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 .
Continued Examination Under 37 CFR 1.114
A request for continued examination under 37 CFR 1.114, including the fee set forth in 37 CFR 1.17(e), was filed in this application after final rejection. Since this application is eligible for continued examination under 37 CFR 1.114, and the fee set forth in 37 CFR 1.17(e) has been timely paid, the finality of the previous Office action has been withdrawn pursuant to 37 CFR 1.114. Applicant's submission filed on 02/26/2026 has been entered. Claims 1-20 are pending. Claims 1, 9 and 14 are currently amended.
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.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over Yavuz et al. (Yavuz), US Patent Application Publication No. US 2021/0375269 A1, in view of Li et al. (Li), US Patent Application Publication No. US 2022/0084204 A1, and further in view of Truong et al. (Truong), US Patent Application Publication No. US 2021/0150332 A1.
As to independent claim 1, Yavuz discloses a computer-implemented method to enable improved learning with few training examples, the method comprising:
obtaining, by a computing system comprising one or more computing devices, a set of unlabeled training data associated with a target task, the set of unlabeled training data comprising a plurality of unlabeled training examples that are in-domain for the target task (Figures 1, 2B and paragraph [0025]: unlabeled dialogues (unlabeled training data) in target domain (in-domain for target task); Figure 2B provides an example data segment of the dialogue data 120 in the target domain of flight bookings, wherein the dialogue 120 in the target domain includes a plurality of dialogue turns 211a-d);
accessing, by the computing system, a first machine-learned model that has been previously trained using a set of labeled training data associated with a pre-training task that is different than the target task, the set of labeled training data comprising a plurality of labeled training examples that are out-of-domain for the target task (Figures 1, 2A and paragraph [0024]: labeled dialogues (labeled training data) in source domain (out-of-domain for the target), and using a pre-trained language model 150 (a first machine-learned model) for dialogue act tagging tasks, wherein the pre-trained language model 150 may be trained with labeled dialogues (labeled training data) in a source domain 110 (out-of-domain for the target task); Figure 2A provides an example data segment of the labeled dialogue in the source domain, which may include multiple dialogue turns 201a-d (plurality of labeled training examples));
processing, by the computing system, each unlabeled training example with the first machine-learned model to respectively generate a synthetic supplement for each unlabeled training example, the plurality of training examples and synthetic supplements forming a set of synthetic training data (Figure 1 and paragraph [0026]: to adapt the pre-trained language model 150 to the target domain, utilizing the pre the pre-trained language model 150 (the first machine-learned model) with the pre-trained parameters 153 to implement mask augmentation (synthetic supplement) of the unlabeled dialogue data in the target domain); and
training, by the computing system, a second, different machine-learned model using the set of synthetic training data (Figure 1 and paragraphs [0026]-[0028]: training the language model 155 (a second, different machine-learned model), which loaded with pretrained parameters 153 from pre-trained language model 150 (first machine-learned model), is trained with the mask augmented data (synthetic supplement). The combination of pre-trained parameters 153 and mask augmented data is considered as synthetic training data.
Yavuz, however, does not disclose wherein the synthetic supplement supplements the plurality of unlabeled training examples, and the plurality of the unlabeled training examples and the synthetic supplement form a set of synthetic training data.
In the same field of endeavor, Li discloses apparatuses, systems, and techniques to generate labels for images using generative adversarial networks (Abstract). Li further discloses receiving in put image 305 (unlabeled training example), using a generative adversarial network (GAN) (first machine-learned model) to generate synthetic version of input image and generate one or more labels (synthetic supplement) corresponding to one or more objects in synthetic version of input image (Figures 3A-3B and paragraphs [0072]-[0087]). Li further discloses in Figure 6 and paragraphs [0076] and [0103]-[0105] that discriminator network B (2nd machine-learned model) receives synthetic image 622 and labels of synthetic image for solving classification problem.
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the system of Yavuz to incorporate “wherein the synthetic supplement supplements the plurality of unlabeled training examples, and the plurality of the unlabeled training examples and the synthetic supplement form a set of synthetic training data”, which taught by Li for the purpose of evaluating/training data.
Yavus and Li, however does not disclose the synthetic supplement supplements the plurality of unlabeled training examples with a synthetic string of tokens.
In the same field of endeavor, Truong discloses providing systems and methods for synthetic data generation, wherein a recurrent neural network can be trained for synthetic data generation by obtaining a sequence of elements and determining, using a classifier, that the sequence corresponds to a token (Abstract). Truong further discloses detecting, in the generated synthetic data, a sequence of elements matching a string represented by one of the tokens, and the synthetic data can be updated by replacing the sequence elements with the one of the tokens (paragraph [0011]). Truong further discloses in Figure 1 and paragraph [0026] generating synthetic data including tokens corresponding to strings, and synthetic data can be stored or consumed by one or more other system (paragraph [0040]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the systems of Yavus and Li to include synthetic data with a synthetic string of tokens, as taught by Truong, for the purpose of updating the generated synthetic data by replacing the tokens with the elements (Truong, paragraphs [0010]-[0011]).
As to dependent claim 2, Yavuz discloses wherein: the set of labeled training data comprises a plurality of labeled natural language inference training examples, each labeled natural language inference training example comprising a first string of tokens, a second string of tokens, and a label that describes a relationship between the first string of tokens and the second string of tokens (Figure 7 and paragraphs [0063], [0064]); and
the first machine-learned model comprises a generative language model that has been trained to process the first string of tokens and the label to predict the second string of tokens (Figure 7 and paragraphs [0065]-[0067]).
As to dependent claim 3, Yavuz discloses wherein: each unlabeled training example in the set of unlabeled training data comprises an unlabeled string of tokens (paragraph [0026]); and
processing, by the computing system, each unlabeled training example with the first machine-learned model to respectively generate a synthetic supplement for each unlabeled training example comprises processing, by the computing system, each unlabeled string of tokens and a supplied label to generate a synthetic string of tokens (paragraph [0026]).
As to dependent claim 4, Yavuz discloses wherein processing, by the computing system, each unlabeled string of tokens and the supplied label to generate the synthetic string of tokens comprises processing, by the computing system, each unlabeled string of tokens and a plurality of different supplied labels to generate a plurality of different synthetic strings of tokens for each unlabeled string of tokens (paragraphs [0026]-[0028]).
As to dependent claim 5, Yavuz discloses further comprising:
using, by the computing system, a third machine-learned model to filter the plurality of different synthetic strings of tokens (paragraphs [0026], [0034]-[0035]).
As to dependent claim 6, Yavuz discloses wherein using, by the computing system, the third machine-learned model to filter the plurality of different synthetic strings of tokens comprises:
for each pair of unlabeled string of tokens and synthetic string of tokens (Figure 5 and paragraph [0049]):
processing, by the computing system, the pair of unlabeled string of tokens and synthetic string of tokens with the third machine-learned model to generate a predicted label (Figure 5 and paragraph [0049]); and
determining, by the computing system, whether the predicted label matches the supplied label that was supplied to generate the synthetic string of tokens (Figure 5 and paragraph [0049]).
As to dependent claim 7, Yavuz discloses wherein using, by the computing system, the third machine-learned model to filter the plurality of different synthetic strings of tokens further comprises, for each pair of unlabeled string of tokens and synthetic string of tokens and when the predicted label matches the supplied label:
determining, by the computing system, whether a confidence value output by the third machine-learned model for the predicted label satisfies a threshold value (paragraphs [0071]-[0077]);
when the confidence value output by the third machine-learned model for the predicted label satisfies the threshold value: maintaining, by the computing system, the pair of unlabeled string of tokens and synthetic string of tokens in the set of synthetic training data (paragraphs [0071]-[0077]); and
when the confidence value output by the third machine-learned model for the predicted label does not satisfy the threshold value: discarding, by the computing system, the pair of unlabeled string of tokens and synthetic string of tokens from the set of synthetic training data (paragraphs [0071]-[0077]).
As to dependent claim 8, Yavuz discloses after training, by the computing system, the second machine-learned model using the set of synthetic training data: training, by the computing system, the second machine-learned model using a second set of labeled training data associated with the target task, the second set of labeled training data comprising a second plurality of labeled training examples that are in-domain for the target task (paragraph [0024]).
As to independent claim 9, Yavuz discloses a computing system configured to perform improved learning with few training examples, the computing system comprising:
one or more processors (Figure 3, paragraph [0030]: computing device 300 includes a processor); and
one or more non-transitory computer-readable media that collectively store instructions that when executed by the one or more processors cause the computing system to perform operations (Figure 3 and paragraph [0033]: memory may include non-transitory, tangible, machine readable media that includes executable code that when run by one or more processor may cause the one or more processors to perform the methods), the operations comprising:
for each of a plurality of training iterations:
accessing a current set of labeled training data associated with a target task, the current set of labeled training data comprising labeled training examples that are in-domain for the target task (Figures 1, 2A and paragraph [0024]: labeled dialogues (labeled training data) in source domain (in-domain for the target), and using a pre-trained language model 150 for dialogue act tagging tasks (target task), wherein the pre-trained language model 150 may be trained with labeled dialogues (labeled training data) in a source domain 110 (in-domain); Figure 2A provides an example data segment of the labeled dialogue in the source domain, which may include multiple dialogue turns 201a-d (labeled training examples));
training a base model using the current set of labeled training data to generate a current student model (paragraphs [0034]-[0035]: language model 335 may be a pre-trained MASK token language model, which may be trained by one or more of a supervised tagging loss (STL) module 331, a masked tagging loss (MTL) module 332, a masked language model loss (MLM) module 333, and/or a disagreement loss module (DAL) 334, and the DAL module utilizes an unsupervised teacher-student training mechanism to control the level and kind of discrete perturbations to achieve augmentation of the input text 340 [0039]-[0041] and Figure 4A: input dialogue history 340 is labeled and converted into a sequence of words which is input into a pre-trained language model (e.g., BERT) 335);
accessing a set of unlabeled training data associated with the target task, the set of unlabeled training data comprising unlabeled training examples that are in-domain for the target task (Figures 1, 2B and paragraph [0025]: unlabeled dialogues (unlabeled training data) in target domain (in-domain for target task); Figure 2B provides an example data segment of the dialogue data 120 in the target domain of flight bookings, wherein the dialogue 120 in the target domain includes a plurality of dialogue turns 211a-d);
processing each unlabeled training data with the current student model to respectively generate a synthetic label for each unlabeled training example, the unlabeled training examples and synthetic labels forming a set of self-labeled training data (paragraphs [0051]-[0053] and Figure 6: a classification distribution tags is generated using the pre-trained model for the generated input representation, for example, the representation of the dialogue data is used as the input to the pre-trained language model and the model computes a probability vector indicating a conditional probability of each specific tag, given the input dialogue history); and
combining some or all of the set of self-labeled training data with an original set of labeled training data to generate the current set of labeled training data for a next training iteration of the plurality of training iterations (Figure 3 and paragraph [0033]: the dialogue tagging module 330 may be used to receive and handle the input of a dialogue history and generate an output of dialogue tags 350, which may appear in the form of classification distributions of different tags. The dialogue act tagging module 330 may also handle the iterative training and/or evaluation of a system or model used for dialogue act tagging); and
after the plurality of training iterations, outputting the current student model as a
output model (Figure 4D and paragraphs [0047]-[0048]: updating the student language model 335b via backpropagation 445b).
Yavuz, however, does not disclose wherein the synthetic supplement supplements the plurality of unlabeled training examples, and the plurality of the unlabeled training examples and the synthetic supplement form a set of synthetic training data.
In the same field of endeavor, Li discloses apparatuses, systems, and techniques to generate labels for images using generative adversarial networks (Abstract). Li further discloses receiving in put image 305 (unlabeled training example), using a generative adversarial network (GAN) (first machine-learned model) to generate synthetic version of input image and generate one or more labels (synthetic supplement) corresponding to one or more objects in synthetic version of input image (Figures 3A-3B and paragraphs [0072]-[0087]). Li further discloses in Figure 6 and paragraphs [0076] and [0103]-[0105] that discriminator network B (2nd machine-learned model) receives synthetic image 622 and labels of synthetic image for solving classification problem.
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the system of Yavuz to incorporate “wherein the synthetic supplement supplements the plurality of unlabeled training examples, and the plurality of the unlabeled training examples and the synthetic supplement form a set of synthetic training data”, which taught by Li for the purpose of evaluating/training data.
Yavus and Li, however does not disclose the synthetic supplement supplements the plurality of unlabeled training examples with a synthetic string of tokens.
In the same field of endeavor, Truong discloses providing systems and methods for synthetic data generation, wherein a recurrent neural network can be trained for synthetic data generation by obtaining a sequence of elements and determining, using a classifier, that the sequence corresponds to a token (Abstract). Truong further discloses detecting, in the generated synthetic data, a sequence of elements matching a string represented by one of the tokens, and the synthetic data can be updated by replacing the sequence elements with the one of the tokens (paragraph [0011]). Truong further discloses in Figure 1 and paragraph [0026] generating synthetic data including tokens corresponding to strings, and synthetic data can be stored or consumed by one or more other system (paragraph [0040]).
It would have been obvious to one of ordinary skill in the art, before the effective filing date of the claimed invention, to modify the systems of Yavus and Li to include synthetic data with a synthetic string of tokens, as taught by Truong, for the purpose of updating the generated synthetic data by replacing the tokens with the elements (Truong, paragraphs [0010]-[0011]).
As to dependent claim 10, Yavuz discloses wherein the base model comprises a self-trained model.
As to dependent claim 11, Yavuz discloses wherein the same base model is used at each of the plurality of training iterations (paragraph [0033]).
As to dependent claim 12, Yavuz discloses wherein combining some or all of the set of self- labeled training data with the original set of labeled training data to generate the current set of labeled training data for the next training iteration comprises combining all of the set of self- labeled training data with the original set of labeled training data to generate the current set of labeled training data for the next training iteration (Figure 1 and paragraphs [0026]-[0028]).
As to dependent claim 13, Yavuz discloses wherein:
the base model comprises a base language model (Figure 3); and
the target task comprises a natural language processing task (paragraph [0040]).
Claims 14-19 and 20 are system claims that contains similar limitations of claims 1-6 and 8, respectively. Therefore, claims 14-19 and 20 are rejected under the same rationale.
Response to Arguments
Applicant’s argument regarding the rejection under 35 U.S.C. 101 (see pages 9-13 of Remarks) is persuasive. Therefore, the rejection of claims 1-20 under 35 U.S.C. 101 is hereby withdrawn.
Applicant’s arguments and amendments filed on 02/26/2026 have been fully considered but they are not deemed fully persuasive. Applicant’s arguments with respect to claims 1-20 have been considered but are moot in view of the new ground(s) of rejection as explained here below, necessitated by Applicant’s substantial amendment (i.e., the synthetic supplement supplements the plurality of unlabeled training examples with a synthetic string of tokens) to the claims which significantly affected the scope thereof. Please see the new ground of the rejection with additional prior art Truong.
Conclusion
Any inquiry concerning this communication should be directed to CHAU T NGUYEN at telephone number (571)272-4092. The examiner can normally be reached on M-F from 8am to 5pm (PT).
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) Form at https://www.uspto.gov/patents/uspto-automated-interview-request-air-form.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Cesar Paula, can be reached at telephone number 5712724128. 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 and the Private Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from Patent Center or Private PAIR. Status information for unpublished applications is available through Patent Center and Private PAIR for authorized users only. Should you have questions about access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
/CHAU T NGUYEN/Primary Examiner, Art Unit 2145