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 .
Claims 1-9, 21-24 are presented per 08/18/2026.
Response to Arguments
In response to arguments presented in Remarks 08/18/2026:
The arguments are fully considered, however they are not persuasive.
Per claims 1 and 7, Applicant argues the references of record Kolekar in view of Truong combination fails to teach or suggest at least the limitation that the data storage request include "an indicator indicative of a degree of involvement of generative models in said conveyed information which represents said data” and that Kolekar’s cited fields, i.e. model specifications, analytic specifications, service operations, do not equate the claimed indicator. Applicant invokes the Specification, ¶0154-0160 to support the point that, un the express language of the claim, the indicator must convey information concerning the degree to which generative models are involved in the conveyed information representing the data.
The examiner respectfully disagrees.
Applicant is reminded that the claim language is construed under BRI without importing limitations from the specification in the claim (MPEP 2111, 2111.01(II); In re Van Geuns, 988F.2d 1181, 26 USPQ2d 1057 (Fed Cir. 1993). Claim 1 is constructed with broad functional language, namely a data storage request that includes (i) conveyed information that represents the data whose storage is requested and (ii) an indicator indicative of a degree of involvement of generative models in that convey information. The Remarks relies on the instant Spec, ¶0154-0160 with regard to the characterization of “degree of involvement” as distinguishing full data/partial data supplemented by a generative model, or a still greater generative role. Those modes are not recited in the claim 1. The Specification may illustrate an indicator as an example, however that does not redefine the claim language. The term “degree of involvement” is broad, and it does not indicate specific set of metric measuring or defining the so-called “degree”, nor does it define what “involvement” entails. Given ordinary meaning, such term can simply be service operation of the ML model or the various specifications which describes how extensive and range of functions of the ML model.
Kolekar discloses a data storage request that includes a payload to be stored (ML model/model address, data) and characterizing information include data/analytics specification, service operation, Dataset tag, storage handling information. These characterizing information indicates the role, scope and manner in which the model/data participates in that request. Under BRI, the reference satisfies the claim’s requirements.
Applicant’s distinction of Nadrf_DataManagement_StorageRequest and the Nadrf_MLModelManagement_Request is noted, however not persuasive. Both are storage requests that carry a payload and corresponding characterizing fields. Claim 1 does not require a particular 3GPP operation name, or any particular encoding/parameters structurally defines the ambiguous phrase “degree of involvement”.
Regarding Applicant’s argument that Truong does not teach the indicator as claimed. Applicant argues Truong’s ¶0186 teaches generative models that can produce partial/full synthetic data. ¶0162 teaches a model request that list model types, schema, data set ID etc. That identifies a model or data set, and not indicating the degree of generator model involvement.
The examiner respectfully disagrees. In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986). The rejection does not use Truong as a 102(a)/(b) reference. It uses Truong to show ML models can be generative model and any request associated with such generative model can include any characterizing data. Kolekar already send a storage request with the so-called indicator of ML model involvement. Truong is in a combination to show that the non-limiting ML models of Kolekar can include generative model without any bearing to the operation steps. Truong is not required to reiterate entire claims’ limitation in a 35 USC 103 combination.
Applicant further argues the Office conflates two facts: Kolekar has storage-request fields, and Truong has generative models, asserting these two do not add up to the claimed invention and that the invention must be considered as a whole.
The examiner respectfully disagrees.
The rejection did not stop at those two facts. Kolekar is not cited only for a request that can have fields. In fact, Kolekar discloses the same request claim 1 use, with more clarity structurally at that: a payload representing what to be stored accompanied by characterizing information, and thus receiving an acknowledgement.
This combination is not a merely stacking of GAN onto a field. Truong was not cited only for a fact that GAN exists alone. Instead, Truong is cited to supplement what Kolekar does not explicitly say: the ML model can/may be a generative model, and a request may carry associated data with respective to that model, i.e. type, data schema etc. The rejection did not merely show a generic request one reference and a GAN in another.
Applicant further argues the references don’t teach the indicator even if a GAN is used in Kolekar. The argument seems to assume the indicator must be a unique new field added after substitution of a GAN. The examiner respectfully asserts that there is no basis for that argument. Claim 1 does not require specific transmit/receive acts to operate differently because the model is generative. Using GAN of Truong is a mere change in what the existing field indicate, not a requirement to add a new field.
Applicant further argues the motivation statement for the combination lacks factual support. The examiner respectfully disagrees. The Office does not need an express statement in either reference to add a degree of involvement indicator so nodes can prepare for GAN, per MPEP 2143. The factual basis exist: Kolekar already sends characterizing information with the storage request so the repository knows how to handle it, and Truong sends model type/data set/schema info with a model request so a receiving system knows what model/what associated data are involved for generative models. POSITA would understand the characterizing information in Kolekar exists for the same reasoning as in Truong so as the receiving system know what being received and thus handling, the same rationale in different wordings, not an otherworldly new fact or leap of logics.
With regard to claim 3:
Applicant argues that claim 3 requires the conveyed information to include a partial data set and generative model configured to create a full data set based on that partial set, and an indicator indicative of that arrangement, which the references allegedly do not appear to disclose.
The examiner respectfully disagrees.
Claim 3 is given plain meaning under BRI with no import from the Specification. Truong teaches in ¶0186 a generative model that creates a full data set from input data. In this interpretation, the input data can be equated to a partial dataset and the model is configured as claim. ¶0186 is not relied upon to disclose the entire request. The rejection does not rely on that alone. ¶0162 teaches request information identifies the model an associated dataset. Under BRI, that information is indicative that the conveyed information involved that dataset and that GAN model. The dataset identifier is not required to use claim 3’s exact wording. ¶0060 is regarding additional request meta data about how data is to be used with the model, not the primary mapping. Kolekar already teaches a storage request that can carry data to be stored, or an ML model plus characterizing fields. The combination is that the request carrying Truong’s generative model together with the input/partial data which facilitate the model to generate the output, i.e. full data set, and information identifying that dataset/model pair.
Applicant’s point that Kolekar describes data storage and model storage separately teach away from the including both in the information conveyed to the same repository. Teaching away requires explicit evidences that expressively criticize or discourage the particular combination, and none is found in Kolekar. Applicant’s point that merely substituting a GAN into Kolekar yield only a stored generative model is correct as far as it goes. The combination is not based on that substitution alone. Truong’s model operates on input data to produce a full data set. A POSITA would have reason to store the associated input with it and to identify association in the request, the same reason/purpose Truong’s dataset/model information already serves.
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 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.
Claim(s) 1-9, 21-24 is/are rejected under 35 U.S.C. 103 as being unpatentable over Kolekar - WO 2023/215720 in view of Truong - EP 3591586 (IDS).
As to claim 1:
Kolekar discloses:
An apparatus comprising;
at least one processor; and at least one memory including computer program code; and at least one interface for communication with at least another apparatus, the at least one processor, (Fig 2, Fig. 12, network entities/functions implemented as computer devices/nodes with intrinsic processors/memory storing instruction) with the at least one memory and computer program code, being configured to cause the apparatus to:
transmit a data storage request requesting storage of data, said data storage request including conveyed information and an indicator indicative of a degree of involvement of machine learning models in said conveyed information which represents said data; (Page 4, lines 6-19, 25-28 sending by a consumer entity a storage request, wherein the request includes conveyed information such as data to be stored, an ML model (“Nadrf_DataManagement_StorageRequestcontaining the ML model or ML model address to be stored”), and several indicators of degree of involvement of the ML models, namely data specifications, analytics specifications, service operation, and other analytics information)
and receive a data storage acknowledge response (Page 4, lines 15-17, receiving an acknowledgement response message regarding the storage status)
Except Kolekar does not explicitly have a specific type of the ML model, namely a generative model whose involvement is indicated by the indicator.
Truong, in a related field of management and storage of ML models, in ¶0186, 0162, wherein the ML models include a generative model that may generate partially or fully synthetic data set, and per ¶0162, A model request may also become submitted indicating one or more of the type of model (e.g., neural network), the data schema, the type of training dataset (loan application data), the model task (prediction), and identifier of a dataset used to generate the data. Also ¶0060, “computing resources 101 can be configured to receive at least some training parameters from model optimizer 107 (e.g., batch size, number of training batches..)”
It would have been obvious to one of ordinary skill in the art before the effective filing time of the invention that, for the system of Koleda, the type of ML models are not restrictive and also can include GANs that can provide new and original data for the operators. Furthermore, the request can convey dataset indicator that indicates the scope of dataset involved in the data generation by the GAN, allowing system transparency to ensure network nodes to properly expect and prepare for GAN operations.
As to claim 2:
Kolekar in view of Truong discloses all limitations in claim 1, wherein said indicator is indicative of that said conveyed information includes a full data set, and said conveyed information includes said full data set. (Truong in ¶0162, 0060 where indicator for dataset to indicate which dataset to be involved, which allow the system/operator to flexibly dictate the scope of data used to be used to generate data, which include either all of the datasets or simply as necessary. Thus, one of ordinary skill in the art would contemplate a possibility of full pools of dataset to be stored herewith).
Examiner’s notes: this claim’s patentable weight is questionable at best because it is merely directed to the choice amount of data to be stored in the request, where the overall process of storing and issuing acknowledgement are not dependent on whether the data being full or partial. This is merely a design choice of how much data to be transferred and stored with no impact on the claimed system and process.
As to claim 3:
Kolekar in view of Truong discloses all limitations in claim 1, wherein said indicator is indicative of that said conveyed information includes a partial data set and at least one generative model configured to create a full data set based on said partial data set, and said conveyed information includes said partial data set and said at least one generative model. (Truong in, ¶0186, generates full dataset from inputs. ¶0162, 0060 where indicator for dataset to indicate which dataset to be involved, which allow the system/operator to flexibly dictate the scope of data used to be used to generate data, which include either all of the datasets or simply as necessary. Thus, one of ordinary skill in the art would contemplate a possibility of full pools of dataset to be stored herewith).
Examiner’s notes: similar to claim 2, this claim’s patentable weight is questionable at best because it is merely directed to the choice amount of data to be stored in the request, where the overall process of storing and issuing acknowledgement are not dependent on whether the data being full or partial. This is merely a design choice of how much data to be transferred and stored with no impact on the claimed system and process.
As to claim 4:
Kolekar in view of Truong discloses all limitations in claim 1, wherein said indicator is indicative of that said conveyed information includes at least one generative model configured to create a full data set without reference to a part of said data set, and said conveyed information includes said at least one generative model.
(Kolekar, Page 4, lines 6-19, 25-28 sending by a consumer entity a storage request, wherein the request includes conveyed information such as data to be stored, an ML model (“Nadrf_DataManagement_StorageRequestcontaining the ML model or ML model address to be stored”). Also, Truong, ¶0186, the GAN generate a partially or fully synthetic dataset. For example, a dataset generator may normalize the input dataset, and generate a synthetic training dataset while preserving the data structure. The data synthesis model may be a machine learning model (e.g., a GAN or recurrent neural network (RNN) model). At step 1806, process 1800 may train the data synthesis model based on a similarity metric value using the normalized input dataset and the synthetic dataset.)
As to claim 5:
Kolekar in view of Truong discloses all limitations in claim 3, wherein said conveyed information includes a generative model description specifying utilization of said at least one generative model. (Kolekar, Page 4, lines 6-19, 25-28 sending by a consumer entity a storage request, wherein the request includes conveyed information such as data to be stored, an ML model (“Nadrf_DataManagement_StorageRequestcontaining the ML model or ML model address to be stored”), and several indicators of degree of involvement of the ML models, namely data specifications, analytics specifications, service operation, and other analytics information))
As to claim 6:
Kolekar in view of Truong discloses all limitations in claim 1, wherein the apparatus is further caused to: create said data utilizing said at least one generative model. (See Truong, ¶0186, At step 1806, process 1800 may train the data synthesis model based on a similarity metric value using the normalized input dataset and the synthetic dataset)
As to claim 7:
Kolekar discloses:
An apparatus comprising;
at least one processor; and at least one memory including computer program code; and at least one interface for communication with at least another apparatus, the at least one processor, with the at least one memory and computer program code, (Fig 2, Fig. 12, network entities/functions having communication interfaces to couple to each other and implemented as computer devices/nodes with intrinsic processors/memory storing instruction) being configured to cause the apparatus to:
receive a data storage request requesting storage of data, said data storage request including conveyed information and an indicator indicative of a degree of involvement of machine learning models in said conveyed information which represents said data; (Page 4, lines 6-19, 25-28 receiving from a consumer entity a storage request, wherein the request includes conveyed information such as data to be stored, one or more ML models (“Nadrf_DataManagement_StorageRequestcontaining the ML model or ML model address to be stored”), and several indicators of degree of involvement of the ML models, namely data specifications, analytics specifications, service operation, and other analytics information)
store said conveyed information; and transmit a data storage acknowledge response. (Page 4, lines 15-17, transmitting an acknowledgement response message regarding the storage status)
As to claim 8:
Kolekar in view of Truong discloses all limitations in claim 7, wherein said indicator is indicative of that said conveyed information includes a full data set, and said conveyed information includes said full data set. (Truong in ¶0162, 0060 where indicator for dataset to indicate which dataset to be involved, which allow the system/operator to flexibly dictate the scope of data used to be used to generate data, which include either all of the datasets or simply as necessary. Thus, one of ordinary skill in the art would contemplate a possibility of full pools of dataset to be stored herewith).
Examiner’s notes: this claim’s patentable weight is questionable at best because it is merely directed to the choice amount of data to be stored in the request, where the overall process of storing and issuing acknowledgement are not dependent on whether the data being full or partial. This is merely a design choice of how much data to be transferred and stored with no impact on the claimed system and process.
As to claim 9:
Kolekar in view of Truong discloses all limitations in claim 7, wherein said indicator is indicative of that said conveyed information includes a partial data set (Truong in, ¶0186, generates full dataset from inputs. ¶0162, 0060 where indicator for dataset to indicate which dataset to be involved, which allow the system/operator to flexibly dictate the scope of data used to be used to generate data, which include either all of the datasets or simply as necessary. Thus, one of ordinary skill in the art would contemplate a possibility of full pools of dataset to be stored herewith), and at least one generative model configured to create a full data set based on said partial data set, and said conveyed information includes said partial data set and said at least one generative model. (Kolekar, Page 4, lines 6-19, 25-28 sending by a consumer entity a storage request, wherein the request includes conveyed information such as data to be stored, an ML model (“Nadrf_DataManagement_StorageRequestcontaining the ML model or ML model address to be stored”). Also, Truong, ¶0186, the GAN generate a partially or fully synthetic dataset. For example, a dataset generator may normalize the input dataset, and generate a synthetic training dataset while preserving the data structure. The data synthesis model may be a machine learning model (e.g., a GAN or recurrent neural network (RNN) model). At step 1806, process 1800 may train the data synthesis model based on a similarity metric value using the normalized input dataset and the synthetic dataset.)
As to claim 21:
Kolekar in view of Truong discloses all limitations in claim 1, wherein the at least another apparatus implements an analytics data repository function (ADRF). (See Kolekar, Fig. 3, and associated text, “The ADRF 1066 exposes Nadrf services (e.g., via an Nadrf interface) for storage and retrieval of data by other NFs (…) data may be stored in the ADRF 1066 by a consumer sending the ADRF 1066 an Nadrf DataManagement StorageRequest containing the data or analytics to be stored. In some examples, the Nadrf DataManagement StorageRequest sent by a service consumer can include the data to be stored, data collection timestamp(s), analytics with timestamp, service operation, analytics specification or data specification, storage handling information, DataSetTag, and/or other suitable information”)
As to claim 22:
Kolekar in view of Truong discloses all limitations in claim 21, wherein the data storage request comprises an Nadrf_DataManagement_StorageRequest service operation. (page 4 of Kolekar, “data may be stored in the ADRF 1066 by a consumer sending the ADRF 1066 an Nadrf DataManagement StorageRequest containing the data or analytics to be stored. In some examples, the Nadrf DataManagement StorageRequest sent by a service consumer can include the data to be stored, data collection timestamp(s), analytics with timestamp, service operation, analytics specification or data specification, storage handling information, DataSetTag, and/or other suitable information”, and similarly for model, “ML models may be stored in the ADRF 1066 by a consumer sending the ADRF 1066 an Nadrf MLModelManagement StorageRequest”)
As to claim 23:
Kolekar in view of Truong discloses all limitations in claim 22, wherein said indicator is indicative that said conveyed information includes a partial data set and at least one generative model, and said conveyed information includes said partial data set and said at least one generative model.
(Claim 23 depends on claim 22 require the Nadrf DataManagement StorageRequest structure. Kolekar, in at least page 4, Fig. 3, discloses the very identical service operation to an ADRF with the same Nadrf_DataManagement_StorageRequest, or similarly Nadrf MLModelManagement StorageRequest, carrying data to be stored and characterizing information (data/analytics specification, service operation, DatasetTag, timestamp), and an acknowledgement. Kolekar further dsiclsoes storing an ML model at the ADRF. Claim 22 recites that information, while claim 23 does not recite any new information element or any structure to that message. Under BRI, claim 23 is a limitation of the contents of the message of Kolekar. Regarding said indicator is indicative that said conveyed information includes a partial data set and at least one generative model, Kolekar’s in page 4 already discloses characterizing fields as stated above. Those fields indicate that the conveyed information includes the associated data and the generative model. Regarding said conveyed information includes said partial data set and said at least one generative model: Truong per ¶0186, the GAN generates a data set from input data, i.e. “partial data set”. Kolekar’s request message already carries data associated with the ML model. In combination that input is in the conveyed information with the model, both in one payload.)
As to claim 24:
Kolekar in view of Truong discloses all limitations in claim 23, wherein said at least one generative model is configured to create a full data set based on said partial data set (See ¶0186, a GAN generates from a set of input data (i.e. partial data set) an output, i.e. fully synthetic data set)
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
WO 2021/136601 - a network entity being configured to receive a model request from a requesting entity over the communication network, the model request requesting the at least one model parameter of the model, to obtain the requested at least one model parameter by at least one of the following: executing a machine-learning model training algorithm, the machine-learning model training algorithm being configured to train the model with input data in order to determine at least one of the requested model parameter; searching a local data base for an existing model; or requesting the at least one model parameter from a further network entity; and to send the requested model parameter towards the requesting entity over the communication network..
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 replying 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 QUAN M HUA whose telephone number is (571)270-7232. The examiner can normally be reached 10:30-6:30.
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, Anthony Addy can be reached at 571-272-7795. 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.
/QUAN M HUA/Primary Examiner, Art Unit 2645
Prosecution Insights