Detailed Action
This action is in response to the application filed 05/29/2023 for application 18/325,041, in which:
Claims 1 and 8 are the independent claims.
Claims 1-10 are currently pending.
Priority
Acknowledgment is made of applicant’s claim for foreign priority under 35 U.S.C. 119 (a)-(d). The certified copy has been filed in parent Application No. CN202310110092.8, filed on 02/02/2023.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 01/12/2026 is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Specification
Applicant is reminded of the proper content of an abstract of the disclosure.
A patent abstract is a concise statement of the technical disclosure of the patent and should include that which is new in the art to which the invention pertains. The abstract should not refer to purported merits or speculative applications of the invention and should not compare the invention with the prior art.
If the patent is of a basic nature, the entire technical disclosure may be new in the art, and the abstract should be directed to the entire disclosure. If the patent is in the nature of an improvement in an old apparatus, process, product, or composition, the abstract should include the technical disclosure of the improvement. The abstract should also mention by way of example any preferred modifications or alternatives.
Where applicable, the abstract should include the following: (1) if a machine or apparatus, its organization and operation; (2) if an article, its method of making; (3) if a chemical compound, its identity and use; (4) if a mixture, its ingredients; (5) if a process, the steps.
Extensive mechanical and design details of an apparatus should not be included in the abstract. The abstract should be in narrative form and generally limited to a single paragraph within the range of 50 to 150 words in length.
See MPEP § 608.01(b) for guidelines for the preparation of patent abstracts.
The abstract of the disclosure is objected to because the content of the abstract should be in narrative form. A corrected abstract of the disclosure is required and must be presented on a separate sheet, apart from any other text. See MPEP § 608.01(b).
Claim Objections
Claim 8 is objected to because of the following informalities: The limitation “… with a Shapley Additive exPlanations SHAP framework” should be “… with a Shapley Additive exPlanations (SHAP) framework”. Appropriate correction is required.
Claim Interpretation
The following is a quotation of 35 U.S.C. 112(f):
(f) Element in Claim for a Combination. – An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The following is a quotation of pre-AIA 35 U.S.C. 112, sixth paragraph:
An element in a claim for a combination may be expressed as a means or step for performing a specified function without the recital of structure, material, or acts in support thereof, and such claim shall be construed to cover the corresponding structure, material, or acts described in the specification and equivalents thereof.
The claims in this application are given their broadest reasonable interpretation using the plain meaning of the claim language in light of the specification as it would be understood by one of ordinary skill in the art. The broadest reasonable interpretation of a claim element (also commonly referred to as a claim limitation) is limited by the description in the specification when 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is invoked.
As explained in MPEP § 2181, subsection I, claim limitations that meet the following three-prong test will be interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph:
(A) the claim limitation uses the term “means” or “step” or a term used as a substitute for “means” that is a generic placeholder (also called a nonce term or a non-structural term having no specific structural meaning) for performing the claimed function;
(B) the term “means” or “step” or the generic placeholder is modified by functional language, typically, but not always linked by the transition word “for” (e.g., “means for”) or another linking word or phrase, such as “configured to” or “so that”; and
(C) the term “means” or “step” or the generic placeholder is not modified by sufficient structure, material, or acts for performing the claimed function.
Use of the word “means” (or “step”) in a claim with functional language creates a rebuttable presumption that the claim limitation is to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites sufficient structure, material, or acts to entirely perform the recited function.
Absence of the word “means” (or “step”) in a claim creates a rebuttable presumption that the claim limitation is not to be treated in accordance with 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph. The presumption that the claim limitation is not interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, is rebutted when the claim limitation recites function without reciting sufficient structure, material or acts to entirely perform the recited function.
Claim limitations in this application that use the word “means” (or “step”) are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action. Conversely, claim limitations in this application that do not use the word “means” (or “step”) are not being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, except as otherwise indicated in an Office action.
This application includes one or more claim limitations that do not use the word “means,” but are nonetheless being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, because the claim limitation(s) uses a generic placeholder that is coupled with functional language without reciting sufficient structure to perform the recited function and the generic placeholder is not preceded by a structural modifier. Such claim limitation(s) is/are: … a first determining module, configured for … , … a clustering module, configured for … , … a first calculating module, configured for … , … a second calculating module, configured for … , … a second determining module, configured for … , and … a training module, configured for … in Claim 8.
Because this/these claim limitation(s) is/are being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, it/they is/are being interpreted to cover the corresponding structure described in the specification as performing the claimed function, and equivalents thereof.
If applicant does not intend to have this/these limitation(s) interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph, applicant may: (1) amend the claim limitation(s) to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph (e.g., by reciting sufficient structure to perform the claimed function); or (2) present a sufficient showing that the claim limitation(s) recite(s) sufficient structure to perform the claimed function so as to avoid it/them being interpreted under 35 U.S.C. 112(f) or pre-AIA 35 U.S.C. 112, sixth paragraph.
Claim Rejections - 35 USC § 112
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 2 and 4 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 applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention.
Claim 2 recites the limitation "the K-Means algorithm" in … performing clustering to the participating nodes by the K-Means algorithm. There is insufficient antecedent basis for this limitation in the claim. For the purpose of applying prior art, “the” terminology has been interpreted as “a”; thus, the limitation is being interpreted as … performing clustering to the participating nodes by a K-Means algorithm.
Claim 4 recites the limitation "the cluster center" in … training the clusters according to the cluster center … There is insufficient antecedent basis for this limitation in the claim. For the purpose of applying prior art, “the” terminology has been interpreted as “a”; thus, the limitation is being interpreted as … performing clustering to the participating nodes by a K-Means algorithm.
The phrase “determining an” in Claim 4 is an unclear phrase which renders the claim indefinite as the limitation does not note what is being determined according to the clusters. For the purpose of applying prior art, “determining an” phrase has been interpreted as “determining an …”; thus, the limitation is being interpreted as “determining an … according to the clusters”.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-10 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an
abstract idea without significantly more.
Regarding Claim 1:
Subject Matter Eligibility Analysis Step 1:
Claim 1 recites a method, thus a process, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
However, Claim 1 further recites the method comprising:
determining participating nodes (a human being can mentally apply evaluation to determine specific nodes)
performing clustering to the participating nodes and determining several clusters (a human being can mentally apply evaluation to cluster specific nodes and determine multiple clusters)
determining a cluster model according to the several clusters and by means of calculation … (a mathematical relationship between variables and/or numbers using a mathematical formula/equations)
determining a cluster key feature set of any one of the clusters according to the cluster model and by means of calculation … (a mathematical relationship between variables and/or numbers using a mathematical formula/equations)
determining a global model according to the cluster key feature set (a human being can mentally apply evaluation to determine a specific type of model according to a specific feature set)
determining the cluster model of the any one of the clusters … (a human being can mentally apply evaluation to determine the cluster model for a specific cluster)
Claim 1 thus recites an abstract idea (that falls into the “mathematical concepts” and “mental processes” group of abstract ideas).
Subject Matter Eligibility Analysis Step 2A Prong 2:
This judicial exception is not integrated into a practical application because the additional elements recited consists of:
A multi-task learning method based on federated learning, comprising (which is restricting the abstract idea to a Particular Technological Environment, by MPEP 2106.05(h))
… with the federated learning (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
… with a SHapley Additive exPlanations (SHAP) framework (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
training the global model according to the any one of the clusters (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
… wherein a plurality of the clusters are used for achieving multi-task learning (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
Subject Matter Eligibility Analysis Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements recited, alone or in combination, do not provide significantly more than the abstract idea itself. Additional element a is only restricting the abstract idea to a Particular Technological Environment (MPEP 2106.05(h)) which cannot provide significantly more. Additional elements b-e are merely applying the abstract idea on a computer (MPEP 2106.05(f)) which cannot provide significantly more. Thus, the claim is subject-matter ineligible.
Regarding Claim 2:
Subject Matter Eligibility Analysis Step 1:
Dependent Claim 2 recites the method of Claim 1. Claim 1 is a method, thus a process, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
However, Claim 2 further recites the method comprising of:
wherein performing clustering to the participating nodes and determining several clusters comprises: determining a node training model … (a human being can mentally apply evaluation to determine a specific training model)
determining a training time and a model weight in response to determining that a preset number of times of training is reached (a human being can mentally apply evaluation to determine specific time and specific weights in response to a determination that a specific times of training has been reached)
Claim 2 thus recites an abstract idea (that falls into the “mental processes” group of abstract ideas).
Subject Matter Eligibility Analysis Step 2A Prong 2:
This judicial exception is not integrated into a practical application because the additional elements recited consists of:
… training the participating nodes by the node training model (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
… according to the training time and the model weight … (which is restricting the abstract idea to a Particular Technological Environment, by MPEP 2106.05(h))
Subject Matter Eligibility Analysis Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements recited, alone or in combination, do not provide significantly more than the abstract idea itself. Additional element a is merely applying the abstract idea on a computer (MPEP 2106.05(f)) which cannot provide significantly more. Additional element b is only restricting the abstract idea to a Particular Technological Environment (MPEP 2106.05(h)) which cannot provide significantly more. Thus, the claim is subject-matter ineligible.
Regarding Claim 3:
Subject Matter Eligibility Analysis Step 1:
Dependent Claim 3 recites the method of Claim 2. Claim 2 is a method, thus a process, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
However, Claim 3 further recites the method comprising of … by the K-Means algorithm (a mathematical relationship between variables and/or numbers using a mathematical formula/equations). Claim 3 thus recites an abstract idea (that falls into the “mathematical concepts” group of abstract ideas).
Subject Matter Eligibility Analysis Step 2A Prong 2:
This judicial exception is not integrated into a practical application because there are no additional elements.
Subject Matter Eligibility Analysis Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because there are no additional elements. The judicial exception alone does not provide significantly more than the abstract idea itself. Thus, the claim is subject-matter ineligible.
Regarding Claim 4:
Subject Matter Eligibility Analysis Step 1:
Dependent Claim 4 recites the method of Claim 1. Claim 1 is a method, thus a process, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
However, Claim 4 further recites the method comprising of:
wherein determining a cluster model according to the several clusters and by means of calculation with the federated learning comprises: determining … an according to the clusters (a human being can mentally apply evaluation to determine … based on the clusters)
determining the cluster model according to the clusters in response to determining that a number of times of training reaches a preset threshold (a human being can mentally apply evaluation to determine a specific model based on clusters in response to a determination that a specific amount of training reaches a predefined threshold)
Claim 4 thus recites an abstract idea (that falls into the “mental processes” group of abstract ideas).
Subject Matter Eligibility Analysis Step 2A Prong 2:
This judicial exception is not integrated into a practical application because the new sole additional element recited consists of training the clusters according to the cluster center and by the federated learning (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f)).
Subject Matter Eligibility Analysis Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the new sole additional element recited, alone or in combination, does not provide significantly more than the abstract idea itself. The additional element falls is merely applying the abstract idea on a computer (MPEP 2106.05(f)) which cannot provide significantly more. Thus, the claim is subject-matter ineligible.
Regarding Claim 5:
Subject Matter Eligibility Analysis Step 1:
Dependent Claim 5 recites the method of Claim 1. Claim 1 is a method, thus a process, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
However, Claim 5 further recites the method comprising of:
wherein determining a cluster key feature set of any one of the clusters according to the cluster model and by means of calculation with the SHAP framework comprises: analyzing the cluster model … (a human being can mentally apply evaluation to analyze the cluster model)
… determining a data feature and a data feature value of any one of the participating nodes in the clusters (a human being can mentally apply evaluation to determine data features/values of a specific node within the clusters)
determining a node key feature set according to the data feature value (a human being can mentally apply evaluation to determine a specific feature set based on a specific feature value)
… determining the cluster key feature set (a human being can mentally apply evaluation to determine a specific feature set)
Claim 5 thus recites an abstract idea (that falls into the “mental processes” group of abstract ideas).
Subject Matter Eligibility Analysis Step 2A Prong 2:
This judicial exception is not integrated into a practical application because the additional elements recited consists of:
… by the SHAP framework … (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
obtaining a union set for a plurality of the node key feature sets … (which is insignificant extra-solution activity of data gathering, by MPEP 2106.05(g)
Subject Matter Eligibility Analysis Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements recited, alone or in combination, do not provide significantly more than the abstract idea itself. Additional element a is merely applying the abstract idea on a computer (MPEP 2106.05(f)) which cannot provide significantly more. Additional element b falls within MPEP 2106.05(d) as well-understood, routine and conventional activities of receiving or transmitting data over a network (MPEP 2106.05(d)(II): buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014)). Thus, the claim is subject-matter ineligible.
Regarding Claim 6:
Subject Matter Eligibility Analysis Step 1:
Dependent Claim 6 recites the method of Claim 5. Claim 5 is a method, thus a process, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
However, Claim 6 further recites the method comprising of:
wherein determining a node key feature set according to the data feature value comprises: determining the data feature corresponding to the data feature value as a key feature in response to determining that the data feature value is higher than a preset threshold (a human being can mentally apply evaluation to determine the data feature corresponding to a specific value as a specific feature in response to a determination that a feature is higher than a predefined threshold)
determining the node key feature set according to the key feature (a human being can mentally apply evaluation to determine a specific feature set based on a specific feature value)
Claim 6 thus recites an abstract idea (that falls into the “mental processes” group of abstract ideas).
Subject Matter Eligibility Analysis Step 2A Prong 2:
This judicial exception is not integrated into a practical application because there are no new additional elements recited.
Subject Matter Eligibility Analysis Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because there are no new additional elements recited. The judicial exception alone does not provide significantly more than the abstract idea itself. Thus, the claim is subject-matter ineligible.
Regarding Claim 7:
Subject Matter Eligibility Analysis Step 1:
Dependent Claim 7 recites the method of Claim 1. Claim 1 is a method, thus a process, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
However, Claim 7 further recites the method comprising of:
wherein determining a global model according to the cluster key feature set comprises: … determining a global key feature set (a human being can mentally apply evaluation to determine a specific feature set)
performing feature masking for data of the participating nodes according to the global key feature set, and determining the global model (a human being can mentally apply evaluation to mask feature data for specific nodes according to a specific feature set and determine the global model)
Claim 7 thus recites an abstract idea (that falls into the “mental processes” group of abstract ideas).
Subject Matter Eligibility Analysis Step 2A Prong 2:
This judicial exception is not integrated into a practical application because the sole additional elements recited consists of obtaining an intersection set for a plurality of the cluster key feature sets … (which is insignificant extra-solution activity of data gathering, by MPEP 2106.05(g)).
Subject Matter Eligibility Analysis Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the new sole additional element recited, alone or in combination, does not provide significantly more than the abstract idea itself. The additional element falls within MPEP 2106.05(d) as well-understood, routine and conventional activities of receiving or transmitting data over a network (MPEP 2106.05(d)(II): buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014)). Thus, the claim is subject-matter ineligible.
Regarding Claim 8:
Subject Matter Eligibility Analysis Step 1:
Claim 8 recites a device, thus a machine, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
However, Claim 8 further recites the device comprising:
… determining participating nodes (a human being can mentally apply evaluation to determine specific nodes)
… performing clustering to the participating nodes and determining several clusters (a human being can mentally apply evaluation to cluster specific nodes and determine multiple clusters)
… determining a cluster model according to the several clusters and by means of calculation … (a mathematical relationship between variables and/or numbers using a mathematical formula/equations)
… determining a cluster key feature set of any one of the clusters according to the cluster model and by means of calculation … (a mathematical relationship between variables and/or numbers using a mathematical formula/equations)
… determining a global model according to the cluster key feature set (a human being can mentally apply evaluation to determine a specific type of model according to a specific feature set)
determining the cluster model of the any one of the clusters … (a human being can mentally apply evaluation to determine the cluster model for a specific cluster)
Claim 8 thus recites an abstract idea (that falls into the “mathematical concepts” and “mental processes” group of abstract ideas).
Subject Matter Eligibility Analysis Step 2A Prong 2:
This judicial exception is not integrated into a practical application because the additional elements recited consists of:
A device for multi-task learning based on federated learning, comprising: (which is restricting the abstract idea to a Particular Technological Environment, by MPEP 2106.05(h))
a first determining module, configured for … (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
a clustering module, configured for … (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
a first calculating module, configured for … with the federated learning (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
a second calculating module, configured for … with a SHapley Additive exPlanations SHAP framework (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
a second determining module, configured for … (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
a training module, configured for training the global model according to the any one of the clusters (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
… wherein a plurality of the clusters are used for achieving multi-task learning (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f))
Subject Matter Eligibility Analysis Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements recited, alone or in combination, do not provide significantly more than the abstract idea itself. Additional element a is only restricting the abstract idea to a Particular Technological Environment (MPEP 2106.05(h)) which cannot provide significantly more. Additional elements b-h are merely applying the abstract idea on a computer (MPEP 2106.05(f)) which cannot provide significantly more. Thus, the claim is subject-matter ineligible.
Regarding Claim 9:
Subject Matter Eligibility Analysis Step 1:
Dependent Claim 9 recites the method of Claim 1. Claim 1 is a method, thus a process, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
However, Claim 9 does not recite any additional abstract ideas and only inherits the abstract ideas from Claim 1. Claim 9 thus recites an abstract idea (that falls into the “mental processes” group of abstract ideas).
Subject Matter Eligibility Analysis Step 2A Prong 2:
This judicial exception is not integrated into a practical application because the new sole additional element recited consists of An electronic device, comprising a memory, a processor, and a computer program which is stored on the memory and can be executed by the processor, wherein the method according to claim 1 is implemented when the processor is executing the computer program (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f)).
Subject Matter Eligibility Analysis Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the new sole additional element recited, alone or in combination, does not provide significantly more than the abstract idea itself. The additional element is merely applying the abstract idea on a computer (MPEP 2106.05(f)) which cannot provide significantly more. Thus, the claim is subject-matter ineligible.
Regarding Claim 10:
Subject Matter Eligibility Analysis Step 1:
Dependent Claim 10 recites the method of Claim 1. Claim 1 is a method, thus a process, one of the four statutory categories of patentable subject matter.
Subject Matter Eligibility Analysis Step 2A Prong 1:
However, Claim 10 does not recite any additional abstract ideas and only inherits the abstract ideas from Claim 1. Claim 10 thus recites an abstract idea (that falls into the “mental processes” group of abstract ideas).
Subject Matter Eligibility Analysis Step 2A Prong 2:
This judicial exception is not integrated into a practical application because the new sole additional element recited consists of A non-transitory computer-readable storage medium, storing a computer instruction, wherein the computer instruction is used to make a computer execute the method according to claim 1 (to perform a mental process and the performance of an abstract idea on a computer is no more than instructions to “apply it” on a computer, by MPEP 2106.05(f)).
Subject Matter Eligibility Analysis Step 2B:
The claim does not include additional elements that are sufficient to amount to significantly more than the judicial exception because the new sole additional element recited, alone or in combination, does not provide significantly more than the abstract idea itself. The additional element is merely applying the abstract idea on a computer (MPEP 2106.05(f)) which cannot provide significantly more. Thus, the claim is subject-matter ineligible.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1 and 8-10 are rejected under 35 U.S.C. 103 as being unpatentable over Ghosh et al., “An Efficient Framework for Clustered Federated Learning”, in view of Zeng et al., “CS Sparse K-means: An Algorithm for Cluster-Specific Feature Selection in High-Dimensional Clustering”, in view of Pathirannehelage et al., “Analysis of Centralized to Federated Learning-based Anomaly Detection in Networks with Explainable AI (XAI)”.
Regarding Claim 1:
Ghosh teaches:
A multi-task learning method based on federated learning, comprising:
(Ghosh, Abstract, “We address the problem of Federated Learning (FL) where users are distributed and partitioned into clusters … We propose a new framework dubbed the Iterative Federated Clustering Algorithm (IFCA) … we propose to train the models by combining IFCA with the weight sharing technique in multi-task learning …”; Page 3, Paragraph 3, “We consider a distributed learning setting where we have one center machine and m worker machines (i.e., each worker machine corresponds to a user in the Federated Learning framework)”; Page 4, Figure 1. IFCA is a method that implements multi-task learning which is federated learning based via different clusters/nodes (where each cluster/node is computing an inference task via worker machines which corresponds to a user in the Federated Learning Framework) which is shown in Figure 1 (which is interpreted by the examiner as each cluster handling a different task for combination; thus, multi-task learning)).
determining participating nodes;
(Ghosh, Page 4, Algorithm 1: Line 4, “Mt random subset of worker machines (participating devices)”; Page 9, Paragraph 1, “In the local model scheme, the model in each node performs gradient descent only on local data available …”; Page 1, Paragraph 2, “In Federated Learning, since the data source and computing nodes are end users’ personal devices …”. The federated learning system implements determining the participating nodes by showing the set of mobile devices (which are the nodes/clients) that will be clustered as shown in Algorithm 1).
performing clustering to the participating nodes and determining several clusters;
(Ghosh, Page 4, Algorithm 1, Paragraph 2, “Here, we call Mt the set of participating devices. Recall that each worker machine is equipped with local empirical loss function Fi(·). Using the received parameter estimates and Fi, the i-th worker machine … estimates its cluster identity via finding the model parameter with lowest loss …”. The system takes a random subset as shown in Algorithm 1 to determine the cluster identities of each participating device (node); thus, the IFCA methodology performs clustering of the participating nodes (devices) and determines cluster identities for each participating node; thus, interpreted by the examiner as determining several clusters).
determining a cluster model according to the several clusters and by means of calculation with the federated learning;
(Ghosh, Page 4, Algorithm 1; Page 2, Paragraph 2, “… we have to simultaneously solve two problems: identifying the cluster membership of each user and optimizing each of the cluster models in a distributed setting”; Page 9, Paragraph 1, “… In the global model scheme, the algorithm tries to learn single global model that can make predictions from all the distributions …”. Algorithm 1 shows the iterative federated clustering algorithm (IFCA) which optimizes each of the cluster models to be able to distribute updates for the single global model according to the several cluster models via federated learning. Thus, determining a global cluster model according to the several local cluster models).
…
determining a global model …; and training the global model according to the any one of the clusters, and determining the cluster model of the any one of the clusters, wherein a plurality of the clusters are used for achieving multi-task learning.
(Ghosh, Page 4, Algorithm 1; Page 2, Paragraph 2, “… we have to simultaneously solve two problems: identifying the cluster membership of each user and optimizing each of the cluster models in a distributed setting”; Page 9, Paragraph 1, “… In the global model scheme, the algorithm tries to learn single global model that can make predictions from all the distributions …”; Page 5, Paragraph 2, “… we propose to use the weight sharing technique in multi-task learning [3] and combine it with IFCA”. The global model is determined via multi-task learning combined with the IFCA methodology; where each local cluster model is considered a different task which updates the final global model (as shown in Algorithm 1).Thus, the global model is trained and learns from the local model clusters which is according to all the local clusters that are participating devices/nodes).
Ghosh teaches the Iterative Federated Clustering Algorithm (IFCA) methodology which utilizes Federated Learning and determining cluster models for multi-task learning but does not explicitly teach:
determining a cluster key feature set of any one of the clusters according to the cluster model and by means of calculation with a SHapley Additive exPlanations (SHAP) framework;
… according to the cluster key feature set
However, Zeng teaches:
determining a cluster key feature set of any one of the clusters according to the cluster model …
(Zeng, Page 7, Figure 4; Page 1, Paragraph 4, “… the weight is calculated with clustering results, and the clustering is based on feature weights …”; Page 6, Algorithm 1. Figure 4/Algorithm 1 show the determining of cluster-specific (CS) features for finding feature importance; thus, Algorithm 1 determines a cluster key feature set for all clusters being evaluated for importance to as each feature pair is assigned weights based on importance/impact).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to utilize Ghosh’s iterative federated learning clustering methodology with the explicit teaching of Zeng’s determination of cluster key feature sets within the cluster specific methodology. One having ordinary skill in the art would have been motivated to implement this change before the effective filing date of the claimed invention, as this leads to improving clustering accuracy, feature analysis, analysis via discovery, and assigning importance weights (see Zeng, Page 8, Paragraph 3, “… In this paper, we developed an algorithm for cluster-specific sparse clustering based on K-means. Our algorithm showed better clustering accuracy than sparse K-means on several simulated or real-world datasets”; Pages 6-7, Figures 3-4).
However, Ghosh/Zeng do not explicitly teach:
determining a cluster key feature set of any one of the clusters according to the cluster model and by means of calculation with a SHapley Additive exPlanations (SHAP) framework;
determining a global model according to the cluster key feature set;
Nevertheless, Pathirannehelage teaches:
determining a cluster key feature set of any one of the clusters according to the cluster model and by means of calculation with a SHapley Additive exPlanations (SHAP) framework;
(Pathirannehelage, Page 23, Figure 2.6: IIoT Sensors, Paragraph 2, “… proposed an anomaly detection system ideal for security services in smaller clusters of IoT”; Page 27, Paragraph 3, “We utilized SHAP as the XAI tool for the provided CL and FL models”; Page 28, Figure 3.1b: XAI; Page 25, Paragraph 4, “… utilized SHAP-based framework to enhance the interpretability of NIDSs, offering key features on local and global explanations, and demonstrates the variation between feature values and SHAP values for a variety of attack types.” Figure 3.1b shows the FL based anomaly detection model which includes XAI; where the XAI tool for the FL based model utilizes the SHAP framework for calculations; Pathirannehelage also mentions smaller clusters of devices/IOTs where Figure 3.1b shows IIOTs; thus, interpreted by the examiner as cluster models utilizing the SHAP framework).
determining a global model according to the cluster key feature set;
(Pathirannehelage, Page 25, Paragraph 4, “… utilized SHAP-based framework to enhance the interpretability of NIDSs, offering key features on local and global explanations, and demonstrates the variation between feature values and SHAP values for a variety of attack types”. Pathirannehelage teaches using SHAP-based frameworks to determine global explanations; thus, interpreted by the examiner as determining a global model according to the cluster key features).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to utilize Ghosh/Zeng’s method of clustering with the explicit teaching of Pathirannehelage’s means of calculations (SHAP/K-means) within their DNN methodology. One having ordinary skill in the art would have been motivated to implement this change before the effective filing date of the claimed invention, as this leads to improving explainability, timing constraints, transparency, modularity, measure significance, global scope, and more (see Pathirannehelage, Page 59, Paragraph 2, “… enable exponential amount of information transactions. This will increase the data handling and protection … Adding explainability will improve the transparency … It has been shown that Shapley values can be used to measure the significance of FL client local data feature without exposing the local data … This work will provide sufficient insight on migrating XAI based ML anomaly detection to XAI based FL anomaly detection”; Page 61, Paragraph 4, “… Same DNN model and datasets were used for the ease of comparison … SHAP was chosen as the XAI method to achieve global scope of the input data … After training … SHAP algorithm was applied to derive the global explanations of the models … SHAP explanations for each of the attack categories in the training data was presented …”; Page 45, Paragraph 1, “… Kernel SHAP explainer was selected. We expected to provide SHAP explanation for the total training set. But computing for such large dataset was not feasible due to high computation time. Using reduced set of training set is also not advisable … Instead, we used a summarized background dataset which is weighted by the k-means algorithm. SHAP k-means algorithm weights the data using number of occurrences”).
Regarding Claim 8:
Ghosh teaches:
A device for multi-task learning based on federated learning, comprising:
(Ghosh, Abstract, “We address the problem of Federated Learning (FL) where users are distributed and partitioned into clusters … We propose a new framework dubbed the Iterative Federated Clustering Algorithm (IFCA) … we propose to train the models by combining IFCA with the weight sharing technique in multi-task learning …”; Page 3, Paragraph 3, “We consider a distributed learning setting where we have one center machine and m worker machines (i.e., each worker machine corresponds to a user in the Federated Learning framework)”; Page 4, Figure 1; Page 1, Footnote 2: “Implementation of our experiments is open sourced at https://github.com/jichan3751/ifca”. IFCA is a method that implements multi-task learning which is federated learning based via different clusters/nodes (where each cluster/node is computing an inference task via worker machines which corresponds to a user in the Federated Learning Framework) which is shown in Figure 1 (which is interpreted by the examiner as each cluster handling a different task for combination; thus, multi-task learning). The IFCA method utilizes open source code within github where a processor, memory, and a non-transitory machine-readable information storage medium are inherent within a device that uses the framework such as IFCA).
a first determining module, configured for determining participating nodes;
(Ghosh, Page 4, Algorithm 1: Line 4, “Mt random subset of worker machines (participating devices)”; Page 9, Paragraph 1, “In the local model scheme, the model in each node performs gradient descent only on local data available …”; Page 1, Paragraph 2, “In Federated Learning, since the data source and computing nodes are end users’ personal devices …”. The federated learning system implements determining the participating nodes by showing the set of mobile devices (which are the nodes/clients) that will be clustered as shown in Algorithm 1; ); thus, the federated learning system containing IFCA is a first determining module).
a clustering module, configured for performing clustering to the participating nodes and determining several clusters;
(Ghosh, Page 4, Algorithm 1, Paragraph 2, “Here, we call Mt the set of participating devices. Recall that each worker machine is equipped with local empirical loss function Fi(·). Using the received parameter estimates and Fi, the i-th worker machine … estimates its cluster identity via finding the model parameter with lowest loss …”. The system takes a random subset as shown in Algorithm 1 to determine the cluster identities of each participating device (node); thus, the IFCA methodology within the system (interpreted as a clustering module) performs clustering of the participating nodes (devices) and determines cluster identities for each participating node; thus, interpreted by the examiner as determining several clusters).
a first calculating module, configured for determining a cluster model according to the several clusters and by means of calculation with the federated learning;
(Ghosh, Page 4, Figure 1, Algorithm 1; Page 2, Paragraph 2, “… we have to simultaneously solve two problems: identifying the cluster membership of each user and optimizing each of the cluster models in a distributed setting”; Page 9, Paragraph 1, “… In the global model scheme, the algorithm tries to learn single global model that can make predictions from all the distributions …”. Algorithm 1 shows the iterative federated clustering algorithm (IFCA) which optimizes each of the cluster models to be able to distribute updates for the single global model according to the several cluster models via federated learning. Thus, determining a global cluster model according to the several local cluster models; where the device containing IFCA (Figure 1d = function for averaging) is interpreted as performing a first calculating module).
…
a second determining module, configured for determining a global model …; and a training module, configured for training the global model according to the any one of the clusters, and determining the cluster model of the any one of the clusters, wherein a plurality of the clusters are used for achieving multi-task learning.
(Ghosh, Page 4, Figure 1, Algorithm 1; Page 2, Paragraph 2, “… we have to simultaneously solve two problems: identifying the cluster membership of each user and optimizing each of the cluster models in a distributed setting”; Page 9, Paragraph 1, “… In the global model scheme, the algorithm tries to learn single global model that can make predictions from all the distributions …”; Page 5, Paragraph 2, “… we propose to use the weight sharing technique in multi-task learning [3] and combine it with IFCA”. The global model is determined via multi-task learning combined with the IFCA methodology; where each local cluster model is considered a different task which updates the final global model (as shown in Algorithm 1). The global model is trained and learns from the local model clusters which is according to all the local clusters that are participating devices/nodes; thus, the server containing the global model is interpreted as a second determining function/module using IFCA to train (training module)).
Ghosh teaches the Iterative Federated Clustering Algorithm (IFCA) methodology which utilizes Federated Learning and determining cluster models for multi-task learning but does not explicitly teach:
a second calculating module, configured for determining a cluster key feature set of any one of the clusters according to the cluster model and by means of calculation with a SHapley Additive exPlanations (SHAP) framework;
… according to the cluster key feature set
However, Zeng teaches:
a second calculating module, configured for determining a cluster key feature set of any one of the clusters according to the cluster model …
(Zeng, Page 7, Figure 4; Page 1, Paragraph 4, “… the weight is calculated with clustering results, and the clustering is based on feature weights …”; Page 6, Algorithm 1. Figure 4/Algorithm 1 show the determining of cluster-specific (CS) features for finding feature importance; thus, Algorithm 1 determines a cluster key feature set for all clusters being evaluated for importance to as each feature pair is assigned weights based on importance/impact; thus, the algorithm is interpreted as a second calculating module/function/process for determining a cluster key feature for each node within a list/set according the cluster model).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to utilize Ghosh’s iterative federated learning clustering methodology with the explicit teaching of Zeng’s determination of cluster key feature sets within the cluster specific methodology. One having ordinary skill in the art would have been motivated to implement this change before the effective filing date of the claimed invention, as this leads to improving clustering accuracy, feature analysis, analysis via discovery, and assigning importance weights (see Zeng, Page 8, Paragraph 3, “… In this paper, we developed an algorithm for cluster-specific sparse clustering based on K-means. Our algorithm showed better clustering accuracy than sparse K-means on several simulated or real-world datasets”; Pages 6-7, Figures 3-4).
However, Ghosh/Zeng do not explicitly teach:
determining a cluster key feature set of any one of the clusters according to the cluster model and by means of calculation with a SHapley Additive exPlanations (SHAP) framework;
determining a global model according to the cluster key feature set;
Nevertheless, Pathirannehelage teaches:
determining a cluster key feature set of any one of the clusters according to the cluster model and by means of calculation with a SHapley Additive exPlanations (SHAP) framework;
(Pathirannehelage, Page 23, Figure 2.6: IIoT Sensors, Paragraph 2, “… proposed an anomaly detection system ideal for security services in smaller clusters of IoT”; Page 27, Paragraph 3, “We utilized SHAP as the XAI tool for the provided CL and FL models”; Page 28, Figure 3.1b: XAI; Page 25, Paragraph 4, “… utilized SHAP-based framework to enhance the interpretability of NIDSs, offering key features on local and global explanations, and demonstrates the variation between feature values and SHAP values for a variety of attack types.” Figure 3.1b shows the FL based anomaly detection model which includes XAI; where the XAI tool for the FL based model utilizes the SHAP framework for calculations; Pathirannehelage also mentions smaller clusters of devices/IOTs where Figure 3.1b shows IIOTs; thus, interpreted by the examiner as cluster models utilizing the SHAP framework).
determining a global model according to the cluster key feature set;
(Pathirannehelage, Page 25, Paragraph 4, “… utilized SHAP-based framework to enhance the interpretability of NIDSs, offering key features on local and global explanations, and demonstrates the variation between feature values and SHAP values for a variety of attack types”. Pathirannehelage teaches using SHAP-based frameworks to determine global explanations; thus, interpreted by the examiner as determining a global model according to the cluster key features).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to utilize Ghosh/Zeng’s method of clustering with the explicit teaching of Pathirannehelage’s means of calculations (SHAP/K-means) within their DNN methodology. One having ordinary skill in the art would have been motivated to implement this change before the effective filing date of the claimed invention, as this leads to improving explainability, timing constraints, transparency, modularity, measure significance, global scope, and more (see Pathirannehelage, Page 59, Paragraph 2, “… enable exponential amount of information transactions. This will increase the data handling and protection … Adding explainability will improve the transparency … It has been shown that Shapley values can be used to measure the significance of FL client local data feature without exposing the local data … This work will provide sufficient insight on migrating XAI based ML anomaly detection to XAI based FL anomaly detection”; Page 61, Paragraph 4, “… Same DNN model and datasets were used for the ease of comparison … SHAP was chosen as the XAI method to achieve global scope of the input data … After training … SHAP algorithm was applied to derive the global explanations of the models … SHAP explanations for each of the attack categories in the training data was presented …”; Page 45, Paragraph 1, “… Kernel SHAP explainer was selected. We expected to provide SHAP explanation for the total training set. But computing for such large dataset was not feasible due to high computation time. Using reduced set of training set is also not advisable … Instead, we used a summarized background dataset which is weighted by the k-means algorithm. SHAP k-means algorithm weights the data using number of occurrences”).
Regarding Claim 9:
Ghosh/Zeng/Pathirannehelage teach the method of Claim 1 and Ghosh further teaches:
An electronic device, comprising a memory, a processor, and a computer program which is stored on the memory and can be executed by the processor, wherein the method according to claim 1 is implemented when the processor is executing the computer program.
(Ghosh, Page 10, Paragraph 1, “The reason is that our algorithm does not require the users to send any of their own personal data to the central server, and the users can still learn a personalized model using their on-device computing power”; Page 4, Figure 1; Page 1, Footnote 2: “Implementation of our experiments is open sourced at https://github.com/jichan3751/ifca”. IFCA is a method that implements multi-task learning which is federated learning based via different clusters/nodes (where each cluster/node is computing an inference task via worker machines which corresponds to a user in the Federated Learning Framework) which is shown in Figure 1 (which is interpreted by the examiner as each cluster handling a different task for combination; thus, multi-task learning). The IFCA method utilizes open source code within github where a processor, memory, and a non-transitory machine-readable information storage medium are inherent within a device that uses the framework such as IFCA to execute a computer program); thus, Claim 9 is rejected for reasons set forth in the rejection of Claim 1.
Regarding Claim 10:
Ghosh/Zeng/Pathirannehelage teach the method of Claim 1 and Ghosh further teaches:
A non-transitory computer-readable storage medium, storing a computer instruction, wherein the computer instruction is used to make a computer execute the method according to claim 1.
(Ghosh, Page 10, Paragraph 1, “The reason is that our algorithm does not require the users to send any of their own personal data to the central server, and the users can still learn a personalized model using their on-device computing power”; Page 4, Figure 1; Page 1, Footnote 2: “Implementation of our experiments is open sourced at https://github.com/jichan3751/ifca”. IFCA is a method that implements multi-task learning which is federated learning based via different clusters/nodes (where each cluster/node is computing an inference task via worker machines which corresponds to a user in the Federated Learning Framework) which is shown in Figure 1 (which is interpreted by the examiner as each cluster handling a different task for combination; thus, multi-task learning). The IFCA method utilizes open source code within github where a processor, memory, and a non-transitory machine-readable information storage medium are inherent within a device that uses the framework such as IFCA to execute computer instructions); thus, Claim 10 is rejected for reasons set forth in the rejection of Claim 1.
Claims 2-4 are rejected under 35 U.S.C. 103 as being unpatentable over Ghosh et al., “An Efficient Framework for Clustered Federated Learning”, in view of Zeng et al., “CS Sparse K-means: An Algorithm for Cluster-Specific Feature Selection in High-Dimensional Clustering”, in view of Pathirannehelage et al., “Analysis of Centralized to Federated Learning-based Anomaly Detection in Networks with Explainable AI (XAI)”., in view of Chai et al., “TiFL: A Tier-based Federated Learning System”.
Regarding Claim 2:
Ghosh/Zeng/Pathirannehelage teach the method of Claim 1 and Ghosh further teaches:
wherein performing clustering to the participating nodes and determining several clusters comprises:
determining a node training model, and training the participating nodes by the node training model;
(Ghosh, Page 3, Figure 1; Page 9, Paragraph 3, “As we run the IFCA algorithm, we observe that we can gradually find the underlying cluster identities of the worker machines, and after the correct cluster is found, each model is trained and tested using data with the same distribution, resulting in better accuracy.”; Page 4, Algorithm 1: Line 4, “Mt random subset of worker machines (participating devices)”; Page 9, Paragraph 1, “In the local model scheme, the model in each node performs gradient descent only on local data available …”; Page 1, Paragraph 2, “In Federated Learning, since the data source and computing nodes are end users’ personal devices …”. IFCA trains nodes via determining a node training model and trains the participating nodes/devices for local model training (as shown in Figure 1); thus, interpreted by the examiner as the node training model).
However, Ghosh/Zeng/Pathirannehelage do not explicitly teach:
determining a training time and a model weight in response to determining that a preset number of times of training is reached; and performing clustering to the participating nodes according to the training time and the model weight, and determining several clusters.
Nevertheless, Chai teaches:
determining a training time and a model weight in response to determining that a preset number of times of training is reached; and performing clustering to the participating nodes according to the training time and the model weight, and determining several clusters.
(Chai, Page 128, Figures 1-2; Page 133, Figure 3; Page 127, Column 1, Paragraph 2, “At the beginning of each round, the aggregator sends the current model weights to a subset of randomly selected clients. Each selected client then trains its local model with its local data and sends back the updated weights to the aggregator after local training … This iterative process keeps on updating the global model until a certain number of rounds are completed or a desired accuracy is reached”. Chai teaches a method named TiFL which utilizes tiers (shown in Figure 2) to perform clustering based on a training time (Very Fast/Fast/…/Very Slow) and a model weight (W); thus, performing clustering to the participating nodes according to the training time and the model weight, and determining several clusters. Figure 3 shows the maximum amount of rounds (which is set to 500); thus, determining a training time, accuracy and a model weight (as the aggregator is receiving updated weights from the local models after local training rounds) in response to reaching 500 rounds (interpreted by the examiner as number of times of training is reached)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to utilize Ghosh/Zeng/Pathirannehelage’s method of clustering with the explicit teaching of Chai’s determination of training times/model weights/preset threshold to perform clustering. One having ordinary skill in the art would have been motivated to implement this change before the effective filing date of the claimed invention, as this leads to improving accuracy, training time, and tackling resource/data heterogeneity (see Chai, Page 136, Column 1, Paragraph 3, “…Based on the observations of our case study, we propose and prototype a Tier-based Federated Learning System called TiFL. Tackling the resource and data heterogeneity, TiFL employs a tier-based approach that groups clients in tiers by their training response latencies … we further design an adaptive tier selection approach that enables TiFL be data heterogeneity aware and outperform conventional FL in various heterogeneous scenarios: resource heterogeneity, data quantity heterogeneity, non-IID data heterogeneity, and their combinations. Specifically, TiFL achieves an improvement over conventional FL by up to 3× speedup in overall training time and by 6% in accuracy”).
Regarding Claim 3:
Ghosh/Zeng/Pathirannehelage/Chai teach the method of Claim 2 and Zeng/Pathirannehelage further teaches:
performing clustering to the participating nodes by the K-Means algorithm.
(see Pathirannehelage, Page 45, Paragraph 1, “Instead, we used a summarized background dataset which is weighted by the k-means algorithm. SHAP k-means algorithm weights the data using number of occurrences”; see Zeng, Page 2, Figure 1. SHAP K-Means algorithm is taught to be used on the data within Pathirannehelage and also can be reviewed within Zeng’s Figure 1).
The motivation of Claim 2’s combination of Ghosh/Zeng/Pathirannehelage/Chai is still maintained.
Regarding Claim 4:
Ghosh/Zeng/Pathirannehelage teach the method of Claim 1 and Ghosh further teaches:
wherein determining a cluster model according to the several clusters and by means of calculation with the federated learning comprises:
determining an according to the clusters;
(Ghosh, Page 4, Figure 1, Algorithm 1: Line 6. Figure 1 shows an overview of the model averaging of the clusters (nodes/devices shown in Figure 1b) where the equation of Figure 1b is for the cluster identity estimate; thus, determining an according to the clusters which is based on the node/device belonging to a cluster).
training the clusters … and by the federated learning; and
(Ghosh, Page 2, Paragraph 2, “In this paper, we study … the clustered Federated Learning … we have to simultaneously solve two problems: identifying the cluster membership of each user and optimizing each of the cluster models in a distributed setting … we propose a framework … (IFCA) for clustered FL … we learn … use IFCA to train separate final layers for each individual cluster”. IFCA trains the clusters via Federated Learning).
Pathirannehelage further teaches
… according to the cluster center …
(Pathirannehelage, Page 45, Paragraph 1, “Instead, we used a summarized background dataset which is weighted by the k-means algorithm. SHAP k-means algorithm weights the data using number of occurrences”. SHAP K-Means algorithm is taught to be used on the data within Pathirannehelage which would be understood by one of ordinary skill in the art as an algorithm to minimize the distance between data points and their cluster centroid (mean); thus, interpreted by the examiner as according to the cluster center within the system of Ghosh/Pathirannehelage which is a clustering model utilizing the SHAP framework and k-means algorithm to train according to the cluster center).
However, Ghosh/Zeng/Pathirannehelage does not explicitly disclose:
determining the cluster model according to the clusters in response to determining that a number of times of training reaches a preset threshold.
Nevertheless, Chai teaches:
determining the cluster model according to the clusters in response to determining that a number of times of training reaches a preset threshold.
(Chai, Page 128, Figures 1-2; Page 133, Figure 3; Page 127, Column 1, Paragraph 2, “… Each selected client then trains its local model with its local data and sends back the updated weights to the aggregator after local training … This iterative process keeps on updating the global model until a certain number of rounds are completed or a desired accuracy is reached”. Figure 3 shows the maximum amount of rounds (which is set to 500); thus, determining the global cluster model according to the local clusters in response to 500 rounds of training (where 500 rounds is interpreted as a preset threshold by the examiner)).
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to utilize Ghosh/Zeng/Pathirannehelage’s method of clustering with the explicit teaching of Chai’s determination of training times/model weights/preset threshold to perform clustering. One having ordinary skill in the art would have been motivated to implement this change before the effective filing date of the claimed invention, as this leads to improving accuracy, training time, and tackling resource/data heterogeneity (see Chai, Page 136, Column 1, Paragraph 3, “…Based on the observations of our case study, we propose and prototype a Tier-based Federated Learning System called TiFL. Tackling the resource and data heterogeneity, TiFL employs a tier-based approach that groups clients in tiers by their training response latencies … we further design an adaptive tier selection approach that enables TiFL be data heterogeneity aware and outperform conventional FL in various heterogeneous scenarios: resource heterogeneity, data quantity heterogeneity, non-IID data heterogeneity, and their combinations. Specifically, TiFL achieves an improvement over conventional FL by up to 3× speedup in overall training time and by 6% in accuracy”).
Examiner Comments
Claims 5-7 are currently rejected under 35 USC § 101 only. A complete and thorough search was performed for these claims; however no prior art was uncovered that teach or fairly suggest the features recited claims. Specifically, none of the prior art of record, either alone or in combination, fairly discloses the limitations of the independent Claims 5-7. In particular, the limitations in:
Claim 5:
… obtaining a union set for a plurality of the node key feature sets, and determining the cluster key feature set.
Claim 6:
determining a node key feature set according to the data feature value comprises: determining the data feature corresponding to the data feature value as a key feature in response to determining that the data feature value is higher than a preset threshold; and determining the node key feature set according to the key feature.
Claim 7:
… obtaining an intersection set for a plurality of the cluster key feature sets, and determining a global key feature set; …
The closest prior art of record is Ghosh et al., “An Efficient Framework for Clustered Federated Learning”, in view of Zeng et al., “CS Sparse K-means: An Algorithm for Cluster-Specific Feature Selection in High-Dimensional Clustering”, in view of Pathirannehelage et al., “Analysis of Centralized to Federated Learning-based Anomaly Detection in Networks with Explainable AI (XAI)”., in view of Chai et al., “TiFL: A Tier-based Federated Learning System” where the methodology of Ghosh/Zeng/Pathirannehelage/Chai teaches a multi-task learning method within federated learning to determine cluster models and analyze local/global models within a SHAPley framework; however, Ghosh/Zeng/Pathirannehelage/Chai do not explicitly disclose obtaining union/intersection sets of specific key features to specific key features/sets. Thus, the combination of these three prior arts do not disclose the expressions defined in Claim 5-7 and it’s dependent claims.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to IBRAHIM RAHMAN whose telephone number is (703)756-1646. The examiner can normally be reached M-F 8am-5pm.
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, Kakali Chaki can be reached at (571) 272-3719. 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.
/I.R./Examiner, Art Unit 2122
/KAKALI CHAKI/Supervisory Patent Examiner, Art Unit 2122
Prosecution Insights