DETAILED ACTION
This Office Action is in response to the amendments made on 02/05/2026.
Claims 4 and 12 are currently cancelled.
Claims 1, 5, 6, 9, 13, 17, and 19 are currently amended.
Claims 1-3, 5-11, and 13-20 are currently pending in this application and have been examined.
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 .
Response to Arguments
In reference to Applicant’s arguments on page(s) 7-19 regarding rejections made under 35 U.S.C. 101:
Claims 1-20 are rejected under 35 U.S.C. 101 because they are allegedly directed toward an abstract idea without significantly more.
Applicant respectfully disagrees.
The Examiner considers that the limitations of claim 1 cover a mathematical calculation and performance of the limitation in the mind (including an observation, evaluation, judgement, or opinion) or with the aid of pencil and paper, and accordingly, claim 1 recites an abstract idea.
Further, the claim does not recite a mental process because the steps are not practically performed in the human mind.
Applicant respectfully submits that:
the claim does not recite any mathematical relationships, formulas, or calculations. While some of the limitations may be based on mathematical concepts, the mathematical concepts are not recited in the claims;
Further, the claim does not recite a mental process because the steps are not practically performed in the human mind.
Clearly, Applicant's claim language is similar to that of the claim in USPTO's Example 39, and therefore the same analysis in USPTO's Example 39 is applicable to Applicant's claim 1, which necessarily leads to the same conclusion that the answer of Step 2A, Prong One is No, and then Step 2A, Prong Two and Step 2B are not applicable.
As described in paragraph [0002] of the subject application, AI models such as deep learning models often have millions or even billions of trainable parameters. "Training such huge models leads to a high memory consumption and computational complexity. Such models also need to be trained on massive datasets which may take a long time. These issues may cause limitations on training AI models on edge devices with limited computational power, and therefore their application in many scenarios becomes limited."
The claimed invention provides a technical solution to this technical problem by selecting a subset of the plurality of data samples based on the sampling probabilities of the plurality of data samples, wherein the sampling probabilities of the plurality of data samples are determined based on the importance metrics thereof, and the importance metrics are calculated based on predictions of the AI model obtained from the plurality of data samples in a plurality of previous training epochs without using labels of the plurality of data samples and without using a learning rate of the AI model.
Thus, the claimed invention provides an improvement to AI technology with various advantages (see, e.g., paragraphs [0117] to [0124]). As the name "artificial intelligence" indicates, such an AI technology is not for human intelligence, and rather is an artificial intelligence particularly for execution by a machine or specifically a computing device.
Therefore, the claimed method, being an improvement to AI technology, is integrated into a practical application that inevitably leads to an improvement in the functioning of a computer.
Accordingly, "when the claim is considered as a whole, the recited judicial exception is integrated into a practical application as determined under either MPEP sections 2106.06(a) or 2106.05(e), such that the claim is not directed to the abstract idea, and thus is patent-eligible."
Therefore, claim 1 (and similarly claims 2-20) are patent eligible under 35 U.S.C. 101. Withdrawal of this rejection is respectfully requested.
Examiner’s response:
Applicant’s arguments have been fully considered but are found to be not persuasive.
Applicant argues that the instant application does not recite mathematical calculations or mental processes. Examiner disagrees. Independent Claims 1, 9, and 17 all recite actions of calculating importance metrics and sampling probabilities. It is explicitly stated in the claims that these calculations are performed based on “previous training epochs” and therefore do not require the use of the machine learning model. The calculating of importance metrics and sampling probabilities are clear mathematical calculations and the claims therefore recite abstract ideas relating to said calculations. Independent Claim 1, 9, and 17 also recite the action of selecting a subset of the data samples based on the previously calculated sampling probabilities, which is simply selecting certain pieces of data that fit a criterion, an action that can be reasonably performed in the human mind and therefore recites and abstract idea.
Applicant argues that the instant application bears similarities to those of Example 39 of the USPTO provided Subject Matter Eligibility Examples. Examiner disagrees. Example 39 was found to be patent eligible because the claim limitations were not found to recite any mathematical relationships, formulas, or calculations and that some of the limitations were merely based on mathematical concepts. This is not the case with the instant application. The instant application explicitly recites steps of calculating values that are used to select a subset of the training data. Not only is this not similar to Example 39, the explicit steps of calculating importance metrics and sampling probabilities very firmly places those limitations into the realm of reciting abstract ideas related to mathematical calculations.
Applicant argues that the instant application provides a technical solution to the problem of training models with too many parameters which leads to a high memory consumption and computational complexity. Examiner disagrees. The solution to the problem of too many data points to process efficiently is not remedied by the instant application because the solution presented in the instant application is to only process a selected number of data points. It is obvious to one skilled in the art that processing less data will lead to lower memory consumption and computational complexity, however, processing less data is not a novel solution to this problem. Furthermore, since the amount of data to be processed relies on the abstract ideas mentioned above, the solution is invalid as an improvement to a technical problem cannot arise from an abstract idea. Applicant cites to PTAB decision to support their claim that this is indeed a technical solution, but fails to recognize that in said decisions that the deciding factor is the additional elements of the claims. The only additional element in support of the instant application is to train the model using the reduced amount of data, which is merely a recitation of the idea of an outcome or a solution with no details as to how the solution to the problem is accomplished.
In light of the amendments made on the claims, the rejections made under 35 U.S.C. 101 are maintained and updated below.
In reference to Applicant’s arguments on page(s) 19 regarding rejections made under 35 U.S.C. 103:
Claims 1-3, 8-11, 16-18, and 20 are rejected under 35 U.S.C. 103 as allegedly being unpatentable over Zhu et al (US 20220383185 Al, hereinafter Zhu), in view of Ma et al (US 20220358347 Al, hereinafter Ma), and in view of Wang et al (Wang, F., Gao, X., Chen, G., & Ye, J. (2017). Accelerate RNN-based Training with Importance Sampling. arXiv [Cs.LG]. Retrieved from http://arxiv.org/abs/1711.00004, hereinafter Wang).
Claims 7 and 15 are rejected under 35 U.S.C. 103 as allegedly being unpatentable over Zhu, Ma, and Wang as applied to claims 1, 9, and 17 above, and further in view of Sharma et al (US 20210374481 Al, hereinafter Sharma).
Applicant respectfully disagrees. Nevertheless, for the purposes of advancing the examination of the subject application, independent claims 1, 9, and 17 are amended by incorporating the limitations of allowable claim 4. Accordingly, claims 1-3, 5-11, and 13-20 are allowable.
Examiner’s response:
Applicant’s arguments have been fully considered and are found to be persuasive.
Applicant rolled an unrejected dependent claim into the independent claims.
In light of the amendments made on the claims, the rejections made under 35 U.S.C. 103 are withdrawn.
Claim Rejections - 35 USC § 101
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claims 1-3, 5-11, and 13-20 rejected under 35 U.S.C. 101 because they are directed to an abstract idea without significantly more.
Step 1 analysis:
Independent claim 1 recites, in part, a method, therefore falling into the statutory category of process. Independent claim 9 recite, in part, one or more processors for performing actions, therefore falling into the statutory category of machine. Independent claim 17 recites, in part, one or more non-transitory computer-readable storage devices comprising computer- executable instructions, wherein the instructions, when executed, cause a processing structure to perform actions, therefore falling into the statutory category of manufacture.
Regarding Claim 1:
Step 2A: Prong 1 analysis:
Claim 1 recites in part:
“calculating importance metrics of a plurality of data samples based on predictions of the AI model obtained from the plurality of data samples in a plurality of previous training epochs without using labels of the plurality of data samples and without using a learning rate of the Al model”. As drafted and under its broadest reasonable interpretation, this limitation covers a mathematical calculation.
“calculating sampling probabilities of the plurality of data samples based on the importance metrics thereof”. As drafted and under its broadest reasonable interpretation, this limitation covers a mathematical calculation.
“selecting a subset of the plurality of data samples based on the sampling probabilities of the of plurality of data samples”. As drafted and under its broadest reasonable interpretation, this limitation covers performance of the limitation in the mind (including an observation, evaluation, judgement, or opinion) or with the aid of pencil and paper. For example, this limitation encompasses selecting data based on previously calculated probabilities.
Accordingly, at Step 2A: Prong 1, the claim is directed to an abstract idea.
Step 2A: Prong 2 analysis:
The judicial exception is not integrated into practical application. In particular, the claim recites the additional elements of:
“training the Al model using the selected subset of the plurality of data samples for one or more epochs”. This additional elements is recited at a high level of generality such that the claim recites only the idea of a solution or outcome (training a model) i.e., the claim fails to recite details of how a solution to a problem is accomplished.
“wherein the importance metric of each data sample of the plurality of data samples is a M-hop divergence of the logits of the Al model obtained from the data sample in the plurality of previous training epochs, where M> 1 is an integer”. This limitation merely indicates a field of use or technological environment in which the judicial exception is performed (m-hop divergence) and thus fails to add an inventive concept to the claims. See MPEP 2106.05(h).
Accordingly at Step 2A: Prong 2, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B analysis:
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception.
As discussed above, the additional element(s) of “training the Al model using the selected subset of the plurality of data samples for one or more epochs” is/are recited at a high-level of generality such that the claim recites only the idea of a solution or outcome (processing data) i.e., the claim fails to recite details of how a solution to a problem is accomplished (See MPEP 2106.05(f)).
The additional element(s) of “wherein the importance metric of each data sample of the plurality of data samples is a M-hop divergence of the logits of the Al model obtained from the data sample in the plurality of previous training epochs, where M> 1 is an integer” is/are directed to particular field(s) of use (m-hop divergence) (MPEP 2106.05(h)) and therefore do not provide significantly more than the abstract idea, and thus the claim is subject-matter ineligible.
Accordingly, at Step 2B, the additional elements individually or in combination do not amount to significantly more than the judicial exception.
Regarding Claim 2:
Due to claim language stating the repetition of steps of Claim 1, Claim 2 is rejected for the same reasons as presented above in the rejection of Claim 1.
Regarding Claim 3:
Step 2A: Prong 1 analysis:Claim 3 recites in part:
“wherein said calculating the importance metrics of the plurality of data samples comprises: calculating the importance metric of each data sample of the plurality of data samples based on logits of the Al model obtained from the data sample in the plurality of previous training epochs”. As drafted and under its broadest reasonable interpretation, this limitation covers a mathematical calculation.
Accordingly, at Step 2A: Prong 1, the claim is directed to an abstract idea.
Step 2A: Prong 2 analysis:
The claim does not recite any additional elements that integrate the judicial exception into a practical application.
Step 2B analysis:
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception.
Regarding Claim 5:
Step 2A: Prong 1 analysis:Claim 5 recites in part:
“wherein the sampling probability of each data sample is a normalized metric calculated from the importance metric of the data sample and shaped using a shaping function”. As drafted and under its broadest reasonable interpretation, this limitation covers a mathematical calculation.
Accordingly, at Step 2A: Prong 1, the claim is directed to an abstract idea.
Step 2A: Prong 2 analysis:
The claim does not recite any additional elements that integrate the judicial exception into a practical application.
Step 2B analysis:
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception.
Regarding Claim 6:
Step 2A: Prong 2 analysis:
The judicial exception is not integrated into practical application. In particular, the claim recites the additional elements of:
“wherein the shaping function is a sharpness-controlling factor or a softmax function”. This limitation merely indicates a field of use or technological environment in which the judicial exception is performed (shaping function) and thus fails to add an inventive concept to the claims. See MPEP 2106.05(h).
Accordingly at Step 2A: Prong 2, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B analysis:
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception.
The additional element(s) of “wherein the shaping function is a sharpness-controlling factor or a softmax function” is/are directed to particular field(s) of use (shaping function) (MPEP 2106.05(h)) and therefore do not provide significantly more than the abstract idea, and thus the claim is subject-matter ineligible.
Accordingly, at Step 2B, the additional elements individually or in combination do not amount to significantly more than the judicial exception.
Regarding Claim 7:
Step 2A: Prong 2 analysis:
The judicial exception is not integrated into practical application. In particular, the claim recites the additional elements of:
“wherein the importance metric of each data sample is an entropy of the predictions of the Al model obtained from the data sample in the plurality of previous training epochs”. This limitation merely indicates a field of use or technological environment in which the judicial exception is performed (importance metrics) and thus fails to add an inventive concept to the claims. See MPEP 2106.05(h).
Accordingly at Step 2A: Prong 2, the additional elements individually or in combination do not integrate the judicial exception into a practical application.
Step 2B analysis:
In accordance with Step 2B, the claim does not include additional elements that are sufficient to amount to significantly more that the judicial exception.
The additional element(s) of “wherein the importance metric of each data sample is an entropy of the predictions of the Al model obtained from the data sample in the plurality of previous training epochs” is/are directed to particular field(s) of use (importance metrics) (MPEP 2106.05(h)) and therefore do not provide significantly more than the abstract idea, and thus the claim is subject-matter ineligible.
Accordingly, at Step 2B, the additional elements individually or in combination do not amount to significantly more than the judicial exception.
Regarding Claim 8:
Due to claim language similar to and stating the repetition of steps of Claim 1, Claim 8 is rejected for the same reasons as presented above in the rejection of Claim 1.
Regarding Claim 9:
Due to claim language similar to that of Claim 1, Claim 9 is rejected for the same reasons as presented above in the rejection of Claim 1.
Regarding Claim 10:
Due to claim language similar to that of Claim 2, Claim 10 is rejected for the same reasons as presented above in the rejection of Claim 2.
Regarding Claim 11:
Due to claim language similar to that of Claim 3, Claim 11 is rejected for the same reasons as presented above in the rejection of Claim 3.
Regarding Claim 13:
Due to claim language similar to that of Claim 5, Claim 13 is rejected for the same reasons as presented above in the rejection of Claim 5.
Regarding Claim 14:
Due to claim language similar to that of Claim 6, Claim 14 is rejected for the same reasons as presented above in the rejection of Claim 6.
Regarding Claim 15:
Due to claim language similar to that of Claim 7, Claim 15 is rejected for the same reasons as presented above in the rejection of Claim 7.
Regarding Claim 16:
Due to claim language similar to that of Claim 8, Claim 16 is rejected for the same reasons as presented above in the rejection of Claim 8.
Regarding Claim 17:
Due to claim language similar to that of Claims 1 and 9, Claim 17 is rejected for the same reasons as presented above in the rejection of Claims 1 and 9.
Regarding Claim 18:
Due to claim language similar to that of Claims 2 and 10, Claim 18 is rejected for the same reasons as presented above in the rejection of Claims 2 and 10.
Regarding Claim 19:
Due to claim language similar to that of Claims 3, 4, 11, and 12, Claim 19 is rejected for the same reasons as presented above in the rejection of Claims 3, 4, 11, and 12.
Regarding Claim 20:
Due to claim language similar to that of Claims 8 and 16, Claim 20 is rejected for the same reasons as presented above in the rejection of Claims 8 and 16.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US 20220383185 A1 – Hessian matrix-free sample-based techniques for model explanations that are faithful to the model
US 20220358347 A1 – a novel framework that provides mechanisms for a Deep & Cross Network (DCN) framework that performs distilled deep prediction for personalized stream ranking on portal websites
Wang, F., Gao, X., Chen, G., & Ye, J. (2017). Accelerate RNN-based Training with Importance Sampling. arXiv [Cs.LG]. Retrieved from http://arxiv.org/abs/1711.00004 – a novel Fast-Importance-Mining algorithm to calculate the importance factor for unstructured data which makes the application of IS in RNN-based applications possible
US 12242947 B2 – a computer-implemented method of processing an input data item
US 20230342607 A1 – a method and a system for training a machine learning system
US 20230186078 A1 – dynamic user interfaces for use in machine-learning and, in particular embodiments, to a system, method and computer program product for dynamic user interfaces for RNN-based deep reinforcement machine-learning models
US 20210312323 A1 – a technique for generating a performance prediction of a machine learning model with uncertainty intervals includes obtaining a first model configured to perform a task and a production dataset
US 20200401916 A1 – generative and inference machine learning models with discrete-variable latent spaces
US 20180018538 A1 – a feature transformation device, a recognition device, a feature transformation method and a computer readable recording medium
US 20150356464 A1 – supervised machine learning and more specifically generating artificial data samples for a minority data class from an imbalanced training data set to train a multi-class classifier model of a supervised machine learning program
THIS ACTION IS MADE FINAL. 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 COREY M SACKALOSKY whose telephone number is (703)756-1590. The examiner can normally be reached M-F 7:30am-3:30pm EST.
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, Omar Fernandez Rivas can be reached at (571) 272-2589. 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.
/COREY M SACKALOSKY/Examiner, Art Unit 2128
/OMAR F FERNANDEZ RIVAS/Supervisory Patent Examiner, Art Unit 2128
Prosecution Insights