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 .
DETAILED ACTION
The following is a non-Final Office Action in response to applicant’s arguments/filing filed on January 14, 2025
Claim 1-20 are pending
Foreign Priority
Acknowledgment is made of applicant's claim for foreign priority under 35 U.S.C. 119(a)-(d). The certified copy has been filed in Instant Application, filed on 1/14/2025.
Information Disclosure Statement
The information disclosure statement (IDS) submitted on 1/14/2026 was filed prior to the mailing date of the first office action. The submission is in compliance with the provisions of 37 CFR 1.97. Accordingly, the information disclosure statement is being considered by the examiner.
Drawings
Acknowledgment is made of applicant’s drawings submitted on 1/14/2025.
Oath/Declaration
Acknowledgment is made of applicant’s oath submitted on 1/14/2025
Application Data Sheet
Acknowledgment is made of applicant’s application data sheet submitted on 1/14/2025.
Claim Rejections - 35 USC § 102
The following is a quotation of the appropriate paragraphs of 35 U.S.C. 102 that form the basis for the rejections under this section made in this Office action:
A person shall be entitled to a patent unless –
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 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.
(a)(1) the claimed invention was patented, described in a printed publication, or in public use, on sale or otherwise available to the public before the effective filing date of the claimed invention.
Claims 1, 3, 5, 7, 8, 11, 13, 15, 17, and 18 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by US 9838407, Oprea
In regards to claim 1, Oprea teaches a method comprising steps of:responsive to training one or more machine learning models for performing classification of domains(US 9838407, Oprea, col. 18, lines 64-66, Classifying all domains observed in an enterprise with high accuracy and low false positives is extremely challenging in our setting due to a number of reasons.), the training including performing one or more optimizations to the one or more machine learning models, receiving a domain(US 9838407, Oprea, col. 19, lines 54-59, we determine the minimum number of features among the 365 in the final list required for optimizing our metrics. In machine learning, simple models with smaller number of features are generally preferred to overly complex models employing large number of features due to the risk of overfitting the model to the training set.);obtaining data associated with the domain including log data from a cloud-based system that performs monitoring of a plurality of users(US 9838407, Oprea, col. 16, lines 63-67, Domain WHOIS information is a useful indication of malicious activity. Following this trail, we issue WHOIS lookups for all the monitored domains and extract registration/update/expiration dates and registrant email for detection.); andanalyzing the domain via the one or more trained machine learning models for classifying the domain(US 9838407, Oprea, col. 8, lines 50-58, In some embodiments, the regression model 114 is trained on a training set that comprises a plurality of benign or unclassified domains and a plurality of domains previously classified as malicious domains but that excludes a global whitelist of popular domains. Such an arrangement facilitates the detection of domains associated with malicious activity from a potentially very large number of unknown domains. Other types of training sets can be used in other embodiments.).
In regards to claim 3, Oprea teaches the method of claim 1, wherein the steps further comprise:performing an action based on the classifying, the action comprising any of blocking the domain, allowing the domain, and isolating the domain(US 9838407, Oprea, fig. 2, col. 7, lines 48-53, Steps 206 and 208 are examples of malicious domain identification and proactive prevention steps assumed to be performed by the malicious domain identifier 116 and proactive malware infection prevention module 118, respectively, of the network security system 105.).
In regards to claim 5, Oprea teaches the method of claim 1, wherein the classifying comprises predicting a likelihood the domain is malicious or benign(US 9838407, Oprea, col. 2, lines 32-37, A subset of the domains are identified based on their respective malicious activity risk scores, and one or more proactive security measures are taken against the identified subset of domains. The malicious activity risk scores illustratively indicate likelihoods that the respective domains are associated with malware.).
In regards to claim 7, Oprea teaches the method of claim 1, wherein the classifying comprises predicting a likelihood the domain is a command and control site(US 9838407, Oprea, col. 2, lines 39-43, As a more particular example, a domain associated with malware may comprise a command-and-control (“C&C”) domain that malware on an enterprise host communicates with in order to receive further instructions.).
In regards to claim 8, Oprea teaches the method of claim 1, wherein the classifying comprises determining a category for the domain(US 9838407, Oprea, col. 9, lines 3-7, With reference now to FIG. 3, an example set of internal features is shown. The internal features in this example set are illustratively organized into seven distinct categories, including communication related features, domain structure related features,).
In regards to claim 11, Oprea teaches a non-transitory computer-readable storage medium having computer readable code stored thereon for programming at least one processor to perform steps of:responsive to training one or more machine learning models for performing classification of domains(US 9838407, Oprea, col. 18, lines 64-66, Classifying all domains observed in an enterprise with high accuracy and low false positives is extremely challenging in our setting due to a number of reasons.), the training including performing one or more optimizations to the one or more machine learning models, receiving a domain(US 9838407, Oprea, col. 19, lines 54-59, we determine the minimum number of features among the 365 in the final list required for optimizing our metrics. In machine learning, simple models with smaller number of features are generally preferred to overly complex models employing large number of features due to the risk of overfitting the model to the training set.);obtaining data associated with the domain including log data from a cloud-based system that performs monitoring of a plurality of users(US 9838407, Oprea, col. 16, lines 63-67, Domain WHOIS information is a useful indication of malicious activity. Following this trail, we issue WHOIS lookups for all the monitored domains and extract registration/update/expiration dates and registrant email for detection.); andanalyzing the domain via the one or more trained machine learning models for classifying the domain(US 9838407, Oprea, col. 8, lines 50-58, In some embodiments, the regression model 114 is trained on a training set that comprises a plurality of benign or unclassified domains and a plurality of domains previously classified as malicious domains but that excludes a global whitelist of popular domains. Such an arrangement facilitates the detection of domains associated with malicious activity from a potentially very large number of unknown domains. Other types of training sets can be used in other embodiments.).
In regards to claim 13, Oprea teaches the non-transitory computer-readable storage medium of claim 11, wherein the steps further comprise:performing an action based on the classifying, the action comprising any of blocking the domain, allowing the domain, and isolating the domain(US 9838407, Oprea, fig. 2, col. 7, lines 48-53, Steps 206 and 208 are examples of malicious domain identification and proactive prevention steps assumed to be performed by the malicious domain identifier 116 and proactive malware infection prevention module 118, respectively, of the network security system 105.).
In regards to claim 15, Oprea teaches the non-transitory computer-readable storage medium of claim 11, wherein the classifying comprises predicting a likelihood the domain is malicious or benign(US 9838407, Oprea, col. 2, lines 32-37, A subset of the domains are identified based on their respective malicious activity risk scores, and one or more proactive security measures are taken against the identified subset of domains. The malicious activity risk scores illustratively indicate likelihoods that the respective domains are associated with malware.).
In regards to claim 17, Oprea teaches the non-transitory computer-readable storage medium of claim 11, wherein the classifying comprises predicting a likelihood the domain is a command and control site(US 9838407, Oprea, col. 2, lines 39-43, As a more particular example, a domain associated with malware may comprise a command-and-control (“C&C”) domain that malware on an enterprise host communicates with in order to receive further instructions.).
In regards to claim 18, Oprea teaches the non-transitory computer-readable storage medium of claim 11, wherein the classifying comprises determining a category for the domain(US 9838407, Oprea, col. 9, lines 3-7, With reference now to FIG. 3, an example set of internal features is shown. The internal features in this example set are illustratively organized into seven distinct categories, including communication related features, domain structure related features,).
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 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 of this title, 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.
Claims 2 and 12 are rejected under 35 U.S.C. 103 as being unpatentable over US 9838407, Oprea in view of US 20180232637, Caicedo
In regards to claim 2, Oprea teaches the method of claim 1. Oprea does not teach wherein the one or more optimizations comprise any of 4-bit quantization of trainable parameters, utilization of bfloat16 memory for non-trainable parameters, utilization of flash attention mechanisms, and utilization of Rotation Positional Encoding (RoPE) However, Caicedo teaches wherein the one or more optimizations comprise any of 4-bit quantization of trainable parameters, utilization of bfloat16 memory for non-trainable parameters, utilization of flash attention mechanisms, and utilization of Rotation Positional Encoding (RoPE) (US 20180232637, Caicedo, para. 0075, Device 150 is configured to train model 120 using the training data by iteratively updating the quantization boundaries. For example, device 150 may use an iterative optimization algorithm such as backpropagation and gradient descent.). 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 teaching of Oprea with the teaching of Caicedo because a user would have been motivated to implement optimized quantization boundaries, taught by Caicedo, in order to improve a prediction for a targeted metric using Artificial Intelligence for the system taught by Oprea(Caicedo, para. 0014)
In regards to claim 12, Oprea teaches the non-transitory computer-readable storage medium of claim 11. Oprea does not teach wherein the one or more optimizations comprise any of 4-bit quantization of trainable parameters, utilization of bfloat16 memory for non-trainable parameters, utilization of flash attention mechanisms, and utilization of Rotation Positional Encoding (RoPE) ) However, Caicedo teaches wherein the one or more optimizations comprise any of 4-bit quantization of trainable parameters, utilization of bfloat16 memory for non-trainable parameters, utilization of flash attention mechanisms, and utilization of Rotation Positional Encoding (RoPE) ) (US 20180232637, Caicedo, para. 0075, Device 150 is configured to train model 120 using the training data by iteratively updating the quantization boundaries. For example, device 150 may use an iterative optimization algorithm such as backpropagation and gradient descent.). 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 teaching of Oprea with the teaching of Caicedo because a user would have been motivated to implement optimized quantization boundaries, taught by Caicedo, in order to improve a prediction for a targeted metric using Artificial Intelligence for the system taught by Oprea(Caicedo, para. 0014)
Claims 4 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over US 9838407, Oprea in view of US 20200382361, Chandrasekhar
In regards to claim 4, Oprea teaches the method of claim 3. Oprea does not teach wherein the steps further comprise:providing an explanation of a classification of the domain via an interactive User Interface (UI) responsive to performing the action However, Chandrasekhar teaches wherein the steps further comprise:providing an explanation of a classification of the domain via an interactive User Interface (UI) responsive to performing the action (US 20200382361, Chandrasekhar, para. 0113, after the RCA 420 determines the root causes for the detected anomalies, an explanation of the root cause as well as the remedial action 422 can be displayed, on a user interface). 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 teaching of Oprea with the teaching of Chandrasekhar because a user would have been motivated to perform discretization of key performance indicators, taught by Chandrasekhar, in order to improve the efficiency of performance management in the machine learning based system taught by Oprea(Chandrasekhar, para 0119)
In regards to claim 14, Oprea teaches the non-transitory computer-readable storage medium of claim 13. Oprea does not teach wherein the steps further comprise: providing an explanation of a classification of the domain via an interactive User Interface (UI) responsive to performing the action However, Chandrasekhar teaches wherein the steps further comprise: providing an explanation of a classification of the domain via an interactive User Interface (UI) responsive to performing the action (US 20200382361, Chandrasekhar, para. 0113, after the RCA 420 determines the root causes for the detected anomalies, an explanation of the root cause as well as the remedial action 422 can be displayed, on a user interface). 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 teaching of Oprea with the teaching of Chandrasekhar because a user would have been motivated to perform discretization of key performance indicators, taught by Chandrasekhar, in order to improve the efficiency of performance management in the machine learning based system taught by Oprea(Chandrasekhar, para 0119)
4.) Claims 6 and 16 are rejected under 35 U.S.C. 103 as being unpatentable over US 9838407, Oprea in view of US 20100043071, Wang
In regards to claim 6, Oprea teaches the method of claim 1. Oprea does not teach wherein the classifying comprises categorizing the domain as phishing However, Wang teaches wherein the classifying comprises categorizing the domain as phishing (US 20100043071, Wang, para. 0022, If the two domain names don't match, then the server 110 classifies the link as including a potential phishing URL.). 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 teaching of Oprea with the teaching of Wang because a user would have been motivated to enhance security, in the system taught by Oprea, by enabling for the classification of phishing attacks, taught by Wang, in order to alert users of potential phishing URLs(Wang, para. 0022)
In regards to claim 16, Oprea teaches the non-transitory computer-readable storage medium of claim 11. Oprea does not teach wherein the classifying comprises categorizing the domain as phishing. However, Wang teaches wherein the classifying comprises categorizing the domain as phishing(US 20100043071, Wang, para. 0022, If the two domain names don't match, then the server 110 classifies the link as including a potential phishing URL.). 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 teaching of Oprea with the teaching of Wang because a user would have been motivated to enhance security, in the system taught by Oprea, by enabling for the classification of phishing attacks, taught by Wang, in order to alert users of potential phishing URLs(Wang, para. 0022)
5.) Claims 9, 10, 19 and 20 are rejected under 35 U.S.C. 103 as being unpatentable over US 9838407, Oprea in view of US 11531846, Bodapati
In regards to claim 9, Oprea teaches the method of claim 1. Oprea does not teach wherein the one or more machine learning models are Large Language Models (LLMs) However, Bodapati teach wherein the one or more machine learning models are Large Language Models (LLMs)(US 11531846, Bodapati, col. 9, lines 10-16, In some embodiments, during transcription an ASR engine 325 searches through a language model (LM) 327 by incrementally generating hypotheses based on incoming audio data. The set of hypotheses is then reduced and simplified by a process called determinization. Since larger LMs [i.e. note: LLM] tend to yield more hypotheses, determinization time increases with LM size.). 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 teaching of Oprea with the teaching of Bodapati because a user would have been motivated to employ a plurality of machine learning models, taught by Bodapati, in order to identify and protect a plurality of sensitive data types in the machine learning system taught by Oprea(Bodapati, col. 4, lines 22-30)
In regards to claim 10, the combination of Oprea and Bodapati teach the method of claim 9, wherein the one or more LLMs are small sized LLMs(US 11531846, Bodapati, col. 9, lines 34-36, Typically, for rescoring a smaller LM [i.e. note: small LLM] is used by the ASR engine 325 in the “first pass” search to construct the initial set of hypotheses.). 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 teaching of Oprea with the teaching of Bodapati because a user would have been motivated to employ a plurality of machine learning models, taught by Bodapati, in order to identify and protect a plurality of sensitive data types in the machine learning system taught by Oprea(Bodapati, col. 4, lines 22-30)
In regards to claim 19, Oprea teaches the non-transitory computer-readable storage medium of claim 11. Oprea does not teach wherein the one or more machine learning models are Large Language Models (LLMs) However, Bodapati teaches wherein the one or more machine learning models are Large Language Models (LLMs)(US 11531846, Bodapati, col. 9, lines 10-16, In some embodiments, during transcription an ASR engine 325 searches through a language model (LM) 327 by incrementally generating hypotheses based on incoming audio data. The set of hypotheses is then reduced and simplified by a process called determinization. Since larger LMs [i.e. note: LLM] tend to yield more hypotheses, determinization time increases with LM size.). 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 teaching of Oprea with the teaching of Bodapati because a user would have been motivated to employ a plurality of machine learning models, taught by Bodapati, in order to identify and protect a plurality of sensitive data types in the machine learning system taught by Oprea(Bodapati, col. 4, lines 22-30)
In regards to claim 20, the combination of Oprea and Bodapati teach the non-transitory computer-readable storage medium of claim 19, wherein the one or more LLMs are small sized LLMs(US 11531846, Bodapati, col. 9, lines 34-36, Typically, for rescoring a smaller LM [i.e. note: small LLM] is used by the ASR engine 325 in the “first pass” search to construct the initial set of hypotheses.). 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 teaching of Oprea with the teaching of Bodapati because a user would have been motivated to employ a plurality of machine learning models, taught by Bodapati, in order to identify and protect a plurality of sensitive data types in the machine learning system taught by Oprea(Bodapati, col. 4, lines 22-30)
CONCLUSION
Any inquiry concerning this communication or earlier communications from the examiner should be directed to GREGORY LANE whose telephone number is (571)270-7469. The examiner can normally be reached on 571 270 7469 from 8:00 AM to 6:00 PM.
If attempts to reach the examiner by telephone are unsuccessful, the examiner's supervisor, Taghi Arani, can be reached on 571 272 3787. 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 the Patent Application Information Retrieval (PAIR) system. Status information for published applications may be obtained from either Private PAIR or Public PAIR. Status information for unpublished applications is available through Private PAIR only. For more information about the PAIR system, see http://pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
/GREGORY A LANE/Examiner, Art Unit 2438
/TAGHI T ARANI/Supervisory Patent Examiner, Art Unit 2438
Prosecution Insights