DETAILED ACTION
Notice of Pre-AIA or AIA Status
1. The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
2. The amendment of claims 1-21, submitted on 2/23/2026, is acknowledged and considered. Claims 1-12 and 14-21 are pending.
Claims 1, 14, and 20 are independent claims.
Claims 13 is cancelled by Applicant.
Priority
3. The present application has relationship to:
KR10-2022-0011670, filing date 1/26/2022
Response to Arguments
4. Applicant’s arguments with respect to claim(s) 1-21 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Claims 15-18 has overcome the objection, necessitated by the current amendment.
Necessitated by the amendment, Applicant’s arguments are moot as the claims are now rejected under Angel and Becker combination.
Claim Objections
5. Claim 4 is objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
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.
6. Claim(s) 1-3, 5-12, and 14-21 is/are rejected under 35 U.S.C. 103 as being unpatentable over Angel, et al. [US 20210174243] in view of Becker, et al. [US 20180337899].
As per claim 1: Angel, et al. teaches a method comprising:
generating a public key and a secret key, based on a master key and a parameter vector indicating parameters of a first layer of a machine learning model; [Angel: para 0027; Machine learning systems are becoming ubiquitous (e.g., recommendation systems, diagnostic/prediction systems, autopilot systems, pattern recognition systems, and so on), and their performance (e.g., optimization of parameters, weights, biases, and so on) can rely on large volumes of training data See also para0041; the key distribution component deliver a common public key or party-specific public key to the participants, based on the given master public/secret keys. The key distribution component take as input the master public/secret keys and a vector y generated by the coordinator (e.g., participant-related weight vector w.sub.p, sample-related weight vector u, and so on)]
generating encrypted data by encrypting, based on the public key, an input vector indicating input data of the machine learning model; [Angel: para 0040; functional encryption for inner-product (FEIP) schemes can be implemented, which can output the inner-product between ciphertext of one or more input vectors and predetermined vector without revealing the plaintext of the one or more input vectors]
generating decrypted data corresponding to an approximation value [Angel: para 0088; two kinds of aggregation computation could be linear or non-linear due to different underlying machine learning algorithms, such as linear classification models and/or non-linear logistic regression and/or tree ensemble models. An approximation approach where non-linear computations are transferred to linear computations by applying Taylor approximation] of an inner product of the parameter vector and the input vector by decrypting the encrypted data based on the secret key [Angel: para 0039-0040; FE techniques belong to a public-key encryption family where the decryptor can be issued a secret key (e.g., a functionally derived key) that allows the decryptor to learn the result of a function over a ciphertext without learning the corresponding plaintext. Without revealing x, where D.sub.sk represents decryption by the functional secret key, E.sub.pk represents encryption by the public key, and ƒ represents the desired output function. Functional encryption for inner-product (FEIP) schemes can be implemented, which can output the inner-product between ciphertext of one or more input vectors and some other given, desired, and/or predetermined vector without revealing the plaintext of the one or more input vectors (e.g., ƒ above can be an inner-product function)], **wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product; and [**rejected under a secondary reference, discussion below]
controlling the machine learning model based on the decrypted data to obtain an output value. [Angel: para 0041; can generate as output an MIFE and/or SIFE function derived key that computes an inner-product between input ciphertext and the vector y generated by the coordinator. The coordinator can accordingly apply the function derived key on data x or set of data {x.sub.i}received from the participants to yield ƒ.sub.SIFE(x. y) or ƒ.sub.MIFE({x.sub.i}, y).]
Angel teaches FE techniques belong to a public-key encryption family where the decryptor can be issued a secret key (e.g., a functionally derived key) that allows the decryptor to learn the result of a function over a ciphertext without learning the corresponding plaintext. Without revealing x, where D.sub.sk represents decryption by the functional secret key, E.sub.pk represents encryption by the public key, and ƒ represents the desired output function. Functional encryption for inner-product (FEIP) schemes can be implemented, which can output the inner-product between ciphertext of one or more input vectors and some other given, desired, and/or predetermined vector without revealing the plaintext of the one or more input vectors (e.g., ƒ above can be an inner-product function). Both single-input functional encryption (SIFE) and multi-input functional encryption (MIFE) techniques can be used to generate inner-product output [0039-0040]. However, Angel did not clearly teach “wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product”.
Becker teaches a modified version of a Learning with Errors (LWE) asymmetric cryptographic system. LWE is a form of lattice-based cryptography that uses a mathematical lattice as the basis for cryptographic primitives that produce a public key and a private key in an asymmetric cryptographic system. More broadly, a reference to LWE cryptography refers to a class of cryptographic systems that rely on the underlying lattice mathematics problems known to the art, such as the shortest vector problem (SVP), to implement an asymmetric cryptographic system with public and private keys. FIG. 9 depicts a simplified prior-art key generation process that produces a public/private key pair that can be used with an LWE cryptographic system. In FIG. 9, the LWE public key is a two-part key that includes a matrix A and a vector b, which are both made publicly available [Becker: para 0028-0029]. Becker discloses the client uses the inner public key to generate an encrypted representation of plaintext data with added noise. The client generates an error vector that is indistinguishable from a random Gaussian noise distribution based on the encrypted data generated using the inner public key. The client uses the partial outer public key, which is equivalent to the matrix A shown above, with the individual secret key and the specially generated error vector to produce an output vector b that forms the rest of a public key that is then transmitted to the untrusted aggregator. Thus, the client generates the vector b that is part of an LWE public key where the vector b also contains an encrypted representation of the plaintext data in the error vector data e [Becker: para 0037]. As such, Angel obviously suggest “wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product”, one would be motivated to prevent attack where the private key is not revealed to the public and is only accessible by trusted devices in the LWE cryptographic system.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Becker with Angel to teach “wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product” for the reason to prevent attack by not revealing the private key to the public and is only accessible by trusted devices in the LWE cryptographic system [Becker: para 0030].
Claim 2: Angel: para 0010, 0041; discussing the method of claim 1, wherein the decrypted data comprise node output values corresponding to nodes of the first layer, and wherein the controlling the machine learning model comprises applying the node output values as input to a second layer of the machine learning model.
Claim 3: Angel: para 0039, 0041 [public/secret keys and a vector]; discussing the method of claim 1, wherein the parameter vector comprises a plurality of node parameter vectors respectively corresponding to a plurality of nodes in the first layer, wherein the secret key comprises a plurality of secret key vectors respectively corresponding to the plurality of nodes in the first layer, wherein the public key comprises a plurality of public key vectors respectively corresponding to the plurality of nodes in the first layer, wherein the encrypted data comprise a plurality of node encrypted data respectively corresponding to the plurality of nodes in the first layer, and wherein the decrypted data comprise a plurality of node output values respectively corresponding to the plurality of nodes in the first layer. [Angel: para 0038; decryption with keys of the participants which may be nodes]
Claim 4: Objected
Claim 5: Angel: para 0027, 0066 [generate (e.g., via type-1 NSA decryption 302) the participant-related weight vector, verify via the inference]; discussing the method of claim 1, further comprising: training the machine learning model based on the input data corresponding to training data; and performing inference of the machine learning model based on the decrypted data corresponding to inference data.
Claim 6: Angel: para 0027, 0034 [Machine learning systems pattern and performance (e.g., optimization of parameters, weights, biases, and so on) rely on large volumes of training data]; discussing the method of claim 5, wherein the parameter vector used in the inference of the machine learning model comprises optimized weight values and optimized bias values of the first layer that are determined based on results of the training after the training is completed.
Claim 7: Angel: para 0080; discussing the method of claim 1, further comprising: training the machine learning model based on first decrypted data corresponding to training data; and performing inference of the machine learning model based on second decrypted data corresponding to inference data.
Claim 8: Angel: para 0027 [Machine learning systems pattern and performance (e.g., optimization of parameters, weights, biases, and so on) rely on large volumes of training data]; discussing the method of claim 7, wherein the parameter vector used in the training comprises initial weight values and initial bias values of the first layer.
Claim 9: Angel: para 0027 [Machine learning systems pattern and performance (e.g., optimization of parameters, weights, biases, and so on) rely on large volumes of training data]; discussing the method of claim 8, wherein weight values and bias values of the first layer are fixed to the initial weight values and the initial bias values during the training.
Claim 10: Angel: para 0030; discussing the method of claim 9, wherein the parameter vector used in the inference of the machine learning model corresponds to the parameter vector used in the training.
Claim 11: Angel: para 0066; discussing the method of claim 8, wherein weight values and bias values of the first layer are updated during the training.
Claim 12: Angel: para 0027, 0034; discussing the method of claim 11, wherein the parameter vector used in the inference of the machine learning model comprises optimized weight values and optimized bias values of the first layer that are determined based on results of the training after the training is completed.
Claim 13: Cancelled
As per claim 14: Angel, et al. teaches a system comprising;
a storage device comprising a storage controller and a nonvolatile memory device, the storage controller comprising a processor configured to control a machine learning model; and [Angel: para 0005]
a host device configured to control the storage device, wherein the system is configured to: [Angel: para 0121]
generate a public key and a secret key, based on a master key and a parameter vector indicating parameters of a first layer of the machine learning model; [Angel: para 0027; Machine learning systems are becoming ubiquitous (e.g., recommendation systems, diagnostic/prediction systems, autopilot systems, pattern recognition systems, and so on), and their performance (e.g., optimization of parameters, weights, biases, and so on) can rely on large volumes of training data See also para0041; the key distribution component deliver a common public key or party-specific public key to the participants, based on the given master public/secret keys. The key distribution component take as input the master public/secret keys and a vector y generated by the coordinator (e.g., participant-related weight vector w.sub.p, sample-related weight vector u, and so on)]
generate encrypted data by encrypting, based on the public key, an input vector indicating input data of the machine learning model; [Angel: para 0040; functional encryption for inner-product (FEIP) schemes can be implemented, which can output the inner-product between ciphertext of one or more input vectors and predetermined vector without revealing the plaintext of the one or more input vectors]
generate decrypted data corresponding to an approximation value [Angel: para 0088; two kinds of aggregation computation could be linear or non-linear due to different underlying machine learning algorithms, such as linear classification models and/or non-linear logistic regression and/or tree ensemble models. An approximation approach where non-linear computations are transferred to linear computations by applying Taylor approximation] of an inner product of the parameter vector and the input vector by decrypting the encrypted data based on the secret key [Angel: para 0039-0040; FE techniques belong to a public-key encryption family where the decryptor can be issued a secret key (e.g., a functionally derived key) that allows the decryptor to learn the result of a function over a ciphertext without learning the corresponding plaintext. Without revealing x, where D.sub.sk represents decryption by the functional secret key, E.sub.pk represents encryption by the public key, and ƒ represents the desired output function. Functional encryption for inner-product (FEIP) schemes can be implemented, which can output the inner-product between ciphertext of one or more input vectors and some other given, desired, and/or predetermined vector without revealing the plaintext of the one or more input vectors (e.g., ƒ above can be an inner-product function)], **wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product; and [**rejected under a secondary reference, discussion below]
control the machine learning model based on the decrypted data to obtain an output value. [Angel: para 0041; can generate as output an MIFE and/or SIFE function derived key that computes an inner-product between input ciphertext and the vector y generated by the coordinator. The coordinator can accordingly apply the function derived key on data x or set of data {x.sub.i}received from the participants to yield ƒ.sub.SIFE(x. y) or ƒ.sub.MIFE({x.sub.i}, y).]
Angel teaches FE techniques belong to a public-key encryption family where the decryptor can be issued a secret key (e.g., a functionally derived key) that allows the decryptor to learn the result of a function over a ciphertext without learning the corresponding plaintext. Without revealing x, where D.sub.sk represents decryption by the functional secret key, E.sub.pk represents encryption by the public key, and ƒ represents the desired output function. Functional encryption for inner-product (FEIP) schemes can be implemented, which can output the inner-product between ciphertext of one or more input vectors and some other given, desired, and/or predetermined vector without revealing the plaintext of the one or more input vectors (e.g., ƒ above can be an inner-product function). Both single-input functional encryption (SIFE) and multi-input functional encryption (MIFE) techniques can be used to generate inner-product output [0039-0040]. However, Angel did not clearly teach “wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product”.
Becker teaches a modified version of a Learning with Errors (LWE) asymmetric cryptographic system. LWE is a form of lattice-based cryptography that uses a mathematical lattice as the basis for cryptographic primitives that produce a public key and a private key in an asymmetric cryptographic system. More broadly, a reference to LWE cryptography refers to a class of cryptographic systems that rely on the underlying lattice mathematics problems known to the art, such as the shortest vector problem (SVP), to implement an asymmetric cryptographic system with public and private keys. FIG. 9 depicts a simplified prior-art key generation process that produces a public/private key pair that can be used with an LWE cryptographic system. In FIG. 9, the LWE public key is a two-part key that includes a matrix A and a vector b, which are both made publicly available [Becker: para 0028-0029]. Becker discloses the client uses the inner public key to generate an encrypted representation of plaintext data with added noise. The client generates an error vector that is indistinguishable from a random Gaussian noise distribution based on the encrypted data generated using the inner public key. The client uses the partial outer public key, which is equivalent to the matrix A shown above, with the individual secret key and the specially generated error vector to produce an output vector b that forms the rest of a public key that is then transmitted to the untrusted aggregator. Thus, the client generates the vector b that is part of an LWE public key where the vector b also contains an encrypted representation of the plaintext data in the error vector data e [Becker: para 0037]. As such, Angel obviously suggest “wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product”, one would be motivated to prevent attack where the private key is not revealed to the public and is only accessible by trusted devices in the LWE cryptographic system.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Becker with Angel to teach “wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product” for the reason to prevent attack by not revealing the private key to the public and is only accessible by trusted devices in the LWE cryptographic system [Becker: para 0030].
Claim 15: Angel: para 0029-0031, 0085 [key generation and encryption]; discussing the system of claim 14, wherein the processor is further configured to control: a key generator to generate the public key and the secret key; an encryptor to receive the input data from the host device, receive the public key from the key generator, and generate the encrypted data by encrypting the input data; and a decryptor to receive the secret key from the key generator, and generate the decrypted data by decrypting the encrypted data. [Angel: para 0039; decryptor]
Claim 16: Angel: para 0029-0031, 0085 [key generation and encryption include input data]; discussing the system of claim 14, wherein the host device comprises a host processor configured to control a key generator to generate the public key and the secret key, and wherein the processor of the storage controller is further configured to control: an encryptor to receive the public key and the input data from the host device, and generate the encrypted data by encrypting the input data; and a decryptor to receive the secret key from the host device, and generate the decrypted data by decrypting the encrypted data. [Angel: para 0038-0039; decryptor]
Claim 17: Angel: para 0038-0039, 0084; discussing the system of claim 14, wherein the host device comprises a host processor configured to control an encryptor to receive the public key from the storage device, and generate the encrypted data by encrypting the input data, and wherein the processor of the storage controller is further configured to control: a key generator to generate the public key and the secret key; and a decryptor to receive the encrypted data from the host device, and generate the decrypted data by decrypting the encrypted data. [Angel: para 0038-0039; decryption]
Claim 18: Angel: para 0029-0031, 0085 [key generation and encryption include input data]; discussing the system of claim 14, wherein the host device comprises a host processor configured to control: a key generator to generate the public key and the secret key; and an encryptor to receive the public key from the key generator, and generate the encrypted data by encrypting the input data, and wherein the processor of the storage controller is further configured to control a decryptor to receive the secret key and the encrypted data from the host device, and generate the decrypted data by decrypting the encrypted data. [Angel: para 0038-0039; decryption]
Claim 19: Angel: para 0112; discussing the system of claim 14, wherein the storage controller is configured to store the encrypted data in the nonvolatile memory device, and generate the decrypted data by reading the encrypted data from the nonvolatile memory device and decrypting the encrypted data.
As per claim 20: Angel, et al. teaches a system comprising:
one or more memories; and [Angel: para 0005]
one or more processors configured to control: [Angel: para 0005]
a key generation device to generate a public key and a secret key, based on a master key and a parameter vector indicating parameters of a first layer of a machine learning model; [Angel: para 0027; Machine learning systems are becoming ubiquitous (e.g., recommendation systems, diagnostic/prediction systems, autopilot systems, pattern recognition systems, and so on), and their performance (e.g., optimization of parameters, weights, biases, and so on) can rely on large volumes of training data See also para0041; the key distribution component deliver a common public key or party-specific public key to the participants, based on the given master public/secret keys. The key distribution component take as input the master public/secret keys and a vector y generated by the coordinator (e.g., participant-related weight vector w.sub.p, sample-related weight vector u, and so on)]
an encryption device to generate encrypted data by encrypting, based on the public key, an input vector indicating input data of the machine learning model; [Angel: para 0040; functional encryption for inner-product (FEIP) schemes can be implemented, which can output the inner-product between ciphertext of one or more input vectors and predetermined vector without revealing the plaintext of the one or more input vectors]
a decryption device to generate decrypted data corresponding to an approximation value [Angel: para 0088; two kinds of aggregation computation could be linear or non-linear due to different underlying machine learning algorithms, such as linear classification models and/or non-linear logistic regression and/or tree ensemble models. An approximation approach where non-linear computations are transferred to linear computations by applying Taylor approximation] of an inner product of the parameter vector and the input vector by decrypting the encrypted data based on the secret key [Angel: para 0039-0040; FE techniques belong to a public-key encryption family where the decryptor can be issued a secret key (e.g., a functionally derived key) that allows the decryptor to learn the result of a function over a ciphertext without learning the corresponding plaintext. Without revealing x, where D.sub.sk represents decryption by the functional secret key, E.sub.pk represents encryption by the public key, and ƒ represents the desired output function. Functional encryption for inner-product (FEIP) schemes can be implemented, which can output the inner-product between ciphertext of one or more input vectors and some other given, desired, and/or predetermined vector without revealing the plaintext of the one or more input vectors (e.g., ƒ above can be an inner-product function)], **wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product; and [**rejected under a secondary reference, discussion below]
a machine learning engine to receive the decrypted data from the decryption device and control the machine learning model based on the decrypted data to obtain an output value. [Angel: para 0041; can generate as output an MIFE and/or SIFE function derived key that computes an inner-product between input ciphertext and the vector y generated by the coordinator. The coordinator can accordingly apply the function derived key on data x or set of data {x.sub.i}received from the participants to yield ƒ.sub.SIFE(x. y) or ƒ.sub.MIFE({x.sub.i}, y).]
Angel teaches FE techniques belong to a public-key encryption family where the decryptor can be issued a secret key (e.g., a functionally derived key) that allows the decryptor to learn the result of a function over a ciphertext without learning the corresponding plaintext. Without revealing x, where D.sub.sk represents decryption by the functional secret key, E.sub.pk represents encryption by the public key, and ƒ represents the desired output function. Functional encryption for inner-product (FEIP) schemes can be implemented, which can output the inner-product between ciphertext of one or more input vectors and some other given, desired, and/or predetermined vector without revealing the plaintext of the one or more input vectors (e.g., ƒ above can be an inner-product function). Both single-input functional encryption (SIFE) and multi-input functional encryption (MIFE) techniques can be used to generate inner-product output [0039-0040]. However, Angel did not clearly teach “wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product”.
Becker teaches a modified version of a Learning with Errors (LWE) asymmetric cryptographic system. LWE is a form of lattice-based cryptography that uses a mathematical lattice as the basis for cryptographic primitives that produce a public key and a private key in an asymmetric cryptographic system. More broadly, a reference to LWE cryptography refers to a class of cryptographic systems that rely on the underlying lattice mathematics problems known to the art, such as the shortest vector problem (SVP), to implement an asymmetric cryptographic system with public and private keys. FIG. 9 depicts a simplified prior-art key generation process that produces a public/private key pair that can be used with an LWE cryptographic system. In FIG. 9, the LWE public key is a two-part key that includes a matrix A and a vector b, which are both made publicly available [Becker: para 0028-0029]. Becker discloses the client uses the inner public key to generate an encrypted representation of plaintext data with added noise. The client generates an error vector that is indistinguishable from a random Gaussian noise distribution based on the encrypted data generated using the inner public key. The client uses the partial outer public key, which is equivalent to the matrix A shown above, with the individual secret key and the specially generated error vector to produce an output vector b that forms the rest of a public key that is then transmitted to the untrusted aggregator. Thus, the client generates the vector b that is part of an LWE public key where the vector b also contains an encrypted representation of the plaintext data in the error vector data e [Becker: para 0037]. As such, Angel obviously suggest “wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product”, one would be motivated to prevent attack where the private key is not revealed to the public and is only accessible by trusted devices in the LWE cryptographic system.
Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to combine Becker with Angel to teach “wherein the public key, the secret key, the encrypted data and the decrypted data are generated based on functional encryption using a learning with error (LWE) problem and a lattice problem, and wherein random noise is included in the decrypted data such that the decrypted data is different from the inner product” for the reason to prevent attack by not revealing the private key to the public and is only accessible by trusted devices in the LWE cryptographic system [Becker: para 0030].
Claim 21: Angel: para 0010, 0122; discussing the system of claim 14, wherein the decrypted data comprise node output values corresponding to nodes of the first layer, and wherein the system is further configured to apply the node output values as input to a second layer of the machine learning model.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
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, Amir Mehrmanesh can be reached at 571-270-3351. 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.
Leynna Truvan
Examiner
Art Unit 2435
/L.TT/Examiner, Art Unit 2435
/EDWARD ZEE/Primary Examiner, Art Unit 2435
Prosecution Insights