DETAILED ACTION
This action is responsive to the following communications: Original Application filed on February 16, 2024. All references to this application refer to the U.S. Patent Application Publication No. 2025/0094840 A1.
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Claims 1-9 are pending in this case. Claim 9 is the independent claim. Claims 1-9 are rejected.
Priority
Receipt is acknowledged of certified copies of papers required by 37 CFR 1.55. Applicants properly claim the benefit of Korean Patent Application No. 10-2023-0122651, filed on September 14, 2023.
Information Disclosure Statement
The information disclosure statement (IDS) filed February 16, 2024, fails to comply with 37 CFR 1.98(a)(3)(i) because it does not include a concise explanation of the relevance, as it is presently understood by the individual designated in 37 CFR 1.56(c) most knowledgeable about the content of the information, of each reference listed that is not in the English language, specifically, KR 10-2023-0068989A.
The IDS indicates that an English translation of the abstract was provided, however, the paper filed does not contain any such English translation. Accordingly, KR 10-2023-0068989A has been ‘struck-through.’
The remaining references have been considered as indicated in the annotated IDS.
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.
Claims 4-9 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being incomplete for omitting essential steps, such omission amounting to a gap between the steps. See MPEP § 2172.01. The omitted step is the recitation of how to make the selection of STL vs MTL models, and the conditions underpinning that determination. This step should be recited in dependent claim 4, or alternatively as a new added claim expressly reciting the omitted step.
Dependent claims 5-9 are rejected solely due to their dependence from a rejected parent claim.
To expedite a complete examination of the instant application, the claims rejected above under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor, or for pre-AIA the applicant regards as the invention, are further rejected as set forth below in anticipation of amendments to these claims to correct the failure.
Examiner’s Note
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.
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.
This application currently names joint inventors. In considering patentability of the claims the Examiner presumes that the subject matter of the various claims was commonly owned as of the effective filing date of the claimed invention(s) absent any evidence to the contrary. Applicants are advised of the obligation under 37 CFR 1.56 to point out the inventor and effective filing dates of each claim that was not commonly owned as of the effective filing date of the later invention in order for the Examiner to consider the applicability of 35 U.S.C. 102(b)(2)(C) for any potential 35 U.S.C. 102(a)(2) prior art against the later invention.
Claims 1 and 2 are rejected under 35 U.S.C. 103 as being unpatentable over U.S. Patent Application Publication No. 2018/0365575 A1, filed by Guttmann e al., on July 30, 2018, and published on December 20, 2018 (hereinafter Guttmann), in view of U.S. Patent Application Publication No. 2024/0370781 A1, filed by Padmanabha Iyer et al., on June 9, 2023, and published on November 7, 2024 (hereinafter Iyer).
With respect to independent claim 1, Guttmann discloses an inference model selection method performed by a computing device comprising a processor, the interference model selection method comprising:
Monitoring computing resources of the computing device; Guttmann discloses monitoring computing resources of the computing device (see Guttman, paragraphs 0145-0148 [describing Fig. 9: Step 910 concerns obtaining information related to available resources (e.g., available memory, size of physical memory, types of memory types, speed/latency of memory, availability of processing units, types of processing units, instructions capable of being performed in a time period, pipeline stages, cache memory size, bus speed, etc.].
Receiving a task; Guttmann discloses receiving one or more tasks (see Guttmann, paragraphs 0106-0108 [describing receipt of tasks awaiting execution, the tsks comprising scheduling information, task priorities, task ordering, preferred execution times, constraints on execution times, preferred execution frequencies, constraints on execution frequencies, further comprising execution requirements, such as required resources]).
Selecting an inference model to perform inference for the received task; Guttmann discloses selection of an inference model for the received task (see Guttmann, paragraphs 0149-0153 [describing the selection of inference models from a plurality of inference models based on the received tasks and available resources, including characteristics of the model]).
Although Guttman discloses scheduling tasks using priorities, ordering, execution times, latency, etc. (see Guttmann, paragraphs 0106-0108, described supra), Guttman fails to expressly disclose inputting the task into a queue of the selected inference model.
However, Iyer teaches using task queues for each inference model, and enqueuing the received task requests into the queue for the selected model (see Iyer, paragraphs 0050-0052 [describing the receipt of task requests and assignment of each request to a designated model associated with the task based on the model having sufficient resources to execute the task request (such as latency, throughput, etc.), and placed onto each model’s queue for scheduling and execution based on task type, the task requests typically executed in batches based on latency or throughput thresholds], 0055 [the dispatcher performs input-aware scheduling to ensure that the batch sizes are more uniform and therefore optimal], 0056 [describing the improvement of using input-aware scheduling, which minimizes padding and ensures that a batch of tasks is performed within 12ms], and 0057 [describing the accumulation of task requests against a threshold batch size and constantly determines if additional tasks increase latency beyond a threshold]).
Accordingly, it would have been obvious to one of ordinary skill in the art, having the teachings of Guttmann and Iyer before him before the effective filing date of the claimed invention, to modify the method of Guttmann to incorporate task request queueing as taught by Iyer. One would have been motivated to make such a combination because this minimizes compute waste and reduces latency, as taught by Iyer (see Iyer, paragraph 0055, described supra).
Guttmann, as modified by Iyer, further teaches performing an inference operation.
Guttmann further teaches performing inference operations (see Guttmann, paragraphs 0154-0157 [describing Step 930 of Fig. 9, in which the inference model performs the inference operation]).
With respect to dependent claim 2, Guttmann, as modified by Iyer, teaches the inference model selection method of claim 1, as described above.
Guttman further teaches the method wherein the monitoring is periodically performed at predetermined periods.
Guttmann further teaches the resource monitoring is periodically performed at predetermined periods (see Guttmann, paragraph 0114 [policies determine the periodicity of resource monitoring]).
Claim 3 is rejected under 35 U.S.C. 103 as being unpatentable over Guttman, in view of Iyer, further in view of U.S. Patent Application Publication No. 2024/0037930 A1, filed by Jacob et al., on July 29, 2022, and published on February 1, 2024 (hereinafter Jacob).
With respect to dependent claim 3, Guttmann, as modified by Iyer, teaches the inference model selection method of claim 2, as described above.
Guttmann further teaches the method wherein the selecting of the inference model comprises selecting an appropriate inference model for tasks received over the predetermined period.
Guttmann further teaches selecting the appropriate inference model based on the tasks received over the period (see Guttmann, paragraphs 0149-0153, described supra, claim 1).
Guttmann and Iyer fail to further teach the inference model includes at least one Single-task Learning (STL) model and at least one Multi-task Learning (MTL) model.
However, Jacob teaches the inference model including both STL and MTL models (see Jacob, Fig. 1; see also, Jacob, paragraphs 0027-0028 [defining the problems of negative transfer and how to combined STL.MTL architecture, which trains MTL and STL models simultaneously overcomes this degradation], 0031 [online distillation includes adaptive feature distillation between STL and MTL models], 0033 [describing the loss function to be minimized using the end-to-end model], and 0048 [describing Fig. 1, which shows the framework as including an MTL model and a plurality of STL models])
Accordingly, it would have been obvious to one of ordinary skill in the art, having the teachings of Guttmann, Iyer, and Jacob before him before the effective filing date of the claimed invention, to modify the method of Guttmann, as modified by Iyer, to incorporate a combined STL and MTL inference model as taught by Jacob. One would have been motivated to make such a combination because this minimizes “negative transfer” due to inappropriate STL/MTL model training, as taught by Jacob (see Jacob, paragraph 0027, described supra).
Claims 4-9 are rejected under 35 U.S.C. 103 as being unpatentable over Guttman, in view of Iyer, further in view of Jacob, further in view of Non-Patent Literature reference entitled “Q-Learning Algorithms: A Comprehensive Classification and Applications,” published by Jang et al., on September 27, 2019 (hereinafter Jang).
With respect to dependent claim 4, Guttmann, as modified by Iyer and Jacob, teaches the inference model selection method of claim 3, as described above.
Guttman, Iyer, and Jacob fail to further teach the method
Wherein the selecting of the inference model comprises solving a Q-function derived as a result of transforming a Markov Decision Process (MDP) model to Q-Learning,
The Q-function is represented as Q*(S,A) = maxπ[Qπ(S,A)], where S denotes a state space, A denotes an action space, and π denotes a policy.
However, Jang teaches solving a Q-Learning function as a function of a state/action pair and a policy (see Jang, Section II.A. [describing the background of MDPs and the state transition probability matrix that using the state space and action pair to determine the reward based on the policy] and Section II.C [describing the Bellman Optimality Equation for determining the maximum expected value]).
Accordingly, it would have been obvious to one of ordinary skill in the art, having the teachings of Guttmann, Iyer, Jacob, and Jang before him before the effective filing date of the claimed invention, to modify the method of Guttmann, as modified by Iyer and Jacob, to incorporate Q-Learning to calculate the reward for the MDP as taught by Jang. One would have been motivated to make such a combination because this minimizes “negative transfer” due to inappropriate STL/MTL model training, as taught by Jang (see Jang, paragraph 0027, described supra).
Examiner’s Note:
Each of dependent claims 5-8 are contingent claims, in that they recite a step that is only performed upon the satisfaction of some condition.
However, the broadest reasonable interpretation of a contingent claim includes an interpretation where the condition is NOT met. See MPEP 2111.04(II) and Ex parte Schulhauser, PTAB Appeal 2013-007847 (PTAB April 28, 2016).1
Accordingly, as each of claims 5-8 are of the format <perform step> if <condition is met> and fails to recite steps to be performed when the condition is NOT met, the broadest reasonable interpretation of the claim requires nothing to be performed.
With respect to dependent claim 5, Guttmann, as modified by Iyer, Jacob, and Jang, teaches the inference model selection method of claim 4, as described above.
Jacob further teaches the method wherein the performing of the inference operation comprises performing the inference operation by inputting a task input to a queue to the STL model without a waiting time if the STL model is selected.
Jacob further teaches using STL models (see Jacob, Fig. 1; see also, Jacob, paragraphs 0027-0028, 0031, 0033, and 0048, described supra, claim 3).
With respect to dependent claim 6, Guttmann, as modified by Iyer, Jacob, and Jang, teaches the inference model selection method of claim 5, as described above.
Jacob further teaches the method wherein the performing of the inference operation comprises performing the inference operation by, when the queue is full of tasks, inputting the task input to the queue to the MTL model if the MTL model is selected.
Jacob further teaches using MTL models (see Jacob, Fig. 1; see also, Jacob, paragraphs 0027-0028, 0031, 0033, and 0048, described supra, claim 3).
With respect to dependent claim 7, Guttmann, as modified by Iyer, Jacob, and Jang, teaches the inference model selection method of claim 5, as described above.
Jacob further teaches the method wherein the performing of the inference operation comprises performing the inference operation by, when a sum of a time elapsed from a point in time at which at least one task stored in the queue is received and a computing time required for inference is equal to a value acquired by subtracting a predetermined period of time from a required delay time, inputting all the tasks stored in the queue to the MTL model, if the MTL model is selected.
Jacob further teaches using MTL models (see Jacob, Fig. 1; see also, Jacob, paragraphs 0027-0028, 0031, 0033, and 0048, described supra, claim 3).
With respect to dependent claim 8, Guttmann, as modified by Iyer, Jacob, and Jang, teaches the inference model selection method of claim 5, as described above.
Jacob further teaches the method wherein the performing of the inference operation comprises performing the inference operation by, when a sum of a waiting time of a task in the queue and a computing time required for inference is equal to a value acquired by subtracting a predetermined period of time from a required delay time, inputting all the tasks stored in the queue to the MTL model, if the MTL model is selected.
Jacob further teaches using MTL models (see Jacob, Fig. 1; see also, Jacob, paragraphs 0027-0028, 0031, 0033, and 0048, described supra, claim 3).
With respect to dependent claim 9, Guttmann, as modified by Iyer, Jacob, and Jang, teaches the inference model selection method of claim 8, as described above.
Guttmann further teaches the method, further comprising: transmitting an inference result.
Guttmann further teaches transmitting an inference result (see Guttmann, paragraphs 0154-0157, described supra, claim 1).
Conclusion
The prior art made of record and not relied upon is considered pertinent to Applicants’ disclosure. See PTO-892.
It is noted that any citation to specific pages, columns, figures, or lines in the prior art references any interpretation of the references should not be considered to be limiting in any way. A reference is relevant for all it contains and may be relied upon for all that it would have reasonably suggested to one having ordinary skill in the art. In re Heck, 699 F.2d 1331-33, 216 USPQ 1038-39 (Fed. Cir. 1983) (quoting In re Lemelson, 397 F.2d 1006, 1009, 158 USPQ 275, 277 (CCPA 1968)).
Any inquiry concerning this communication or earlier communications from the Examiner should be directed to ERIC J. BYCER whose telephone number is (571) 270-3741. The Examiner can normally be reached Monday - Thursday 9am-6pm, and alternate Fridays 9am-5pm.
Examiner interviews are available via a variety of formats. See MPEP § 713.01. To schedule an interview, Applicants are encouraged to use the USPTO Automated Interview Request (AIR) Form at https://www.uspto.gov/InterviewPractice.
If attempts to reach the Examiner by telephone are unsuccessful, the Examiner’s supervisor, MATT ELL can be reached on (571) 270-3264. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from Patent Center. Status information for published applications may be obtained from Patent Center. Status information for unpublished applications is available through Patent Center to authorized users only. Should you have questions about access to the USPTO patent electronic filing system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
/ERIC J. BYCER/
Primary Examiner
Art Unit 2141
1 “When analyzing the claimed method as a whole, the PTAB determined that giving the claim its broadest reasonable interpretation, "[i]f the condition for performing a contingent step is not satisfied, the performance recited by the step need not be carried out in order for the claimed method to be performed" (quotation omitted). Schulhauser at 10.” MPEP 2111.04(II).
Prosecution Insights