DETAILED ACTION
This office action is in response to application filed on 9/24/2024.
Claims 1 – 14 and 25 – 27 are pending.
Priority is claimed as national stage application of PCT/IB2022/053814 (filed on 4/25/2022).
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claim 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 10 is 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 10 recites the limitation "the predetermined accuracy threshold". There is insufficient antecedent basis for this limitation in the claim. Applicant is advised to amend claim 10 to be dependent on claim 9 to overcome this deficiency.
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.
Claim(s) 1 – 4, 7, 8, 12 – 14 and 25 – 27 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li et al (US 20200076884, hereinafter Li), in view of Bathen et al (US 20180343175, hereinafter Bathen).
As per claim 1, Li discloses: A method for computing resource allocation in a collaborative machine learning, ML, system, the method comprising:
receiving, at a resource management controller, registration requests from one or more computing devices seeking to participate in the collaborative ML system, wherein the registration requests comprise smart contracts based on a blockchain system; (Li [0123]: “In step 708 the new arrival node performs node registration, e.g., the new arrival node registers itself to the blockchain… said account including account information, said account information being an input to a resource allocation smart contract… step 712, in which the new arrival node broadcasts the account information on blockchain. Since each node in the provider network owns the same copy of the smart contact, along with the requesting node, e.g. node A, registration, each of the nodes in the network are updating the same smart contract. For example node A 504, node C 508, node D 510 are able to update their copy of the smart contract based on the information broadcast in step 712 by new arrival node B 506.”)
registering the one or more computing devices; (Li [0124]: “In step 714 the new arrival node receives a service request for a computing task, e.g., a machine learning (ML) training task. For example, the received service request is a service request from node A 504 which is a service request for a computing task through the blockchain, e.g. a machine learning task, This request is an input signal to the resource allocation smart contract. The request information communicated in the received request reflects the needed (estimated) computational load, For example, in machine learning training task in some embodiments, the request includes the sizes of the neural network, dataset, and the format of the data, e.g., images, voices, etc.”; [0125]: “In step 716 the new arrival node writes the received service request into the resource allocation smart contract.”)
allocating computing resources provided by the one or more computing devices to the collaborative ML system, and tracking the allocation of resources using the smart contracts; (Li [0126]: “In step 718 the new arrival node runs a resource allocation method to select a number of operation nodes for the computing task. Having the information for the service request and the registration information from each of the registered nodes in the network each node including the new arrival node, e.g., node B, runs the same copy of the resource allocation contract to select a number of nodes that will together complete the computing task. This node selection process is essentially based on system capability of each registered node. The output of the smart contract is a set of node IDs, such as blockchain account (i.e., a user specific public key). Step 718 includes step 720 in which the new arrival node selects a number of nodes that will together complete the computing task. Step 720 include step 722 in which the new arrival node determines a set of node IDs, e.g., blockchain account IDs, corresponding to the selected nodes that will perform the computing task. Operation proceeds from step 718, via connecting node A 724 to step 726.”)
Li did not explicitly disclose:
receiving updated information on at least one of: the available computing resources provided by the one or more computing devices, and the collaborative ML system resource requirements;
and updating the allocation of computing resources to the collaborative ML system based on the updated information, and tracking the updated allocation of resources using the smart contracts.
However, Bathen teaches:
receiving updated information on at least one of: the available computing resources provided by the one or more computing devices, and the collaborative ML system resource requirements; and updating the allocation of computing resources to the collaborative ML system based on the updated information, and tracking the updated allocation of resources using the smart contracts. (Bathen figure 4B and [0029] – [0030])
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of Bathen into that of Li in order to receive updated information on at least one of: the available computing resources provided by the one or more computing devices, and the collaborative ML system resource requirements; and updating the allocation of computing resources to the collaborative ML system based on the updated information, and tracking the updated allocation of resources using the smart contracts. Li figure 7 and 8 teaches the smart contracts are automatically updated whenever a new arrival node wish to be registered to the blockchain. However, one of ordinary skill in the art can easily see that other criteria that can lead to update of the smart contract in blockchain can equally be used here without deviating from the teaching of Li, such as change of resource usage based on an optimization goal as taught by Bathen, applicants here have merely claimed the combination of known parts in the field to achieve predictable results and is therefore rejected under 35 USC 103.
As per claim 2, the combination of Li and Bathen further teach:
The method of claim 1, wherein the one or more registration requests comprise information on the computing devices. (Li [0123]: “the account information includes, e.g., user information, e.g., necessary user information, for conducting distributed computing tasks, such as, e.g.: i) computing capability, e.g., GPU, CPU, etc., ii) connectivity capability, e.g., turn-on duration, bandwidth, etc., and iii) storage capability, e.g., memory size, buffer size, etc.”)
As per claim 3, the combination of Li and Bathen further teach:
The method of claim 2, wherein the information on the computing devices comprises at least one of: a device type of the computing device; a computation capacity of the computing device; a storage capacity of the computing device; a network interface capacity of the computing device; a physical location of the computing device; a type of energy used to power the computing device; and an availability of the computing device. (Li [0123]: “the account information includes, e.g., user information, e.g., necessary user information, for conducting distributed computing tasks, such as, e.g.: i) computing capability, e.g., GPU, CPU, etc., ii) connectivity capability, e.g., turn-on duration, bandwidth, etc., and iii) storage capability, e.g., memory size, buffer size, etc.”)
As per claim 4, the combination of Li and Bathen further teach:
The method of claim 2 further comprising, following the receipt of the registration requests, configuring the collaborative ML system. (Li [0150]: “In step 810 the existing node, e.g., Node C 508, updates the states of the smart contract, wherein the smart contract is a resource allocation contract or multiple related contracts, and wherein the states of the smart contract are attributes for the methods in the smart contract, e.g., the number of active nodes, registration information of the new registered node, etc.”)
As per claim 7, the combination of Li and Bathen further teach:
The method of claim 1, wherein the resource management controller monitors the one or more computing devices to receive the updated information. (Bathen figure 4B and [0029] – [0030])
As per claim 8, the combination of Li and Bathen further teach:
The method of claim 7, wherein the monitoring comprises receiving one or more Key Performance Indicators, KPI, from the one or more computing devices. (Bathen figure 4B and [0029] – [0030])
As per claim 12, the combination of Li and Bathen further teach:
The method of claim 1, wherein the one or more computing devices comprise at least one of: an Internet of Things, IoT, device; an edge server; and a database. (Bathen [0030])
As per claim 13, the combination of Li and Bathen further teach:
The method of claim 1, wherein at least one of the one or more computing devices is part of a communications network. (Bathen [0030])
As per claim 14, it is the system variant of claim 1 and is therefore rejected under the same rationale.
As per claim 25, Li discloses: A collaborative machine learning, ML, system comprising a resource management controller for computer resource allocation and one or more computing devices, wherein the resource management controller comprises processing circuitry, one or more interfaces and a memory containing instructions executable by the processing circuitry, whereby the resource management controller is operable to:
receive registration requests from one or more computing devices seeking to participate in the collaborative ML system, wherein the registration requests comprise smart contracts based on a blockchain system; (Li [0123]: “In step 708 the new arrival node performs node registration, e.g., the new arrival node registers itself to the blockchain… said account including account information, said account information being an input to a resource allocation smart contract… step 712, in which the new arrival node broadcasts the account information on blockchain. Since each node in the provider network owns the same copy of the smart contact, along with the requesting node, e.g. node A, registration, each of the nodes in the network are updating the same smart contract. For example node A 504, node C 508, node D 510 are able to update their copy of the smart contract based on the information broadcast in step 712 by new arrival node B 506.”)
register the one or more computing devices; (Li [0124]: “In step 714 the new arrival node receives a service request for a computing task, e.g., a machine learning (ML) training task. For example, the received service request is a service request from node A 504 which is a service request for a computing task through the blockchain, e.g. a machine learning task, This request is an input signal to the resource allocation smart contract. The request information communicated in the received request reflects the needed (estimated) computational load, For example, in machine learning training task in some embodiments, the request includes the sizes of the neural network, dataset, and the format of the data, e.g., images, voices, etc.”; [0125]: “In step 716 the new arrival node writes the received service request into the resource allocation smart contract.”)
allocate computing resources provided by the one or more computing devices to the collaborative ML system, and tracking the allocation of resources using the smart contracts; (Li [0126]: “In step 718 the new arrival node runs a resource allocation method to select a number of operation nodes for the computing task. Having the information for the service request and the registration information from each of the registered nodes in the network each node including the new arrival node, e.g., node B, runs the same copy of the resource allocation contract to select a number of nodes that will together complete the computing task. This node selection process is essentially based on system capability of each registered node. The output of the smart contract is a set of node IDs, such as blockchain account (i.e., a user specific public key). Step 718 includes step 720 in which the new arrival node selects a number of nodes that will together complete the computing task. Step 720 include step 722 in which the new arrival node determines a set of node IDs, e.g., blockchain account IDs, corresponding to the selected nodes that will perform the computing task. Operation proceeds from step 718, via connecting node A 724 to step 726.”)
Li did not explicitly disclose:
wherein the one or more computing devices comprise at least one of: an Internet of Things, IoT, device; an edge server; and a database,
receive updated information on at least one of: the available computing resources provided by the one or more computing devices, and the collaborative ML system resource requirements;
and update the allocation of computing resources to the collaborative ML system based on the updated information, and tracking the updated allocation of resources using the smart contracts.
However, Bathen teaches:
wherein the one or more computing devices comprise at least one of: an Internet of Things, IoT (Bathen [0030]), device; an edge server; and a database,
receiving updated information on at least one of: the available computing resources provided by the one or more computing devices, and the collaborative ML system resource requirements; and updating the allocation of computing resources to the collaborative ML system based on the updated information, and tracking the updated allocation of resources using the smart contracts. (Bathen figure 4B and [0029] – [0030])
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of Bathen into that of Li in order to receive updated information on at least one of: the available computing resources provided by the one or more computing devices, and the collaborative ML system resource requirements; and updating the allocation of computing resources to the collaborative ML system based on the updated information, and tracking the updated allocation of resources using the smart contracts. Li figure 7 and 8 teaches the smart contracts are automatically updated whenever a new arrival node wish to be registered to the blockchain. However, one of ordinary skill in the art can easily see that other criteria that can lead to update of the smart contract in blockchain can equally be used here without deviating from the teaching of Li, such as change of resource usage based on an optimization goal as taught by Bathen, applicants here have merely claimed the combination of known parts in the field to achieve predictable results and is therefore rejected under 35 USC 103.
As per claim 26, the combination of Li and Bathen further teach:
The system of claim 25, wherein at least one of the one or more computing devices is part of a communications network. (Bathen [0030])
As per claim 27, the combination of Li and Bathen further teach:
A non-transitory computer-readable medium comprising instructions which, when executed on a computer, cause the computer to perform the method of claim 1. (Li [0186] – [0187])
Claim(s) 5 and 6 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li and Bathen, and further in view of Li et al (US 20220060390, hereinafter Li X).
As per claim 5, the combination of Li and Bathen did not teach:
The method of claim 4, wherein the collaborative ML system utilizes split learning, and wherein: the configuration of the collaborative ML system comprises determining a cutting layer, and communicating the cutting layer to the one or more computing devices participating in the collaborative ML system; the updated information on the collaborative ML system resource requirements comprises an indication that a training iteration has been completed; and the updated allocation of computing resources to the collaborative ML system comprises updated device and ML model information to be used for a next training iteration.
However, Li X teaches:
The method of claim 4, wherein the collaborative ML system utilizes split learning, and wherein: the configuration of the collaborative ML system comprises determining a cutting layer, and communicating the cutting layer to the one or more computing devices participating in the collaborative ML system; the updated information on the collaborative ML system resource requirements comprises an indication that a training iteration has been completed; and the updated allocation of computing resources to the collaborative ML system comprises updated device and ML model information to be used for a next training iteration. (Li X [0008])
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of Li X into that of Li and Bathen in order to have the collaborative ML system utilizes split learning, and wherein: the configuration of the collaborative ML system comprises determining a cutting layer, and communicating the cutting layer to the one or more computing devices participating in the collaborative ML system; the updated information on the collaborative ML system resource requirements comprises an indication that a training iteration has been completed; and the updated allocation of computing resources to the collaborative ML system comprises updated device and ML model information to be used for a next training iteration. Li X has shown the claimed limitations are commonly known in the field for ML model deployment, applicants here have merely claimed the combination of known parts in the field to achieve predictable results and is therefore rejected under 35 USC 103.
As per claim 6, the combination of Li and Bathen did not teach:
The method of claim 4, wherein the collaborative ML system utilizes federated learning and the one or more computing devices comprise a plurality of computing devices, and wherein: the configuration of the collaborative ML system comprises identifying, among the plurality of computing devices, at least one client device for performing local ML model training and at least one leader device for performing global model derivation; the updated information on the collaborative ML system resource requirements comprises an indication that a training iteration has been completed; and the updated allocation of computing resources to the collaborative ML system comprises updated device and ML model information to be used for a next training iteration.
However, Li X teaches:
The method of claim 4, wherein the collaborative ML system utilizes federated learning and the one or more computing devices comprise a plurality of computing devices, and wherein: the configuration of the collaborative ML system comprises identifying, among the plurality of computing devices, at least one client device for performing local ML model training and at least one leader device for performing global model derivation; the updated information on the collaborative ML system resource requirements comprises an indication that a training iteration has been completed; and the updated allocation of computing resources to the collaborative ML system comprises updated device and ML model information to be used for a next training iteration. (Li X [0008])
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of Li X into that of Li and Bathen in order to have the collaborative ML system utilizes federated learning and the one or more computing devices comprise a plurality of computing devices, and wherein: the configuration of the collaborative ML system comprises identifying, among the plurality of computing devices, at least one client device for performing local ML model training and at least one leader device for performing global model derivation; the updated information on the collaborative ML system resource requirements comprises an indication that a training iteration has been completed; and the updated allocation of computing resources to the collaborative ML system comprises updated device and ML model information to be used for a next training iteration. Li X has shown the claimed limitations are commonly known in the field for ML model deployment, applicants here have merely claimed the combination of known parts in the field to achieve predictable results and is therefore rejected under 35 USC 103.
Claim(s) 9 – 11 is/are rejected under 35 U.S.C. 103 as being unpatentable over Li and Bathen, and further in view of Hanna et al (US 20210383261, hereinafter Hanna).
As per claim 9, the combination of Li and Bathen did not teach:
The method of claim 1 further comprising receiving a determination that the accuracy of a ML model trained using the collaborative ML system has reached a predetermined threshold accuracy.
However, Hanna teaches:
The method of claim 1 further comprising receiving a determination that the accuracy of a ML model trained using the collaborative ML system has reached a predetermined threshold accuracy. (Hanna [0117])
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of Hanna into that of Li and Bathen in order to receive a determination that the accuracy of a ML model trained using the collaborative ML system has reached a predetermined threshold accuracy. Li [0080] teaches training of the ML, Hannah [0117] teaches that it is commonly known in the field that models can be trained up to a threshold prior to deployment, applicants here have merely claimed the combination of known parts in the field to achieve predictable results and is therefore rejected under 35 USC 103.
As per claim 10, the combination of Li and Bathen did not teach:
The method of claim 8 further comprising, when the ML model is determined to have reached the predetermined accuracy threshold, storing the ML model in a database.
However, Hanna teaches:
The method of claim 8 further comprising, when the ML model is determined to have reached the predetermined accuracy threshold, storing the ML model in a database. (Hanna [0117])
It would have been obvious for one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate the teaching of Hanna into that of Li and Bathen in order to have the ML model is determined to have reached the predetermined accuracy threshold, storing the ML model in a database. Li [0080] teaches training of the ML, Hannah [0117] teaches that it is commonly known in the field that models can be trained up to a threshold prior to deployment, applicants here have merely claimed the combination of known parts in the field to achieve predictable results and is therefore rejected under 35 USC 103.
As per claim 11, the combination of Li, Bathen and Hanna further teach:
The method of claim 9 further comprising: receiving a request from a user for a trained specific ML model, wherein the specific ML model is to be trained using the collaborative ML system, and deploying a model delivery smart contract; and when the specific ML model is determined to have been trained to reach the predetermined accuracy threshold, providing the trained specific ML model to the user. (Li [0123] – [0124] and Hannah [0117])
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Cheng et al (US 20210182872) teaches “storing a smart contract in a blockchain of the blockchain network. The smart contract corresponds to a logistic process. A transaction information about a transaction in the logistic process is verified and validated based on the smart contract and is hashed into the blockchain upon consensus being concluded.”;
Zhao et al (US 20200174840) teaches “receiving a client request to provision resources for executing a computing job, provisioning accelerator resources of one or more accelerator server nodes in the distributed computing system to perform tasks associated with an execution flow of the computing job, and provisioning a logical nodes within the distributed computing system to compose a data flow pipeline which is configured to perform data flow operations associated with the computing job for providing data to the provisioned accelerator resources to perform the tasks associated with the execution flow of the computing job. The data flow operations include, e.g., data storage input/output operations, data pre-processing operations, and data staging operations, which are decoupled from the execution flow of the computing job.”;
Ramasamy et al (US 20190333030) teaches “Data corresponding to digital tokens may be stored on a blockchain associated with a decentralized peer-to-peer (P2P) network. Smart contracts for collecting digital tokens in exchange for service provider tokens may be deployed to the blockchain. Smart contracts for distributing digital tokens based on credit tokens may be deployed to the blockchain. Through execution of the smart contracts, digital tokens may be utilized by users in various ways.”.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to CHARLES M SWIFT whose telephone number is (571)270-7756. The examiner can normally be reached Monday - Friday: 9:30 AM - 7PM.
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, April Blair can be reached at 5712701014. 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.
/CHARLES M SWIFT/Primary Examiner, Art Unit 2196
Prosecution Insights