DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Status of Claims
Claims 1 and 5-20 are pending in this Office Action.
Response to Arguments
Applicant's arguments filed in the amendment filed on 03/02/2026, have been fully considered but they are moot in view of the new grounds of rejections. The reasons are set forth below.
Drawings
The formal drawings received on 03/27/2023 have been entered.
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.
Claim(s) 15-20 is/are rejected under 35 U.S.C. 101 because the claimed invention is directed to non-statutory subject matter. The claim(s) does/do not fall within at least one of the four categories of patent eligible subject matter because claim 15 recites “A computer program product comprising a computer readable program stored on a computer readable storage medium ….” The computer program product comprises a computer readable program and does not include a computer readable storage medium. The computer readable storage medium is interpreted as being external to the computer program product. As such the claim may be fairly construed to constitute software only/software per se, i.e. non-statutory subject matter that falls outside of the four categories of patent eligible subject matter. Therefore, the claim(s) does/do not fall within at least one of the four categories of patent eligible subject matter because the claim(s) is/are not directed to either a process, machine, manufacture, or composition of matter.
Claim Rejections - 35 USC § 112
The following is a quotation of the first paragraph of 35 U.S.C. 112(a):
(a) IN GENERAL.—The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor or joint inventor of carrying out the invention.
The following is a quotation of the first paragraph of pre-AIA 35 U.S.C. 112:
The specification shall contain a written description of the invention, and of the manner and process of making and using it, in such full, clear, concise, and exact terms as to enable any person skilled in the art to which it pertains, or with which it is most nearly connected, to make and use the same, and shall set forth the best mode contemplated by the inventor of carrying out his invention.
Claims 1 and 5-20 are rejected under 35 U.S.C. 112(a) or 35 U.S.C. 112 (pre-AIA ), first paragraph, as failing to comply with the written description requirement. The claim(s) contains subject matter which was not described in the specification in such a way as to reasonably convey to one skilled in the relevant art that the inventor or a joint inventor, or for applications subject to pre-AIA 35 U.S.C. 112, the inventor(s), at the time the application was filed, had possession of the claimed invention.
Claims 1, 8, 15 recite “generating a reconfigured version of the set of compressed parameters of the ML model by replacing the PGCD with the one or more PGCD identifiers…,” which are not supported by the applicant’s original disclosure. The Examiner reviewed the applicant’s disclosure and remarks but was not able to find the support for the limitations in paragraphs [0004], [0030], [0063], [0064]. The specification neither discloses the concept of reconfiguration version nor generating a reconfigured version of the set of compressed parameters.
Claims 1, 8, 15 recite “… transmitting the reconfigured version of the set of compressed parameters …,” which are not supported by the applicant’s original disclosure. The Examiner reviewed the applicant’s disclosure but was not able to find the support for the limitations in paragraphs [0004], [0030], [0063], [0064]. The specification neither discloses the concept of reconfigured version nor transmitting the reconfigured version of the set of compressed parameters.
Claim Rejections - 35 USC § 103
In the event the determination of the status of the application as subject to AIA 35 U.S.C. 102 and 103 (or as subject to pre-AIA 35 U.S.C. 102 and 103) is incorrect, any correction of the statutory basis (i.e., changing from AIA to pre-AIA ) for the rejection will not be considered a new ground of rejection if the prior art relied upon, and the rationale supporting the rejection, would be the same under either status.
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claim(s) 1, 5, 8-12, 15-19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Lee (US 20230196205) in view of Denoue (US 20100268694), and further in view of Aasted (US 20100057750).
Claims 15, 8. Lee teaches:
A computer program product comprising a computer readable program stored on a computer readable storage medium, wherein the computer readable program, when executed on a computing environment comprising a server electronically coupled to one or more hardware computing devices, causes the server to perform operations comprising: – in paragraphs [0008]-[0186] (The computing system 1000 according to an embodiment of the present invention may include at least one processor 1100 and a memory 1200 stores instructions for invoking the at least one processor 1100 to perform at least one step of the method according to an embodiment of the present invention. The at least one step of the method may be performed by the at least one processor 1100 loading and executing instructions from the memory 1200.)
determining, using the server, a set of compressed parameters of a machine learning (ML) model; – in paragraphs [0008]-[0186] (Generating, by the parameter server 100, compressed global model parameters by compressing the parameters of the global model 110 in step S340.)
wherein the one or more hardware computing devices comprise a target device; – in paragraphs [0008]-[0186] (Transmitting, by the parameter server 100, the compressed global model parameters to the local devices 200 in step S350. The computer-readable recording medium may also include a hardware device specially configured to store and execute program instructions, such as a read only memory (ROM), a random access memory (RAM), and a flash memory.)
Lee does not explicitly teach:
determining, using the server, that the set of compressed parameters of the ML model comprises new compressed data (NCD) and previously-generated compressed data (PGCD); wherein the NCD comprises compressed data segments that were not previously transmitted to the target device; wherein the PGCD comprises compressed data segments that were previously transmitted to the target device; computing one or more PGCD identifiers based at least in part on the PGCD; the reconfigured version of the set of comprised parameters of the ML model comprising the PGCD identifiers and the NCD.
However, Denoue teaches:
determining, using the server, that the set of compressed parameters of the ML model comprises new compressed data (NCD) and previously-generated compressed data (PGCD); – in paragraphs [0010]-[0172] (The presenter computer system 602 then determines (614) whether the signature of the compressed tile is different than a previously computed signature of the previous version of the compressed tile. In response to determining that the signature for the compressed tile is different than the previously computed signature of the previous version of the compressed tile (616, yes), the presenter computer system determines (618) that the compressed tile has changed. The presenter computer system 602 marks the tiles that have changed so that the presenter computer system 602 can send the tiles marked as changed to the server 604. After step 618 or in response to determining that the signature for the tile is not different than the previously computed signature of the previous version of the tile (616, no), the presenter computer system 602 sends (620) tiles that have changed to the server 604.)
wherein the NCD comprises compressed data segments that were not previously transmitted to the target device; – in paragraphs [0010]-[0172] (The presenter computer system 602 then determines (614) whether the signature of the compressed tile is different than a previously computed signature of the previous version of the compressed tile. In response to determining that the signature for the compressed tile is different than the previously computed signature of the previous version of the compressed tile (616, yes), the presenter computer system determines (618) that the compressed tile has changed. The presenter computer system 602 marks the tiles that have changed so that the presenter computer system 602 can send the tiles marked as changed to the server 604.)
wherein the PGCD comprises compressed data segments that were previously transmitted to the target device; – in paragraphs [0010]-[0172] (After step 618 or in response to determining that the signature for the tile is not different than the previously computed signature of the previous version of the tile (616, no), the presenter computer system 602 sends (620) tiles that have changed to the server 604.)
computing one or more PGCD identifiers based at least in part on the PGCD; – in paragraphs [0010]-[0172] (The presenter computer system 602 computes the signature of the compressed tile by computing a hash value of the compressed tile using a hash function. The presenter computer system 602 then determines (614) whether the signature of the compressed tile is different than a previously computed signature of the previous version of the compressed tile.)
the reconfigured version of the set of comprised parameters of the ML model comprising the PGCD identifiers and the NCD; and – in paragraphs [0010]-[0172] (The presenter computer system 602 computes the signature of the compressed tile by computing a hash value of the compressed tile using a hash function. The presenter computer system 602 then determines (614) whether the signature of the compressed tile is different than a previously computed signature of the previous version of the compressed tile. The presenter computer system 602 then determines (614) whether the signature of the compressed tile is different than a previously computed signature of the previous version of the compressed tile. In response to determining that the signature for the compressed tile is different than the previously computed signature of the previous version of the compressed tile (616, yes), the presenter computer system determines (618) that the compressed tile has changed. The presenter computer system 602 marks the tiles that have changed so that the presenter computer system 602 can send the tiles marked as changed to the server 604.)
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Lee with Denoue to include determining, using the server, that the set of compressed parameters of the ML model comprises new compressed data (NCD) and previously-generated compressed data (PGCD); wherein the NCD comprises compressed data segments that were not previously transmitted to the target device; wherein the PGCD comprises compressed data segments that were previously transmitted to the target device; computing one or more PGCD identifiers based at least in part on the PGCD; the reconfigured version of the set of comprised parameters of the ML model comprising the PGCD identifiers and the NCD, as taught by Denoue, in paragraphs [0001]-[0068], to process and transmit data using a technique which does not require a substantial amount of bandwidth, thereby reducing bandwidth usage and storage requirements.
Combination of Lee and Denoue does not explicitly teach:
generating a reconfigured version of the set of compressed parameters of the ML model by replacing the PGCD with the one or more PGCD identifiers, instead of transmitting the set of compressed parameters of the ML model to the target device, transmitting the reconfigured version of the set of compressed parameters of the ML model to the target device.
However, Aasted teaches:
generating a reconfigured version of the set of compressed parameters of the ML model by replacing the PGCD with the one or more PGCD identifiers, – in paragraphs [0005]-[0056] (Generating file hashes from said compressed file blocks. Substituting standard-block identifiers (SBIDs) for at least some blocks of a file to be transferred. SBIDs 29 represent respective uncompressed standard blocks or "SU-blocks" 30, and, thus, compressed versions of those blocks, namely, compressed standard blocks or "S-blocks" 31, which are stored at standard block storage 32. Source system encoder 21 is responsible for including SBIDs in recipe file 23. Encoder 21 first compresses the blocks and then generates hashes from the compressed blocks.)
instead of transmitting the set of compressed parameters of the ML model to the target device, transmitting the reconfigured version of the set of compressed parameters of the ML model to the target device. – in paragraphs [0005]-[0056] (Recipe file 23 on source system 11 is transferred (instead of original file 19) to target system 13. A recipe file including SBIDs for matching file blocks and compressed versions of non-matching file blocks is transferred from a source system to a target system.)
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Lee and Denoue with Aasted to include generating a reconfigured version of the set of compressed parameters of the ML model by replacing the PGCD with the one or more PGCD identifiers, instead of transmitting the set of compressed parameters of the ML model to the target device, transmitting the reconfigured version of the set of compressed parameters of the ML model to the target device, as taught by Lee, in paragraphs [0002]-[0051], to fulfill a demand for technology for reducing the cost required to transmit and receive parameters in the process of training artificial intelligence distributed in a number of mobile devices.
Claim 1. Claim is substantially similar to claims 8 and 15, except for the following limitations:
Combination of Lee and Denoue teaches Claims 8 and 15.
Aasted teaches:
wherein a data size of the one or more PGCD identifiers is less than a data size of the PGCD; – in paragraphs [0005]-[0056] (An encoder 21 of source system 11 encodes relatively large original file 19 to yield a relatively small recipe file 23 on source system 11. A recipe file including SBIDs for matching file blocks and compressed versions of non-matching file blocks is transferred from a source system to a target system.)
wherein each of the one or more PGCD identifiers transmitted to the target device comprises a hash value; and – in paragraphs [0005]-[0056] (Recipe file 23 on source system 11 is transferred (instead of original file 19) to target system 13. Source system encoder 21 is responsible for including SBIDs in recipe file 23. Encoder 21 first compresses the blocks and then generates hashes from the compressed blocks.)
wherein the target device is configured to determine a local version of the PGCD based at least in part on the one or more PGCD identifiers. – in paragraphs [0005]-[0056] (A recipe file including SBIDs for matching file blocks and compressed versions of non-matching file blocks is transferred from a source system to a target system. A duplicate of the original file is constructed on the target system in part by exchanging transferred SBIDs for the standard blocks they represent.)
Claims 5, 12, 19. The server of claim 1, – refer to the indicated claim for reference(s).
Lee teaches:
wherein the target device is configured to use the local version the PGCD and the NCD to generate a local version of the set of compressed parameters of the ML model. – in paragraphs [0008]-[0186] (Updating, by each of the local devices 200, parameters of a corresponding local model 210 by training the local model 210 based on a corresponding local dataset D1, D2, or D3 in step S320. Generating, by the local device 200, compressed gradient information by compressing the gradients of the parameters of the local model 210 in step S320.)
Claims 9, 16. Claims 9 and 16 are substantially similar to claim 1.
Claims 10, 17. Claims 10 and 17 are substantially similar to claim 1.
Claims 11, 18. Claims 11 and 18 are substantially similar to claim 1.
Claim(s) 6, 7, 13, 14, 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Lee (US 20230196205) in view of Denoue (US 20100268694), and further in view of Aasted (US 20100057750) and Ben-Itzhak (US 20240296317).
Claims 6, 13. The server of claim 5, – refer to the indicated claim for reference(s).
Combination of Lee, Denoue, and Aasted does not explicitly teach:
wherein the target device is further configured to decompress the local version of the set of compressed parameters of the ML model to generate reconstructed parameters of the ML model.
However, Ben-Itzhak teaches:
wherein the target device is further configured to decompress the local version of the set of compressed parameters of the ML model to generate reconstructed parameters of the ML model. – in paragraphs [0025]-[0034] (At step 618, each participating client can receive the compressed global gradient and decompress (i.e., de-quantize) it using the same linear quantization technique applied at step 610. This step can involve recovering the real coordinates of the global gradient from the quantized coordinates. The client can also apply an inverse transform to the global gradient after decompression if a transform was originally applied to the outgoing gradient at step 608 (step 620).)
It would have been obvious for one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify Lee, Denoue, and Aasted with Ben-Itzhak to include wherein the target device is further configured to decompress the local version of the set of compressed parameters of the ML model to generate reconstructed parameters of the ML model, as taught by Ben-Itzhak, in paragraphs [0001]-[0014], to provide machine learning techniques that allow multiple networked computing devices/systems, referred to as clients, to collaboratively train an artificial neural network (ANN) with the help of a central server, referred to as a parameter server.
Claims 7, 14. The server of claim 6, – refer to the indicated claim for reference(s).
Lee teaches:
wherein the target device is further configured to: use the reconstructed parameters of the ML model to update a local instance of the ML model; and – in paragraphs [0008]-[0186] (Reconstructing, by the parameter server 100, gradients of the parameters of the local model 210 based on the compressed gradient information received from the local device 200 in step S340)
use local data of the target device to further train the updated local instance of the ML model. – in paragraphs [0008]-[0186] (Updating, by the parameter server 100, parameters of a global model 110 based on the reconstructed gradients of the parameters of the local model 210 in step S340. Updating, by each of the local devices 200, parameters of a corresponding local model 210 by training the local model 210 based on a corresponding local dataset D1, D2, or D3 in step S320.)
Claim 20. Claim 20 is substantially similar to claims 6, 7, 13, and 14.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MUHAMMAD RAZA whose telephone number is (571)272-7734. The examiner can normally be reached Monday-Friday, 7:00 A.M.-5:00 P.M..
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, Vivek Srivastava can be reached on (571)272-7304. 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.
/MUHAMMAD RAZA/Primary Examiner, Art Unit 2449