ 2, Definition 2.2, pp. 3–4, identifies minimal feature subsets capable of producing a desired alternative outcome).
Claim 7. Amoukou and Liu disclose the method of claim 6, Amoukou further discloses wherein the predetermined optimality criteria include a feasibility criterion that requires selection of one of the identified at least one rule according to a proportion of data points within the identified at least one rule when more than one of the identified at least one rule satisfy the sparsity criterion (§ 2, Definition 2.3, maximizes the probability mass of a counterfactual rectangle subject to a required probability of reaching the desired outcome…further supplies a density-based plausibility objective. Selecting between equally sparse rules by empirical data coverage is a possible implementation, but the cited definition does not expressly establish the claimed tie-breaking procedure).
Claim 9. Amoukou and Liu disclose the method of claim 1, Amoukou further discloses wherein the target model includes at least one from among a deep learning model, a neural network model, a machine learning model, a mathematical model, and a process model (§ 2, expressly discusses recourse for predictions of a given model and is applicable to machine-learning predictors).
Claims 10, 12-16 and 18-19 represent the device and medium of claims 1, 3-7, and 9, respectively and are rejected along the same rationale.
5. Claims 2, 11 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over Amoukou et al. (Local and Regional Counterfactual Rules: Summarized and Robust Recourses, September 29, 2022) in view of Liu et al. (US 7,853,615) and further in view of Rawal et al. (Beyond Individualized Recourse: Interpretable and Interactive Summaries of Actionable Recourses, 2020).
Claim 2. Amoukou and Liu disclose the method of claim 1, further comprising: generating, by the at least one processor using the surrogate model, at least one regional rule that summarizes a plurality of recourse options based on the generated at least one region of optimality (rectangular local and regional rules and further describes generalizing overlapping local counterfactual rules into regional rules)( § 2), but fail to explicitly disclose wherein the at least one regional rule is assembled into a global textual summary of the plurality of recourse options; and wherein the global textual summary includes at least one conditional statement.
However, Rawal discloses wherein the at least one regional rule is assembled into a global textual summary of the plurality of recourse options; and wherein the global textual summary includes at least one conditional statement (a global summary comprising interpretable recourse rules…recourse rules expressed through conditions on subgroup membership and recommended feature changes) (§ 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 Amoukou further in view of Rawal to incorporate the above cited features. One would have been motivated to do so to organize regional rules into conditional statements that identify affected subgroups and their available recourse actions.
Claim 11 and 20 represent the device and medium of claim 2 and are rejected along the same rationale.
6. Claims 8 and 17 are rejected under 35 U.S.C. 103 as being unpatentable over Amoukou et al. (Local and Regional Counterfactual Rules: Summarized and Robust Recourses, September 29, 2022) in view of Liu et al. (US 7,853,615) and further in view of in view of Kanamori et al. (Counterfactual Explanation Trees: Transparent and Consistent Actionable Recourse with Decision Trees, March 2022).
Claim 8. Amoukou and Liu disclose the method of claim 1, Amoukou further discloses comprising: receiving, by the at least one processor, at least one request to generate a local counterfactual explanation for a single instance of the output (fig. 1);
Liu further discloses searching, by the at least one processor, each of the generated at least one region of optimality for the single instance of the output (col. 7, lines 49-60), but fail to explicitly disclose wherein the generated at least one region of optimality that corresponds to the single instance of the output includes a precalculated and optimized counterfactual rule; and Amoukou further discloses generating, by the at least one processor, a report in response to each of the at least one request, the report including the precalculated and optimized counterfactual rule for the single instance of the output )(§ 4).
However, Kanamori discloses wherein the generated at least one region of optimality that corresponds to the single instance of the output includes a precalculated and optimized counterfactual rule (describes assigning optimized actions to decision-tree leaves. This supports storing subgroup-specific actions, but does not establish a precalculated rule in the particular regions of optimality)(§ 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 Amoukou further in view of Kanamori to incorporate the above cited features. One would have been motivated to do so to associate recourse actions with interpretable subgroups.
Claim 17 represents the device of claim 8 and is rejected along the same rationale.
Conclusion
9. 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