DETAILED ACTION
This Office Action is in response to the application filed on 06/10/2025.
Claims 1-20 are presented for examination.
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 .
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 –
(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.
Claim(s) 1, 3 is rejected under 35 U.S.C. 102(a)(1) as being anticipated by Uniyal et al. (hereinafter Uniyal, US 2024/0135154 A1).
Regarding Claim 1, Uniyal discloses A computer system (Uniual: para.0005 system) comprising: one or more processors (Uniyal: para.0041); and a computer-readable storage media storing computer-executable instructions, wherein the computer-executable instructions, when executed by the one or more processors, cause the computer system to perform (Uniyal: para.0046):
receiving historical root cause analysis (RCA) documents specific to a network service provider (Uniyal: para.0015 “ The information in the second database may, for example, include usage logs or any other suitable information that includes historical data relating to performance issues associated with the plurality of digital applications. The information may preferably be text-based.” Para.0035 “In some embodiments, the system may be further configured to store, in the second database, historical solution data. The historical solution data may include information relating to actions that successfully resolved the historical performance issues. As part of the training, the system may be configured to feed the hybrid neural network with the historical solution data.” Para.0006 “The hybrid neural network may be trained by: generating a first representation of the information in the first database; generating a second representation of the information in the second database; generating, via an embedding algorithm, a first information vector from the first representation; generating, via the embedding algorithm, a second information vector from the second representation; feeding, using a masking algorithm, the first and the second information vectors to the transformer block neural network; and feeding the second representation to the NLP engine.” The second database stores both historical network data and historical solution data. The historical solution data is obtained, i.e. RCA documents are received, from the database and used by the system to train a hybrid neural network in para.0023 using a vector format. These documents are specific to a particular service provider for a set of applications.);
receiving network data associated with the RCA documents (Uniyal: para.0015 “ The information in the second database may, for example, include usage logs or any other suitable information that includes historical data relating to performance issues associated with the plurality of digital applications. The information may preferably be text-based.”Para.0006 “The hybrid neural network may be trained by: generating a first representation of the information in the first database; generating a second representation of the information in the second database; generating, via an embedding algorithm, a first information vector from the first representation; generating, via the embedding algorithm, a second information vector from the second representation; feeding, using a masking algorithm, the first and the second information vectors to the transformer block neural network; and feeding the second representation to the NLP engine.” Para.0060 “Data may be passed to a data collection API (Application Programming Interface) at step 411. The data may be training data. The data may include event information, identification of human or non-human source of the event, and a time stamp of the events. Training data fed into the system may also be provided from data lake 413. The data in the data lake may include historical data relating to applications and issues associated with the applications. For example, the data may include user information 415, security issue database 417, application database 419, infrastructure database 421, and network database 423.” The second database stores both historical network data and historical solution data. The historical network data is obtained and used by the system to train a hybrid neural network in para.0023 using a vector format. See also Fig. 3 301 listing potential past network issues related to an application.);
processing the RCA documents to generate RCA data based on the RCA documents and the network data (Uniyal: Para.0006 “The hybrid neural network may be trained by: generating a first representation of the information in the first database; generating a second representation of the information in the second database; generating, via an embedding algorithm, a first information vector from the first representation; generating, via the embedding algorithm, a second information vector from the second representation; feeding, using a masking algorithm, the first and the second information vectors to the transformer block neural network; and feeding the second representation to the NLP engine.” para.0023 “The hybrid neural network may be trained by generating a first representation of the information in the first database and generating a second representation of the information in the second database. The first and second representations may include vectors that contain the underlying information. For example, the information may be converted to a character string which may in turn be converted to a vector containing the character string.” The data from the second database, including both RCA documents, the historical solution data para.0035, and historical network data in para.0015, is processed into a second representation);
generating one or more vectors based on the RCA data (Uniyal: Para.0006 “The hybrid neural network may be trained by: generating a first representation of the information in the first database; generating a second representation of the information in the second database; generating, via an embedding algorithm, a first information vector from the first representation; generating, via the embedding algorithm, a second information vector from the second representation; feeding, using a masking algorithm, the first and the second information vectors to the transformer block neural network; and feeding the second representation to the NLP engine.” para.0023 “The hybrid neural network may be trained by generating a first representation of the information in the first database and generating a second representation of the information in the second database. The first and second representations may include vectors that contain the underlying information. For example, the information may be converted to a character string which may in turn be converted to a vector containing the character string.” The data from the second database, including both RCA documents, the historical solution data para.0035, and historical network data in para.0015, is processed into a second representation, which is then converted into a second information vector.); and
constructing and training one or more Al models based on the RCA data (Uniyal: para.0023 “The hybrid neural network may be trained by generating a first representation of the information in the first database and generating a second representation of the information in the second database.” Para.0024 “The training may also include generating, via an embedding algorithm, a first information vector from the first representation and generating, via the embedding algorithm, a second information vector from the second representation.” para.0025 “The training may also include feeding, using a masking algorithm, the first and the second information vectors to the transformer block neural network.” An AI model, in this case neural network model, is constructed via training using the information vectors generated from the first and second representations.).
Regarding Claim 3, Uniyal discloses claim 1 as set forth above.
Uniyal further discloses wherein the instructions when executed by the one or more processors, further cause the computer system to perform: receiving a query from a network operator of the network service provider (Uniyal: para.0062 “Diagram 400 shows an illustrative example of a manual trigger event according to aspects of the disclosure. Developer/application team 403 may include application owner 405, application user 407, and developer lead 409. Any one of the team may submit a request for a prediction from the system. The request may be submitted along with documentation of the event for which the prediction is sought.” An operator of the network service provider, in this case a developer/application team 403 for the service being provided on the network, sends a query.),
the query indicating a suspicious network incident and including network data pertaining to the suspicious network incident (Uniyal: para.0062 “Any one of the team may submit a request for a prediction from the system. The request may be submitted along with documentation of the event for which the prediction is sought.” Para.0030 “In some embodiments, the information relating to the trigger event may include documentation describing the trigger event or a log describing the trigger event.” A predication of a potential event and documentation of the event is provided. Para.0059 “Trigger events may also include new or updated technology, network change, server change, add, or update, maximum threshold reached in infrastructure (e.g., maximum utilization reached or exceeded), integration with an application, change in data provider for a database and/or cloud, or any other suitable event that may increase the potential for a system failure.” The type of events include suspicious network activity such as any of the trigger events above.);
identifying/verifying one or more network incidents using the AI models, based on the network data (Uniyal: para.0060 “Training data fed into the system may also be provided from data lake 413. The data in the data lake may include historical data relating to applications and issues associated with the applications. For example, the data may include user information 415, security issue database 417, application database 419, infrastructure database 421, and network database 423…. This data may be used to train ML model 427 using various vectors and feature transforms, as described in detail elsewhere herein.” Para.0061 “Using the aforementioned data to train ML model 427 may provide a model that is able to track relationships between historical issues with an application, a user or developer or general information of an application, and from similar/correlated applications and user/developers as well. For example, the ML model may create relationships and store a close distance between a particular developer of a first application and the same developer of a second application. If the developer made a mistake and caused an application failure with the first application, the system may calculate a high probability of the same developer causing a failure when they update the second application.” Para.0062 “The model may generate output 431, which may be a table whose rows represent a list of various potential issues. Columns of the table may display information relating to the potential issues, such as app ID, source of the issue, a description of the issue, a recommended resolution (which may be a preventative measure), and a time for the potential issue.” Historical data is used by the ML model that verifies/identifies events based on the query from the operator.);
determining one or more root causes of the identified/verified network incidents using the AI models (Uniyal: Para.0062 “The model may generate output 431, which may be a table whose rows represent a list of various potential issues. Columns of the table may display information relating to the potential issues, such as app ID, source of the issue, a description of the issue, a recommended resolution (which may be a preventative measure), and a time for the potential issue.” Source of the identified issue may be identified.);
recommending one or more actions to resolve the network incidents (Uniyal: Para.0062 “The model may generate output 431, which may be a table whose rows represent a list of various potential issues. Columns of the table may display information relating to the potential issues, such as app ID, source of the issue, a description of the issue, a recommended resolution (which may be a preventative measure), and a time for the potential issue.” A recommended resolution provided); and
generating a response for output (Uniyal: para.0062 “The model may generate output 431, which may be a table whose rows represent a list of various potential issues. Columns of the table may display information relating to the potential issues, such as app ID, source of the issue, a description of the issue, a recommended resolution (which may be a preventative measure), and a time for the potential issue.” The response is in the form of a table comprising the issues and recommendations.).
Claim(s) 4-7 and 13-15, 17 are rejected under 35 U.S.C. 102(a)(1) as being anticipated by Yokel (US 2018/0131810 A1).
Regarding Claim 4, Yokel discloses A method for determining a root cause of a cellular network errors (Yokel: para.0032 “In various embodiments, the support service engine 104 may use a machine learning algorithm to determine a root cause of the service issue.” Para.0014 “ For example, a trouble report may simply recite that cellular calls are being dropped.” Root cause of a service issue in a cellular network.),
comprising: receiving, by a computing system, an error log indicating an error within a cellular network (Yokel: para.0040 “ The processors 204 and the memory 206 of the computing devices 108 may implement an operating system 210. In turn, the operating system 210 may provide an execution environment for the support service engine 104 and the support session engine 106.” para.0073 “At block 302, the support service engine 104 may receive a trouble report from a customer of a wireless telecommunication network 102 via an online chat session or a telephone call.” An issue in the cellular network can be received by system 108 in fig. 1 and fig. 2., para.0014 “cellular calls…cellular coverage”);
providing, by the computing system, the error log to a machine learning module (MLM) (Yokel: para.0074 “At block 304, the support service engine 104 may determine a service issue associated with the trouble report via a machine learning classification algorithm.” Para.0046 “ The issue classification module 214 analyze a trouble report from a customer to determine an underlying service issue that affects the customer. In various embodiments, the issue classification module 214 may use a machine learning algorithm to assign a service issue to the trouble report.” System 108 provides this information to its support service engine 104 running on the computing device.),
the MLM configured to determine a root cause of the error by: parsing, by the MLM, the error log to identify one or more datapoints (Yokel: para.0074 “In various embodiments, the machine learning classification algorithm may match specific words or phrases that the customer used in the trouble report to a specific service issue.” Para.0046 “ In various embodiments, the issue classification module 214 may use a machine learning algorithm to assign a service issue to the trouble report. The assignment may be performed based on descriptive words and/or phrases used by the customer in the trouble report, the relevant contextual data from the operation databases 118 and/or the third-party databases 122, and/or so forth.” Datapoint of a classification, i.e. data point, of the issue is identified by parsing the trouble report for words and phrases by the classification module of the MLM, support service engine 104..);
determining, by the MLM, a root cause of the error by utilizing the one or more datapoints as inputs to an artificial intelligence engine configured to associate the one or more datapoints with the root cause (Yokel: Fig. 2 218, para.0056 “The recommendation module 218 may use a machine learning-based recommendation algorithm to generate potential solutions for service issues that are classified from trouble reports. A potential solution for a service issue may be presented by a support person to a customer during a support session in the form of an online chat session or a telephone call. The machine learning-based recommendation algorithm may be a naïve Bayes algorithm, a Bayesian network algorithm, a decision tree algorithm, a neural network algorithm, a support vector machine, and/or so forth. In operation, the recommendation module 218 may use a machine learning algorithm to determine a root cause for a service issue. Subsequently, the recommendation module 218 may find one or more matching potential solutions for the root cause from a solutions database. For example, if the root cause of a service issue is the lack of network coverage, the potential solution for the root cause may be the activation of Wi-Fi calling and/or the installation of a Wi-Fi calling capable router. In another example, if the root cause of the service issue is user device software that is incompatible with the wireless telecommunication network 102, the potential solution may be an upgrade to the software of the user device.” A root cause can be associated with the classification of the issue via a recommendation module 218, the artificial intelligence engine of the support service engine.); and
determining, by the MLM, a corrective action to be taken such that the error is corrected (Yokel: Fig. 3 312, para.0078 “At block 312, the support service engine 104 may generate a potential solution for the service issue based on the problem summary using a machine learning-based recommendation algorithm.” The support service engine determines a corrective action to be taken, potential solution.); and
outputting, by the computing system, data indicating at least one of the error, the root cause, or the corrective action (Yokel: poara.0032 “The root cause and the potential solutions may be surfaced to the initial support person 126. In turn, the initial support person 126 may present the solution to the customer in the support session.” The root cause and the solution may be presented.).
Regarding Claim 5, Yokel disclose claim 4 as set forth above.
Yokel further discloses wherein the MLM comprises at least one of a large language model or a support vector machine (Yokel: Fig. 1-2, 104, para.0056 “The machine learning-based recommendation algorithm may be a naïve Bayes algorithm, a Bayesian network algorithm, a decision tree algorithm, a neural network algorithm, a support vector machine,” the support service engine 104 comprises the recommendation module having a machine learning based recommendation algorithm SVM.).
Regarding Claim 6, Yokel discloses claim 4 as set forth above.
Yokel further discloses wherein the artificial intelligence engine comprises a neural network (Yokel: Fig. 1-2, 104, para.0056 “The machine learning-based recommendation algorithm may be a naïve Bayes algorithm, a Bayesian network algorithm, a decision tree algorithm, a neural network algorithm, a support vector machine,” the support service engine 104 comprises the recommendation module having a machine learning based recommendation algorithm neural network ).
Regarding Claim 7, Yokel discloses claim 4 as set forth above.
Yokel further discloses wherein the error log comprises at least one of geographic data, software data, hardware data, user equipment (UE) data, an error type (Yokel: para.0040 “ The processors 204 and the memory 206 of the computing devices 108 may implement an operating system 210. In turn, the operating system 210 may provide an execution environment for the support service engine 104 and the support session engine 106.” para.0073 “At block 302, the support service engine 104 may receive a trouble report from a customer of a wireless telecommunication network 102 via an online chat session or a telephone call.” Para.0014 “The trouble report may contain an indication that service is unsatisfactory to the customer. The trouble report is distinct from the specific service issue or issues, which comprise the specific technical causes of the unsatisfactory service. For example, a trouble report may simply recite that cellular calls are being dropped. The specific service issues may be that the nearest base station is overloaded and/or weather conditions are interfering with cellular coverage.” Error type is indicated in the error log), or an error rate.
Regarding Claim 13-15, 17, it teaches all of the same elements as claim 4-7 but in A system for analyzing error logs, comprising: one or more processors; and a computer-memory comprising instructions that, when executed by the one or more processors, cause the system to: (Yokel: para.0038). Therefore the supporting rationale for the rejection to claims 4-7 apply equally as well to that of claims 13-15 and 17.
Claim Rejections - 35 USC § 103
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.
Claim(s) 2 is/are rejected under 35 U.S.C. 103 as being unpatentable over Uniyal et al. (hereinafter Uniyal, US 2024/0135154 A1) in view of Arora et al. (hereinafter Arora, US 2020/0175971 A1).
Regarding Claim 2, Uniyal discloses claim 1 as set forth above.
Uniyal further discloses wherein the instructions when executed by the one or more processors, further cause the computer system to perform: receiving a query from a network operator of the network service provider (Uniyal: para.0062 “Diagram 400 shows an illustrative example of a manual trigger event according to aspects of the disclosure. Developer/application team 403 may include application owner 405, application user 407, and developer lead 409. Any one of the team may submit a request for a prediction from the system. The request may be submitted along with documentation of the event for which the prediction is sought.” An operator of the network service provider, in this case a developer/application team 403 for the service being provided on the network, sends a query.);
identifying one or more of the historical RCA documents pertaining to the query using the Al models (Uniyal: para.0060 “Training data fed into the system may also be provided from data lake 413. The data in the data lake may include historical data relating to applications and issues associated with the applications. For example, the data may include user information 415, security issue database 417, application database 419, infrastructure database 421, and network database 423…. This data may be used to train ML model 427 using various vectors and feature transforms, as described in detail elsewhere herein.” Para.0061 “Using the aforementioned data to train ML model 427 may provide a model that is able to track relationships between historical issues with an application, a user or developer or general information of an application, and from similar/correlated applications and user/developers as well. For example, the ML model may create relationships and store a close distance between a particular developer of a first application and the same developer of a second application. If the developer made a mistake and caused an application failure with the first application, the system may calculate a high probability of the same developer causing a failure when they update the second application.” Para.0062 “The model may generate output 431, which may be a table whose rows represent a list of various potential issues. Columns of the table may display information relating to the potential issues, such as app ID, source of the issue, a description of the issue, a recommended resolution (which may be a preventative measure), and a time for the potential issue.” Historical issues are tracked via the ML model, therefore in operation, when creating its output for potential issues based on the query, historical RCA documents, i.e. past events stored in the data lake, are identified. The historical issues being the same data that is used to train the model.);
analyzing the query to extract one or more intents of the network operator (Uniyal: para.0062 “The request may be submitted via app server 433 which may generate the appropriate input form and feed the input to ML model 427.” The server transforms the query into a particular format, thereby analyzing the query and obtaining an intent of the query used to generate the appropriate form.);
generating contents using the Al models, the contents comprising data associated with the identified historical RCA documents and pertaining to the extracted intent (Uniyal: para.0062 “The model may generate output 431, which may be a table whose rows represent a list of various potential issues. Columns of the table may display information relating to the potential issues, such as app ID, source of the issue, a description of the issue, a recommended resolution (which may be a preventative measure), and a time for the potential issue.” The ai model is used to then determine content associated with the previous events and the query to obtain an output of potential issues.); and
generating a response comprising the contents for output (Uniyal: para.0062 “The model may generate output 431, which may be a table whose rows represent a list of various potential issues. Columns of the table may display information relating to the potential issues, such as app ID, source of the issue, a description of the issue, a recommended resolution (which may be a preventative measure), and a time for the potential issue.” The response is in the form of a table comprising the potential issues.).
However Uniyal does not explicitly disclose analyzing the query using the Al models to extract one or more intents of the network operator.
Arora discloses analyzing the query using the Al models to extract one or more intents of the network operator (Arora: para.0031 “The VNA receives from a voice server 150 a transcript and a general intent of a voice query 103 that is spoken by an operator 102.” para.0064 “ In some implementations, the machine learning models use context data 123 that includes historical network data for the communication system, historical data for other similar networks, data related to the operator 102, and/or other context data to refine the query intent. For example, if the current voice query 103 asks “Is there a problem with beam A?” the machine learning models may determine from historical network data that problems with beam A are often associated with a lower-than-normal level of beam A traffic. As a result, the machine learning models may determine that the refined intent of the query 103 is “determine whether the current beam A traffic is outside of a normal range for beam A traffic at this time of day.”” Para.0094 “If the module 270 has generated a query command for a database, after interpreting the query, the computer system retrieves the query results 227 (228). For example, the module 270 may retrieve current or historical performance data for the network from a network database.” A operators query can be analyzed and an intent may be extracted and refined.).
Therefore it would have been obvious to one of ordinary skill in the art before the effective filing date toe combine Uniyal with that of Arora in order to incorporate analyzing the query using the Al models to extract one or more intents of the network operator, such that when the system formats the user request in Uniyal, a machine learning model may be used to accurately determine the intent of the users request.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of accurately interpreting the query such that results of the query may be better aligned to the users intent (Arora: para.0062-0063).
Claim(s) 8, 16 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yokel (US 2018/0131810 A1) in view of Ting et al. (hereinafter Ting, US 2022/0217196 A1).
Regarding Claim 8, Yokel discloses claim 4 as set forth above.
Yokel further discloses one or more nodes of the MLM (Yokel: Fig. 1-2, 104, para.0056 “The machine learning-based recommendation algorithm may be a naïve Bayes algorithm, a Bayesian network algorithm, a decision tree algorithm, a neural network algorithm, a support vector machine,” the support service engine 104 comprises the recommendation module having a machine learning based recommendation algorithm neural network, which comprises of nodes)
However Yokel does not explicitly disclose receiving, by the computing system, retraining data based at least in part on the data indicating at least one of the error, the root cause, or the corrective action; and providing, by the computing system, the retraining data to the MLM such that one or more nodes of the MLM are reconfigured, and an accuracy of the MLM is increased when determining a future root cause.
Ting discloses receiving, by the computing system, retraining data based at least in part on the data indicating at least one of the error, the root cause, or the corrective action (Ting: para.0087 “ Moreover, machine-learning module 522 may provide corrective information 526 (such as call quality analysis and root-cause analysis) to an administrator of the network via a Web browser interface 412 and/or to the user of electronic device 114-1.” para.0098 “During the training, the neural network model may be repeatedly subjected to the training input set and retrained using a backpropagation technique, in which errors in the output from the neural network model may be used as feedback to adjust the parameters in the neural network model until the output error in the measurement results reaches some minimum level (such as 1, 5, 10, 20 or 30%).” Machine learning model performs root cause analysis, and it can be retrained based on outputs of the model, i.e. root cause result. Therefore retraining data is received from the neural network as an output of the neural network.); and
providing, by the computing system, the retraining data to the MLM such that one or more nodes of the MLM are reconfigured, and an accuracy of the MLM is increased when determining a future root cause (Ting: para.0087 “ Moreover, machine-learning module 522 may provide corrective information 526 (such as call quality analysis and root-cause analysis) to an administrator of the network via a Web browser interface 412 and/or to the user of electronic device 114-1.” para.0098 “During the training, the neural network model may be repeatedly subjected to the training input set and retrained using a backpropagation technique, in which errors in the output from the neural network model may be used as feedback to adjust the parameters in the neural network model until the output error in the measurement results reaches some minimum level (such as 1, 5, 10, 20 or 30%).” Retraining is performed to reach a threshold error rate, i..e accuracy is improved. This retraining of a neural network would alter the neurons of the neural network, para.0094).
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 Yokel with that of Ting in order to incorporate receiving, by the computing system, retraining data based at least in part on the data indicating at least one of the error, the root cause, or the corrective action; and providing, by the computing system, the retraining data to the MLM such that one or more nodes of the MLM are reconfigured, and an accuracy of the MLM is increased when determining a future root cause.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of improved accuracy of the model (Ting: para.0098).
Regarding Claim 16, Yokel discloses claim 13 as set forth above.
However Yokel does not explicitly disclose wherein the telecommunications network comprises a standalone SG cellular network.
Ting disclose wherein the telecommunications network comprises a standalone SG cellular network (Ting: para.0087 “ Moreover, machine-learning module 522 may provide corrective information 526 (such as call quality analysis and root-cause analysis) to an administrator of the network via a Web browser interface 412 and/or to the user of electronic device 114-1.” para.0098 “During the training, the neural network model may be repeatedly subjected to the training input set and retrained using a backpropagation technique, in which errors in the output from the neural network model may be used as feedback to adjust the parameters in the neural network model until the output error in the measurement results reaches some minimum level (such as 1, 5, 10, 20 or 30%).” Para.0103 “For example, networking subsystem 714 can include a Bluetooth networking system, a cellular networking system (e.g., a 3G/4G/5G network such as UMTS, LTE, etc.),” the root cause analysis operates for a telecommunication network comprising a 5G network.).
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 Yokel with that of Ting in order to incorporate r wherein the telecommunications network comprises a standalone SG cellular network.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of known advantages of a 5G network such as faster data speed (Ting: Para.0103).
Claim(s) 9 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yokel (US 2018/0131810 A1) Uniyal et al. (hereinafter Uniyal, US 2024/0135154 A1) in view of Garapati et al. (hereinafter Garapati, US 2025/0077851 A1).
Regarding Claim 9, Yokel discloses claim 4 as set forth above.
However Yokel does not explicitly disclose wherein the MLM comprises a large language model (LLM), the method further comprising: receiving, by the computing system, a training dataset comprising historical error logs; generating, by the computing system, a modified training dataset wherein the modified training dataset comprises transformed data of the training dataset; vectorizing, by the computing system, the training dataset and the modified training dataset to generate a preprocessed dataset; and providing, by the computing system, the preprocessed dataset to the MLM such that an accuracy of the LLM is increased when parsing a future error log.
Uniyal discloses the method further comprising: receiving, by the computing system, a training dataset comprising historical error logs (Uniyal: para.0015 “ The information in the second database may, for example, include usage logs or any other suitable information that includes historical data relating to performance issues associated with the plurality of digital applications. The information may preferably be text-based.” Para.0035 “In some embodiments, the system may be further configured to store, in the second database, historical solution data. The historical solution data may include information relating to actions that successfully resolved the historical performance issues. As part of the training, the system may be configured to feed the hybrid neural network with the historical solution data.” Para.0006 “The hybrid neural network may be trained by: generating a first representation of the information in the first database; generating a second representation of the information in the second database; generating, via an embedding algorithm, a first information vector from the first representation; generating, via the embedding algorithm, a second information vector from the second representation; feeding, using a masking algorithm, the first and the second information vectors to the transformer block neural network; and feeding the second representation to the NLP engine.” Historical logs of network issues are obtained.);
generating, by the computing system, a modified training dataset wherein the modified training dataset comprises transformed data of the training dataset (Uniyal: Para.0006 “The hybrid neural network may be trained by: generating a first representation of the information in the first database; generating a second representation of the information in the second database; generating, via an embedding algorithm, a first information vector from the first representation; generating, via the embedding algorithm, a second information vector from the second representation; feeding, using a masking algorithm, the first and the second information vectors to the transformer block neural network; and feeding the second representation to the NLP engine.” para.0023 “The hybrid neural network may be trained by generating a first representation of the information in the first database and generating a second representation of the information in the second database. The first and second representations may include vectors that contain the underlying information. For example, the information may be converted to a character string which may in turn be converted to a vector containing the character string.” The data from the second database, including both RCA documents, the historical solution data para.0035, and historical network data in para.0015, is processed into a second representation);
vectorizing, by the computing system, the training dataset and the modified training dataset to generate a preprocessed dataset (Uniyal: Para.0006 “The hybrid neural network may be trained by: generating a first representation of the information in the first database; generating a second representation of the information in the second database; generating, via an embedding algorithm, a first information vector from the first representation; generating, via the embedding algorithm, a second information vector from the second representation; feeding, using a masking algorithm, the first and the second information vectors to the transformer block neural network; and feeding the second representation to the NLP engine.” para.0023 “The hybrid neural network may be trained by generating a first representation of the information in the first database and generating a second representation of the information in the second database. The first and second representations may include vectors that contain the underlying information. For example, the information may be converted to a character string which may in turn be converted to a vector containing the character string.” The data from the second database, including both RCA documents, the historical solution data para.0035, and historical network data in para.0015, is processed into a second representation, which is then converted into a second information vector.); and
providing, by the computing system, the preprocessed dataset to the MLM such that an accuracy of the model is increased when parsing a future error log (Uniyal: para.0023 “The hybrid neural network may be trained by generating a first representation of the information in the first database and generating a second representation of the information in the second database.” Para.0024 “The training may also include generating, via an embedding algorithm, a first information vector from the first representation and generating, via the embedding algorithm, a second information vector from the second representation.” para.0025 “The training may also include feeding, using a masking algorithm, the first and the second information vectors to the transformer block neural network.” An AI model, in this case neural network model, is constructed via training using the information vectors generated from the first and second representations. Training of a model gradually increases its accuracy., para.0064).
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 Yokel with Uniyal in order to incorporate the method further comprising: receiving, by the computing system, a training dataset comprising historical error logs; generating, by the computing system, a modified training dataset wherein the modified training dataset comprises transformed data of the training dataset; vectorizing, by the computing system, the training dataset and the modified training dataset to generate a preprocessed dataset; and providing, by the computing system, the preprocessed dataset to the MLM such that an accuracy of the model is increased when parsing a future error log.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of improved accuracy of the model (Uniyal: para.0006, para.0024).
However Yokel-Uniyal does not explicitly disclose wherein the MLM comprises a large language model (LLM), providing, by the computing system, the preprocessed dataset to the MLM such that an accuracy of the LLM is increased when parsing a future error log
Garapati discloses wherein the MLM comprises a large language model (LLM) (Garapati: para.0046 “ As further illustrated, the LLM 153 may include one or more topological context adapter(s) 154, and may be configured to process, e.g., the event graph 146a and associated event text 146c, along with the network context 125, to generate a corresponding situation narrative 156, which may include root cause identification and explanation for the event graph 146a.” the LLM is trained for root cause analysis),
and providing, by the computing system, the preprocessed dataset to the MLM such that an accuracy of the LLM is increased when parsing a future error log (Garapati: para.0047 “As described in detail, below, the topological context adapter 154 may be trained using a training engine 160 and associated training data 162 to enable the topological context adaptor(s) 154 to provide a desired outcome, such as the situation narrative 156 or the remediation 158. For example, when training for generating the situation narrative 156, the training data 162 may include previously determined narratives associated with similar or related event graphs and associated situations, including root cause identification and explanation. When training for generating actionable remediations for resolving situations, the training data 162 may include previously determined remediations, worklogs, and other data associated with resolving previous IT situations.” Para.0048 “Similar comments apply when the topological context adaptor(s) 154 are trained to generate the remediation 158, which may, in those scenarios, be fed back to the training data 162 to obtain up-to-date, accurate, and evolving remediations for future situations” Previous datasets are fed to train the model to increase accuracy).
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 Yokel-Uniyal with Garapati in order to incorporate wherein the MLM comprises a large language model (LLM), providing, by the computing system, the preprocessed dataset to the MLM such that an accuracy of the LLM is increased when parsing a future error log.
One of ordinary skill in the art would have been motivate to combine because of the expected benefit of accurate remediation in IT situations (Garapati: para.0041).
Claim(s) 10, 19 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yokel (US 2018/0131810 A1) in view of Vichare et al. (hereinafter Vichare, US 2020/0005093 A1).
Regarding Claim 10, Yokel discloses claim 4 as set forth above.
However Yokel does not explicitly disclose generating, by the MLM, instructions based at least in part on the output indicating the corrective action; and transmitting, by the computing system, the instructions to one or more network components such that upon execution of the corrective action, at least a portion of the root cause is resolved.
Vichare discloses generating, by the MLM, instructions based at least in part on the output indicating the corrective action (Vichare: para.0034 “then the server may determine, using machine learning (e.g., K-means clustering or the like), what is causing the abnormally large number of events and select one or more remediation actions.” Para.0033 “ If more than a predetermined number of computing devices are generating the same (or similar) anomalous events, the particular anomaly may be re-classified as a bug. If more than a predetermined number of the same (or similar) anomalous events are generated by the same computing device, the server may select a remediation action, such as instructing the agent or software application to stop sending data or temporarily deactivating the API key.” Instructions may be generated to perform a remediation action.); and
transmitting, by the computing system, the instructions to one or more network components such that upon execution of the corrective action, at least a portion of the root cause is resolved (Vichare: para.0024 “The server may …The remediation action may include at least one of: (1) sending a first remediation instruction to the computing device instructing the computing device to: uninstall an application, download a latest version of the application, and install the latest version of the application,” the server sends instructions to computing devices to perform a corrective action, thereby resolving a cause of a remediable issue.).
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 Yokel with Vichare in order to incorporate generating, by the MLM, instructions based at least in part on the output indicating the corrective action; and transmitting, by the computing system, the instructions to one or more network components such that upon execution of the corrective action, at least a portion of the root cause is resolved.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of correcting a detected issue (Vichare: para.0024).
Regarding Claim 19, it does not teach nor further define over the limitations of claim 10, therefore the supporting rationale for the rejection to claim 10 applies equally as well to that of claim 19.
Claim(s) 11, 20 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yokel (US 2018/0131810 A1) in view of Vichare et al. (hereinafter Vichare, US 2020/0005093 A1) in view of Venkatachalam et al. (hereinafter Vent, US 2025/0165330 A1).
Regarding Claim 11, Yokel-Vichare discloses claim 10 as set forth above.
However Yokel-Vichare does not explicitly disclose wherein the MLM comprises a generative AI model.
Vent discloses wherein the MLM comprises a generative AI model (Vent: para.0043 “ Issue identification module 304 may receive the adverse trends 210. In some instances, issue identification module 304 may also include one or more AI models, such as a generative AI model, an LLM, such as GPT-4 or its variants in some examples. ” Generative ai may be used for issue identification).
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 Yokel-Vichare with Vent in order to incorporate wherein the MLM comprises a generative AI model.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of improved calculations by using a plurality of different models (Vent: para.0046).
Regarding Claim 20, it does not teach nor further define over the limitations of claim 11 therefore the supporting rationale for the rejection to claim 11 applies equally as well to that of claim 20.
Claim(s) 12, 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Yokel (US 2018/0131810 A1) in view of Bickford et al. (hereinafter Bickford, US 2018/0027416 A1).
Regarding Claim 12, Yokel discloses claim 4 as set forth above.
Yokel further discloses determining, by the MLM, one or more network components associated with the root cause (Yokel: para.0056 “Subsequently, the recommendation module 218 may find one or more matching potential solutions for the root cause from a solutions database. For example, if the root cause of a service issue is the lack of network coverage, the potential solution for the root cause may be the activation of Wi-Fi calling and/or the installation of a Wi-Fi calling capable router. In another example, if the root cause of the service issue is user device software that is incompatible with the wireless telecommunication network 102, the potential solution may be an upgrade to the software of the user device.” A network component can be identified for the root cause, a router or a user device in this case.).
However Yokel does not explicitly disclose determining, by the MLM, a respective entity associated with each of the one or more network components; and transmitting, by the computing system, the data to the respective entities.
Bickford discloses determining, by the MLM, a respective entity associated with each of the one or more network components (Bickford: para.0065 “In step 310, the processor receives a notification from a network service provider as to an anomaly associated with a particular mobile software application executing on a mobile endpoint device. For example, the network service provider has detected a mobile software application operating on the user's mobile endpoint device as causing an anomaly as discussed above. In one example, the notification is sent directly to the user operating the mobile endpoint device. Alternatively, the notification is sent to an entity responsible for the mobile endpoint device, e.g., an enterprise customer such as a business entity or a governmental agency, a guardian, a parent and so on. Alternatively, the notification is sent to an entity responsible for the mobile software application, e.g., a developer of the mobile software application.” An entity responsible for the device that is subject to an anomaly is identified.); and
transmitting, by the computing system, the data to the respective entities (Bickford: para.0065 “In step 310, the processor receives a notification from a network service provider as to an anomaly associated with a particular mobile software application executing on a mobile endpoint device. For example, the network service provider has detected a mobile software application operating on the user's mobile endpoint device as causing an anomaly as discussed above. In one example, the notification is sent directly to the user operating the mobile endpoint device. Alternatively, the notification is sent to an entity responsible for the mobile endpoint device, e.g., an enterprise customer such as a business entity or a governmental agency, a guardian, a parent and so on. Alternatively, the notification is sent to an entity responsible for the mobile software application, e.g., a developer of the mobile software application.” The data of the anomaly is sent as a notification to the respective entities.).
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 Yokel with Bickford in order to incorporate determining, by the MLM, a respective entity associated with each of the one or more network components; and transmitting, by the computing system, the data to the respective entities.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of resolving the issue by transmitting a notification to related parties (Bickford: para.0065).
Regarding Claim 18, Yokel discloses claim 13 as set forth above.
Yokel further discloses wherein the error comprises a hardware component error (Yokel: para.0056 “Subsequently, the recommendation module 218 may find one or more matching potential solutions for the root cause from a solutions database. For example, if the root cause of a service issue is the lack of network coverage, the potential solution for the root cause may be the activation of Wi-Fi calling and/or the installation of a Wi-Fi calling capable router. In another example, if the root cause of the service issue is user device software that is incompatible with the wireless telecommunication network 102, the potential solution may be an upgrade to the software of the user device.” A network component can be identified for the root cause, in this case missing a router.),
However Yokel does not explicitly disclose and the computing system determines an entity associated with the hardware component and transmits the data indicating at least one of the error, the root cause, or the corrective action to the entity.
Bickford discloses the computing system determines an entity associated with the hardware component (Bickford: para.0065 “In step 310, the processor receives a notification from a network service provider as to an anomaly associated with a particular mobile software application executing on a mobile endpoint device. For example, the network service provider has detected a mobile software application operating on the user's mobile endpoint device as causing an anomaly as discussed above. In one example, the notification is sent directly to the user operating the mobile endpoint device. Alternatively, the notification is sent to an entity responsible for the mobile endpoint device, e.g., an enterprise customer such as a business entity or a governmental agency, a guardian, a parent and so on. Alternatively, the notification is sent to an entity responsible for the mobile software application, e.g., a developer of the mobile software application.” An entity responsible for the device that is subject to an anomaly is identified.); and
transmits the data indicating at least one of the error, the root cause, or the corrective action to the entity (Bickford: para.0065 “In step 310, the processor receives a notification from a network service provider as to an anomaly associated with a particular mobile software application executing on a mobile endpoint device. For example, the network service provider has detected a mobile software application operating on the user's mobile endpoint device as causing an anomaly as discussed above. In one example, the notification is sent directly to the user operating the mobile endpoint device. Alternatively, the notification is sent to an entity responsible for the mobile endpoint device, e.g., an enterprise customer such as a business entity or a governmental agency, a guardian, a parent and so on. Alternatively, the notification is sent to an entity responsible for the mobile software application, e.g., a developer of the mobile software application.” The data of the anomaly is sent as a notification to the respective entities.).
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 Yokel with Bickford in order to incorporate the computing system determines an entity associated with the hardware component and transmits the data indicating at least one of the error, the root cause, or the corrective action to the entity.
One of ordinary skill in the art would have been motivated to combine because of the expected benefit of resolving the issue by transmitting a notification to related parties (Bickford: para.0065).
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
Barnes et al. US 2024/0362105 A1 para.0025 showing vector-based NLP for root cause analysis
Rajagopal US 2020/0226401 para.0011 artificial intelligent mode for root cause analysis, vector similarity.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to EUI H KIM whose telephone number is (571)272-8133. The examiner can normally be reached 7:30-5 M-R, M-F alternating.
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, Kamal B Divecha can be reached at 5712725863. 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.
/EUI H KIM/ Examiner, Art Unit 2453
/KAMAL B DIVECHA/ Supervisory Patent Examiner, Art Unit 2453
Prosecution Insights