DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
This office correspondence is in response to “Amendment and Response under 37 C.F.R. 1.111 filed on May 12, 2026 in response to a non-final office action issued on February 12, 2026.
Claims 1, 3 – 14, and 16 – 20 are pending.
Claims 1, 3 – 5, 8, 11, 14, and 16 – 18 are amended.
Claims 2 and 15 are cancelled.
Claims 1, 3 – 14, and 16 – 20 are rejected.
Applicant’s arguments filed on May 12, 2026 have been fully considered:
Response to Arguments
The applicant responds to the rejection of claims 1 – 20 as rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract idea without significantly more.
The applicant states:
“ . . . Under Alice/Mayo Step 2A, Prong 1, the claims are not properly characterized as a mental process or organizing human activity. The claims are directed to distributed, closed-loop control of intent knowledge in an intent-driven network management system. For example, claim 11 requires a knowledge consumption network element to determine performance evaluation information indicating a change value of a performance indicator of at least one target network device in a target time period, where the target network device executes the first intent operation and executes no other intent operation during that target time period. Claims 1, 8, and 14 likewise recite coordinated exchange of knowledge-quality information and policy information between knowledge provision and network management network elements, followed by updating of intent knowledge based on the received policy information.
These are not steps a human mind is practically equipped to perform in the claimed network-management environment, and the Office's characterization disregards the claimed machine-collected KPI change values, target network devices, target time periods, and policy- controlled updating loop. Moreover, assertion that the foregoing can simply be performed by the human mind disregards the stated benefits that the knowledge provision network element can feed back the quality of the intent knowledge to the network management network element, the network management network element can deliver a quality-based update policy, and the knowledge provision network element can then perform subsequent updating based on that policy "to improve quality of updated knowledge." See, e.g., Specification, paras. [0005]-[0007]. The foregoing also disregards benefits that the recited methods and apparatus can provide to overcoming problems associated with use of low-quality updated intent knowledge that can cause a system to execute an inappropriate intent operation in a subsequent intent execution procedure, as noted in the Specification at para. [0004]. Further, when a network device executes multiple intent operations in the same time period, a KPI change value may result from multiple operations, so intent knowledge generated for one intent operation may be affected by another operation and may cause inappropriate subsequent intent operations, as discussed in the Specification at para. [0091].
The specification further describes⁴ the benefit of the claimed policy-driven update loop in more detail. Policy information can indicate how to determine the target network device and target time period, improving flexibility of knowledge update. Updating based on performance evaluation information corresponding to a time period and network device that execute only the relevant intent operation improves accuracy of updated knowledge.⁵ These activities clearly meet the requirements set forth in SRI Int'l, Inc. V. Cisco Sys., Inc., 930 F.3d 1295, 1304 (Fed. Cir. 2019), as the level of collection and analysis from a practical standpoint far exceed the capability of the human mind, particularly in a network environment where a network device executes multiple intent operations in the same time period, a performance-indicator change value may reflect multiple operations, so intent knowledge for one operation may be affected by another operation and may cause an inappropriate subsequent intent operation.
The dependent claims further confirm the practical application and technological improvement. The Specification describes confidence information indicating a trustworthiness level of performance evaluation information, and explains that the knowledge provision network element may determine whether to update intent knowledge based on that confidence information or may weight historical intent knowledge and current performance evaluation information based on trustworthiness. 6 The resulting feedback loop allows the knowledge provision network element to feedback quality of intent knowledge, allows the network management network element to deliver a manner of updating based on that quality, and allows subsequent updating based on the policy information to improve quality of the updated knowledge. 7 These claimed operations are concrete controls over an intent-driven network-management technology, not insignificant extra- solution activity or a mere field-of-use limitation.
In view of the foregoing distinctions and as the claims as a whole recite a specific network- management improvement and integrate any alleged abstract idea into a practical application, the Applicant respectfully submits that claims 1-20 as presented herein recite patent-eligible subject matter and, as such, the rejection of claims 1-20 under 35 U.S.C. §101 should be withdrawn . . .” (Applicant’s remarks pages 12 – 13)
In response to the applicant’s argument:
The applicant’s argument that the claims provide significantly more than the abstract idea and herein recite patent-eligible subject matter under 35 USC 101 is persuasive. The applicant amended independent claims 1, 8, 11 and 14 to add more detail concerning how performance evaluation information is used to generate and update intent information. In this regard, it would not be reasonable to interpret the claims as a mental process or organizing human activity under Alice step 2A prong 1. The applicant amendments enable a showing to provide network management improvements that when integrated into the abstract idea creates a practical application for knowledge updates in a computer network, and enables the claims to recite patent eligible subject matter. Therein, the rejections under 35 U.S.C. 101 are withdrawn.
The applicant is advised that the claims as now constructed have introduced 35 U.S.C. 112(b) deficiencies and as such introduced rejections for said claims that are described below.
In regard to claims 1, 8, and 14 the applicant argues the prior art combination of Geddes and R fails to anticipate, disclose or teach the claim elements
The applicant states:
“ . . . Claims 1, 8, and 14 have been amended to clarify that the claimed update of intent
knowledge is based on performance evaluation information indicating a change value of a performance indicator of at least one target network device in a target time period, and that, during the target time period, the target network device executes the first intent operation and executes no other intent operation. These amendments are supported at least by the Specification's disclosure that performance evaluation information indicates a change value of a performance indicator of at least one target network device in a target time period, that the target network device executes the
first intent operation and does not execute an intent operation other than the first intent operation during that target time period, and that intent knowledge is updated based on the policy information and the performance evaluation information. . . . In view of the foregoing distinctions, the Applicant respectfully submits that each of claims 1, 8 and 14 is patentable over the proposed combination of Geddes in view of R and, as such, the 35 U.S.C. §103 rejection based thereon should be withdrawn. . . .” (Applicant’s remarks 14 – 17)
In response to the applicant’s argument:
The applicant’s amendments triggered a new search and consideration that found prior art reference (Shrivastava et al. U.S. 2019/0378048 to teach the elements added by the applicant’s amendments , and when combined with Geddes and R, teaches said claims. The applicant is referred to the claim rejections described below.
In regard to claims 2,3, 11, 15 and 16, the applicant argues the prior art combination of Geddes, R and Abdelkader fails to anticipate, disclose or teach the claim elements
The applicant states:
“ . . . As noted above, claims 2 and 15 have been cancelled. The rejection as it applies to remaining claims 3, 11 and 16 is respectfully traversed.
Claim 3 depends from claim 1 and incorporates all of the features recited therein. Likewise, claim 16 depends from independent claim 14. Patentability of independent claims 1 and 14 over the proposed combination of Geddes and R has been discussed at length above. The proposed addition of Abdelkader does not overcome the deficiencies of Geddes in combination with R. Instead, as concerns claims 3 and 16, Abdelkader was applied generally for its disclosure of a feedback report and "network control operations" to supplement policy information limitations for feedback reports, utility functions and network control operations for improving future intent fulfillment, as noted in paras. 15 and 18 of the Office Action. Thus, Abdelkader does not overcome the deficiencies noted above with respect to Geddes and R as concerns the failure of these references to disclose or suggest the feature of "obtaining, by the knowledge provision network element, performance evaluation information indicating a change value of a performance indicator of at least one target network device in a target time period, wherein the target network first intent operation during the target time period" and that the updating, by the knowledge provision network element, the intent knowledge based on the policy information is also based on "the performance evaluation information", as recited in independent claims 1 and 14. (Emphasis added). Accordingly, the 35 U.S.C. §103 based upon the combination of Geddes, R and Abdelkader should be withdrawn.
Claim 11 as amended herein is distinguishable over the proposed combination of references for at least the reason that the rejection does not disclose or suggest the recited aspect of policy- based determination of isolated performance evaluation information. In this regard, claim 11 recites the knowledge consumption network element receives policy information from the knowledge provision network element, determines performance evaluation information based on the policy information, and sends the performance evaluation information to the knowledge provision network element. The performance evaluation information must indicate a change value of a performance indicator of at least one target network device in a target time period, where the target network device executes the first intent operation and executes no other intent operation during the target time period.
The Office Action relies on Abdelkader's intent-fulfillment satisfaction workflow for this limitation. See Office Action para. 16. However, Abdelkader merely describes a consumer submitting an intent request, an IDNMS¹² fulfilling the intent, and the consumer calculating a satisfaction level after collecting data such as KPIs. See Abdelkader paras. [0063]-[0066]. Abdelkader does not disclose determining performance evaluation information for a target network device during a target time period selected so that the target network device executes only the first intent operation and no other intent operation. Nor does Abdelkader disclose that such performance evaluation information is determined based on policy information received from a knowledge provision network element. The proposed combination of Geddes, R and Abdelkader therefore fails to disclose or suggest at least the determination and transmission of performance evaluation information required by claim 11. In view of the foregoing amendments to claim 11 and related remarks, the 35 U.S.C. §103 rejection of claim 11 should be withdrawn. (applicant’s remarks pages 18-19)
In response to the applicant’s argument:
The applicant’s amendments triggered a new search and consideration that found prior art reference (Shrivastava et al. U.S. 2019/0378048 to teach the elements added by the applicant’s amendments , and when combined with Geddes R, and Abdelkader teaches said claims. The applicant is referred to the claim rejections described below.
In regard to claims 4, 9, 12,and 17, the applicant argues the prior art combination of Geddes, R, Abdelkader, and Desai fails to anticipate, disclose or teach the claim elements
The applicant states:
“ . . . Claim 4 depends ultimately from claim 1; claim 9 depends from independent claim 8; claim 12 depends from independent claim 11, and claim 17 depends ultimately from independent claim 14. As each of the foregoing independent claims is patentable over Geddes in view of R, and the addition of Abdelkader does not overcome the deficiencies of Geddes in combination with R, it stands to follow that Desai - applied for its disclosure of network health factors 13 - does not overcome the failure of Geddes in combination with R to disclose the recited feature of the claims of "obtaining, by the knowledge provision network element, performance evaluation information indicating a change value of a performance indicator of at least one target network device in a target time period, wherein the target network first intent operation during the target time period". (Emphasis added). For at least these reasons, each of claims 4, 9, 12 and 17 is patentable over the proposed combination of Geddes in view of R, Abdelkader and Desai, and the 35 U.S.C. § 103 rejection of the claims based thereon should be withdrawn. . . .” (applicant’s remarks page 19)
In response to the applicant’s argument:
The applicant’s amendments triggered a new search and consideration that found prior art reference (Shrivastava et al. U.S. 2019/0378048 to teach the elements added by the applicant’s amendments , and when combined with Geddes R, Abdelkader, and Desai teaches said claims. The applicant is referred to the claim rejections described below.
In regard to claims 5 – 7, 10, 13, and 18 – 20, the applicant argues the prior art combination of Geddes, R, Abdelkader, and Wang fails to anticipate, disclose or teach the claim elements
The applicant states:
“ . . . Wang was applied for its perceived disclosure of "confidence information " in conjunction with a confidence score relating to a conversational AI context for each potential intent and objective. This rejection is respectfully traversed, as Wang fails to cure the deficiencies noted in Sections A-C of this paper as concerns the base combination of Geddes in view of R. Moreover, the Applicant respectfully notes that Wang's application of a "confidence score" is in the context of conversational AI intent/objective interpretation, not a "trustworthiness level of performance evaluation information" derived from network-device KPI change values and used to gate receipt/use of that performance evaluation information and, as such, is not analogous to the subject matter of the claims.
Each of claims 5-7 ultimately depends from claim 1, which is patentable over the base combination of Geddes in view of R for all the reasons noted above with respect to the rejection of claim 1. Claim 10 depends ultimately from independent claim 8, which is patentable over the base combination of Geddes in view of R for all the reasons noted above with respect to the rejection of claim 8. The same holds true for claim 13 (which depends ultimately from independent claim 11) and claims 18-20, which depend ultimately from independent claim 14. Each of claims 5-7, 10, 13 and 18-20 is patentable over the proposed combination of Geddes in view of R, Abdelkader and Wang for at least their respective dependencies from an allowable base claim as well as recitation of features in addition to those recited in the base claims. For at least these reasons, the 35 U.S.C. §103 rejection of claims 5-7, 10, 13 and 18-20 should be withdrawn.
In response to the applicant’s argument:
The applicant’s amendments triggered a new search and consideration that found prior art reference (Shrivastava et al. U.S. 2019/0378048 to teach the elements added by the applicant’s amendments , and when combined with Geddes R, Abdelkader, and Wang teaches said claims. The applicant is referred to the claim rejections described below.
The examiner recommends that the applicant review the specification for disclosure that if integrated into the independent claims would distinguish the amended claims from the cited prior art. The applicant is invited to contact the examiner for an interview to discuss how to move the prosecution forward.
Authorization for Internet Communications
The examiner encourages Applicant to submit an authorization to communicate with the examiner via the Internet by making the following statement (from MPEP 502.03):
“Recognizing that Internet communications are not secure, I hereby authorize the USPTO to communicate with the undersigned and practitioners in accordance with 37 CFR 1.33 and 37 CFR 1.34 concerning any subject matter of this application by video conferencing, instant messaging, or electronic mail. I understand that a copy of these communications will be made of record in the application file.”
Please note that the above statement can only be submitted via Central Fax (not Examiner's Fax), Regular postal mail, or EFS Web using PTO/SB/439.
Priority
This Application is a continuation of International Application No. PCT/CN2023/080147 filed on Mar. 7, 2023, which claims priority to Chinese Patent Application No. 202210227347.4 filed on Mar. 8, 2022. The applicant is entitled to a priority date of 3/8/2022.
Claim Rejections - 35 USC § 112
The following is a quotation of 35 U.S.C. 112(b):
(b) CONCLUSION.—The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the inventor or a joint inventor regards as the invention.
The following is a quotation of 35 U.S.C. 112 (pre-AIA ), second paragraph:
The specification shall conclude with one or more claims particularly pointing out and distinctly claiming the subject matter which the applicant regards as his invention.
Claims 1,8, 11 and 14 are rejected under 35 U.S.C. 112(b) or 35 U.S.C. 112 (pre-AIA ), second paragraph, as being indefinite for failing to particularly point out and distinctly claim the subject matter which the inventor or a joint inventor (or for applications subject to pre-AIA 35 U.S.C. 112, the applicant), regards as the invention. The said claims are incomplete for omitting essential structural cooperative relationships of elements, such omission amounting to a gap between the necessary structural connections. See MPEP § 2172.01. The omitted structural cooperative relationships are: described in the instant specification that document and accompany Figure 1 (see USPGPUB 2025/0007792 A1; ¶¶ [0084-0091]) so that the structure and relationships of the three elements (Network Management element 120, Knowledge provision network element 110, and Knowledge consumption network element 130) can be given proper representation in the claim recitations. The current independent claims are overtly broad so it is difficult to find a proper bounds of the claim, and the claim reads on applications which exceed a cognitive network management system, which without a through reading of the specification that defines the direction of the claimed invention, a practitioner of the art would not be able to practice the invention as currently claimed.
Claims 3 – 7 which are dependent on claim 1 are also indefinite as they provide no additional structural guidance to the invention
Claims 9 – 10 which are dependent on claim 8 are also indefinite as they provide no additional structural guidance to the invention.
Claims 12 – 13 which are dependent on claim 11 are also indefinite as they provide no additional structural guidance to the invention.
Claims 16 – 20 which are dependent on claim 14 are also indefinite as they provide no additional structural guidance to the invention.
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 1, 8 and 14 are rejected under 35 U.S.C. 103 as being un-patentable over Geddes (U.S. 2004/0205182 A1; herein referred to as Geddes) in view of R et al. (U.S. 2021/0028980 A1; herein referred to as R) in further view of Shrivastava et al. (U.S. 2019/0378048 A1; herein referred to Shrivastava).
In regard to claim 1, Geddes teaches A knowledge update method (see abstract “ . . . A network management system and method is disclosed for providing an intelligent decision support system for assisting human operators monitor, maintain, and diagnose and solve problems with a network. According to one embodiment, a system and method is provided that uses a partial order planner and a knowledge base to implement intelligent decision support using knowledge stored as plan and goal graphs, concept graphs, and/or scripts . . .”) , comprising:
sending, by a knowledge provision network element (e. g. see Fig.2 inference engine 202) knowledge information to a network management network element (e.g. operator monitor 205) (see ¶¶ [0054-0055] “ . . . Inference engine 202 is also coupled to the network management collector 204. This component monitors the communications network using conventional technology such as by sending pings or SNMP requests to network devices. The data collected is incorporated into knowledge base 203 in order to assist the inference engine 202 in performing its functions. Finally, inference engine 202 is coupled to a graphical user interface 201 component that facilitates input and output with a network operator. The graphical user interface 201 output is dependent on the current state of the system as determined by the inference engine 202. . . .” see ¶ [0058] “ . . . The system remains in the all clear state until the network management collector 204 detects a change in the status of a device, or until the inference engine 202 detects a new intention of the network operator. For example, if the network operator begins viewing the firewall 103 log files and logs on to public servers 101, the inference engine may determine that the operator's activity has changed from nominal network operations to respond to security event. In that case, the graphical user interface 201 changes the display to show information that may help the network operator diagnose and resolve security problems . . .” ), wherein the knowledge information indicates quality of intent knowledge of a first intent operation (see Fig. 3 ¶ [0063] “ . . . One embodiment of the present invention is implemented using an inference engine 301 such as the one described in FIG. 3 including one or more planners 302, an intent interpreter 303, an information manager 304, a script performer 305, a knowledge base 306, and a situation assessor 307 . . .” see ¶ [0069] “ . . . to create an intelligent interface, the system monitors a network operator's actions to determine what the operator is trying to accomplish. The intent interpreter does this using a task-analytic decomposition of the purposes of network operators within the communications network domain. This decomposition is represented as a plan and goal graph (PGG), an acyclic, directed graph that represents the hierarchy of possible goals that may be pursued to achieve an intention and the methods (or plans) that can be used to satisfy each goal. Additionally, intent interpreter 303 uses knowledge represented as scripts. These scripts are sequences of primitive actions whose execution may be dependent on the state of the execution context. Other embodiments may use scripts that may include non-primitive actions (e.g., recursive script calls or additional script calls). Scripts represent standard procedures or processes that are routinely used to perform specific network activities described by plan sub-elements. Such standard operational procedures may include standard responses to both normal and abnormal events and operating conditions within a network. The intent interpreter 303 uses reasoning on the PGG to represent problem solving behaviors that are necessary when existing network management procedures defined by scripts are not appropriate for the situation. Using assertions made by the other components of the system together with domain knowledge stored in knowledge base 206, the intent interpreter determines the most likely intent of a network operator. This determined intent is then used to update the information being displayed to the network operator and to generate one or more plans to satisfy the interpreted goals of the operator. . . .”) ;
Geddes fails to explicitly teach,
However R teaches
receiving, by the knowledge provision network element (see Fig. 1 controller device 10) (see ¶ [0004] “ . . Network management systems (NMSs) and NMS devices, also referred to as controllers or controller devices, may support these services such that an administrator can easily create and manage these high-level network configuration services. . . .”), policy information from the network management network element, wherein the policy information indicates a manner of updating the intent knowledge (see ¶ ¶ [0005-0006] “ . . . In particular, user configuration of devices may be referred to as “intents.” An intent-based networking system lets administrators describe the intended network/compute/storage state. User intents can be categorized as business policies or stateless intents. Business policies, or stateful intents, may be resolved based on the current state of a network. Stateless intents may be fully declarative ways of describing an intended network/compute/storage state, without concern for a current network state. Intents may be represented as intent data models, which may be modeled using unified graphs. Intent data models may be represented as connected graphs, so that business policies can be implemented across intent data models. For example, data models may be represented using connected graphs having vertices connected with has-edges and reference (ref) edges. Controller devices may model intent data models as unified graphs, so that the intend models can be represented as connected. In this manner, business policies can be implemented across intent data models. When Intents are modeled using a unified graph model, extending new intent support needs to extend the graph model and compilation logic. ;
updating, by the knowledge provision network element (see Fig. 1 administrator 12), the intent knowledge based on the policy information (see ¶ ¶ [0028-0029] “ . . . Administrator 12 uses controller device 10 to configure elements 14 to specify certain operational characteristics that further the objectives of administrator 12. For example, administrator 12 may specify for an element 14 a particular operational policy regarding security, device accessibility, traffic engineering, quality of service (QoS), network address translation (NAT), packet filtering, packet forwarding, rate limiting, or other policies. Controller device 10 uses one or more network management protocols designed for management of configuration data within managed network elements 14, such as the SNMP protocol or the Network Configuration Protocol (NETCONF) protocol or a derivative thereof, such as the Juniper Device Management Interface, to perform the configuration. In general, NETCONF provides mechanisms for configuring network devices and uses an Extensible Markup Language (XML)-based data encoding for configuration data, which may include policy data . . . Controller device 10 may be configured to accept high-level configuration data, or intents, from administrator 12 (which may be expressed as structured input parameters . . . “) and the performance evaluation information (e.g. state of network) (see ¶ [0032] “ . . . customer environments are configured to allow customers (e.g., administrators 12) to control intent realization and assure programmed intents through controller device 10. Controller device 10 may maintain backups of device-level configuration information for elements 14 based on intents, such that controller device 10 may restore any of elements 14 to a previous state (e.g., after the network or device enters a bad state). For example, new intents provided by administrator 12 to controller device 10 may impact the functionality of one or more existing intents for one or more elements 14—putting the network in a bad state. To restore the network into a good state, controller device 10 may restore one or more elements 14 using backed-up intent configuration information. . . .”).
It would have been obvious to one with ordinary skill in the art before the effective filing date of the applicant’s application to incorporate systems and methods for the management of network devices using a controller device that maintains and updates knowledge information on the configuration of the network device and previous device-level intent configuration information, as taught by R, into systems and methods for providing intelligent decision support in network management by collecting data and incorporating the data into a knowledge base that can be used to infer the intent for a network device, as taught by Geddes. Such incorporation enables knowledge of the network to be updated from multiple sources and used to predict operational intent of the network.
The combination of Geddes and R fails to expclitly teach,
However Shrivastava teaches
obtaining, by the knowledge provision network element (see Fig. 1 ¶ [0027] “ . . . the performance predicting computing system 100 may allocate the training data 104 into intervals of a time series associated with the corresponding entity . . .”), performance evaluation information (see ¶ [0031] “ . . . the performance predicting computing system 100 can operate in a training stage to train a machine learning model using extracted entity metadata, generated query intents, and performance results pertaining to selected training entities. The trained machine learning model may then be used in a prediction stage to predict performance of a target entity using extracted entity metadata and generated query intents pertaining to that entity, thereby generating performance results for the target entity. It should be understood that even during the prediction stage, the training may continue using supplemental information about the training entities or feedback on the performance predictions of the target entity. . . .”) indicating a change value of a performance indicator of at least one target network device in a target time period (see Fig. 4 ¶ [0049] “ . . . During a training stage, a time-factored aggregator 406 receives the query intent time series 452 and time-dependent performance results for the training entities and allocates them in training intervals in the time series relative to a training milestone. For each training milestone, aggregated pre-milestone intents and aggregated post-milestone performance results are submitted to the machine learning model 408 to train the machine learning model 408. . . .”), wherein the target network device (e.g. target entity) executes the first intent operation and executes no other intent operation other than the first intent operation during the target time period (see ¶ [0050] “ . . . During a prediction stage, the time-factored aggregator 406 receives the query intent time series 452 for the target entity and allocates the query intents to target intervals in the time series relative to a target milestone. For each target milestone, aggregated pre-milestone intents are submitted to the machine learning model 408 to predict whether the target entity will satisfy a success condition based on its query intent time series. Accordingly, in the prediction stage, the machine learning model is being used to determine labels (e.g., “performance results” or the likelihood of achieving performance results that satisfy a success condition) for the pre-milestone observations (e.g., target query intents) of the target entity. The machine learning model 408 outputs a performance confidence score 418 representing the likelihood that the performance result is accurate or the likelihood of achieving performance results that satisfy a success condition . . .”;)
It would have been obvious to one with ordinary skill in the art before the effective filing date of the applicant’s application to incorporate systems and methods for generating performance predictions for a target entity using training intents for the target entity over a time interval, as taught by Shrivastava, into systems and methods for providing intelligent decision support in network management by collecting data and incorporating the data into a knowledge base that can be used to infer the intent for a network device, using a controller device that maintains and updates knowledge information on the configuration of the network device and previous device-level intent configuration information, as taught by the combination of Geddes and R. Such incorporation provides a means to predict performance of an entity over a time period.
In regard to claim 8, Geddes teaches A knowledge update method (see abstract as described for the rejection of claim 1 and is incorporated herein), comprising:
receiving, by a network management network element (e.g. operator monitor 205), knowledge information from a knowledge provision network element (e. g. see Fig.2 inference engine 202) (see ¶¶ [0054-0055] as described for the rejection of claim 1 and is incorporated herein) , wherein the knowledge information indicates quality of intent knowledge of a first intent operation (see Fig. 3 ¶ [0063] ¶ [0069] as described for the rejection of claim 1 and is incorporated herein) ;
Geddes fails to explicitly teach,
However R teaches and sending, by the network management network element (see ¶ [0004] as described for the rejection of claim 1 and is incorporated herein), policy information to the knowledge provision network element, wherein the policy information indicates a manner of updating the intent knowledge (see ¶ ¶ [0005-0006], ¶ ¶ [0028-0029] as described for the rejection of claim 1 and is incorporated herein)
The motivation to combine R with Geddes is described for the rejection of claim 1 and is incorporated herein.
The combination of Geddes and R fails to explicitly teach,
However Shrivastava teaches based on performance evaluation information (see Fig. 1 ¶ [0027], ¶ [0031] as described for the rejection of claim1 and is incorporated herein) indicating a change value of a performance indicator of at least one target network device in a target time period (see Fig. 4 ¶ [0049] as described for the rejection of claim1 and is incorporated herein), wherein the target network device (e.g. target entity) executes the first intent operation and executes no other intent operation other than the first intent operation during the target time period (see ¶ [0050] as described for the rejection of claim1 and is incorporated herein).
The motivation to combine Shrivastava with the combination of Geddes and R is described for the rejection of claim 1 and is incorporated herein.
In regard to claim 14, Geddes teaches A knowledge update method (see abstract as described for the rejection of claim 1 and is incorporated herein) , comprising:
sending, by a knowledge provision network element (e. g. see Fig.2 inference engine 202) , knowledge information to a network management network element (e.g. operator monitor 205) (see ¶¶ [0054-0055], ¶ [0058] as described for the rejection of claim 1 and is incorporated herein) and receiving, by the network management network element, the knowledge information from the knowledge provision network element, wherein the knowledge information indicates quality of intent knowledge of a first intent operation (see Fig. 3 ¶ [0063], ¶ [0069] as described for the rejection of claim 1 and is incorporated herein);
Geddes fails to explicitly teach,
However R teaches
sending, by the network management network element, policy information to the knowledge provision network element (see ¶ [0004] as described for the rejection of claim 1 and is incorporated herein) ;
receiving, by the knowledge prov1s1on network element, the policy information from the network management network element, wherein the policy information indicates a manner of updating the intent knowledge (see ¶ ¶ [0005-0006] as described for the rejection of claim 1 and is incorporated herein) ; and
updating, by the knowledge provision network element (see Fig. 1 administrator 12) , the intent knowledge based on the policy information (see ¶ ¶ [0028-0029] as described for the rejection of claim 1 and is incorporated herein) and the performance evaluation information (e.g. state of network) (see ¶ [0032] as described for the rejection of claim 1 and is incorporated herein)
The combination of Geddes and R fails to expclitly teach,
However Shrivastava teaches
obtaining, by the knowledge provision network element (see Fig. 1 ¶ [0027] as described for the rejection of claim 1 and is incorporated herein) , performance evaluation information (see ¶ [0031] as described for the rejection of claim 1 and is incorporated herein) indicating a change value of a performance indicator of at least one target network device in a target time period (see Fig. 4 ¶ [0049] as described for the rejection of claim 1 and is incorporated herein) , wherein the target network device (e.g. target entity) executes the first intent operation and executes no other intent operation other than the first intent operation during the target time period (see ¶ [0050] as described for the rejection of claim 1 and is incorporated herein)
The motivation to combine Shrivastava with the combination of Geddes and R is described for the rejection of claim 1 and is incorporated herein.
Claims 3, 11, and 16 are rejected under 35 U.S.C. 103 as being un-patentable over Geddes (U.S. 2004/0205182 A1; herein referred to as Geddes) in view of R et al. (U.S. 2021/0028980 A1; herein referred to as R) in further view of Shrivastava et al. (U.S. 2019/0378048 A1; herein referred to Shrivastava) as applied to claims 1, 8, and 14 in further view of Abdelkader et al. (U.S. 2024/0214287 A1; herein referred to as Adelkader).
In regard to claim 3, the combination of Geddes, R, and Shrivastava teaches wherein the policy information further indicates a manner of determining the at least one target network device and the target time period (see R ¶ [0028] “ . . . Administrator 12 uses controller device 10 to configure elements 14 to specify certain operational characteristics that further the objectives of administrator 12. For example, administrator 12 may specify for an element 14 a particular operational policy regarding security, device accessibility, traffic engineering, quality of service (QoS), network address translation (NAT), packet filtering, packet forwarding, rate limiting, or other policies. Controller device 10 uses one or more network management protocols designed for management of configuration data within managed network elements 14, such as the SNMP protocol or the Network Configuration Protocol (NETCONF) protocol or a derivative thereof, such as the Juniper Device Management Interface, to perform the configuration. In general, NETCONF provides mechanisms for configuring network devices and uses an Extensible Markup Language (XML)-based data encoding for configuration data, which may include policy data . . .”; see R ¶ [0049] “ . . . at time t2 containing one or more OOB configuration changes (“device config” in the expression for v.sub.2 in FIG. 3). In one or more aspects, the duration of time between each of time t.sub.0, t.sub.1, and t.sub.2 may be the same (or substantially the same) (e.g., 30 seconds, 5 minutes, 30 minutes, 2 hours, or any other interval of time). . . .”) ;
The combination of Geddes, R, and Shrivastava fails to expclitly teach,
However Abdelkader teaches the obtaining, by the knowledge provision network element, performance evaluation information comprises: determining, by the knowledge provision network element, the performance evaluation information (e.g. satisfaction) based on the policy information (e.g. network control operations) (see Abdelkader ¶ [0039] “ . . . According to at least some example embodiments, step S203 further includes providing, e.g. via interface 140 shown in FIG. 1, a second feedback report which comprises at least one of the following information: an indication of first network control operations to achieve a better or higher level of satisfaction with the fulfillment of the intent, a utility function used for calculating the level of satisfaction indicated by the measurement, and a result of an evaluation of second network control operations as to whether these will achieve a better or higher level of satisfaction with the fulfillment of the intent. . . “); or
sending, by the knowledge provision network element, the policy information (e.g. network control operations) to a knowledge consumption network element and receiving, by the knowledge provision network element, the performance evaluation information(e.g. satisfaction) from the knowledge consumption network element (see Abdelkader ¶ [0041] “ . . . a feedback report is evaluated. The feedback report comprises at least one of the following information: [0042] a measurement that indicates a level of satisfaction with a fulfillment of an intent which has been submitted to an intent-driven network management system, [0043] an identification of a consumer entity (e.g. consumer 120 of FIG. 1) that has submitted the intent, [0044] an identification of the intent for which the first feedback report is being provided, [0045] an identification of a service related to the intent, [0046] an indication of first network control operations to achieve a higher level of satisfaction with the fulfillment of the intent, [0047] a utility function used for calculating the level of satisfaction indicated by the measurement, [0048] a result of an evaluation of second network control operations as to whether these will achieve a higher level of satisfaction with the fulfillment of the intent. . . .”).
It would have been obvious to one with ordinary skill in the art before the effective filing date of the applicant’s application to incorporate systems and methods to evaluate in a Cognitive Autonomous Networks (CANs) in 5G networks, e.g. radio access networks, and other (e.g. future) generations of wireless/mobile networks and, specifically, the use of intents in managing networks. In particular, relate to intent performance of a device used by a consumer to measure satisfaction, as taught by Abdelkader, into systems and methods for providing intelligent decision support in network management by collecting data and incorporating the data into a knowledge base that can be used to infer the intent for a network device, using a controller device that maintains and updates knowledge information on the configuration of the network device and previous device-level intent configuration information, and performance over a time interval as taught by the combination of Geddes, R, and Shrivastava. Such incorporation provides a means to predict changes in intent and performance for a network device.
In regard to claim 11, Geddes teaches A knowledge update method (see abstract as described for the rejection of claim 1 and is incorporated herein) ,
Geddes fails to explicitly teach,
However R teaches comprising: receiving, by a knowledge consumption network element (see Fig. 1 controller device 10) (see ¶ [0004] “ . . Network management systems (NMSs) and NMS devices, also referred to as controllers or controller devices, may support these services such that an administrator can easily create and manage these high-level network configuration services. . . .”), policy information from a knowledge provision network element, wherein the policy information indicates a manner of updating intent knowledge (see R ¶ ¶ [0005-0006] as described for the rejection of claim 1 and is incorporated herein.) ;
The motivation to combine R with Geddes is described for the rejection of claim 1 and is incorporated herein.
The combination of Geddes and R fails to explicitly teach,
However Shrivastava teaches wherein the policy information further indicates a manner of determining at least one target network device (e.g. target entity) and a target time period (e.g. interval) (see ¶ [0003] “ . . . The described technology provides prediction of performance of a target entity using a machine learning model. Training query intents generated from a query-URL click graph are allocated for multiple training entities into training time intervals in a time series based on a corresponding query intent time for each training query intent. The training time intervals in the time series are separated by training milestones in the time series. Training performance results for the multiple training entities are allocated into the training time intervals in the time series based on a corresponding performance time of each training performance result. The machine learning model for a training milestone of the time series is trained based on the training query intents allocated to a training time interval prior to the training milestone and the training performance results allocated to a training time interval after the training milestone. Target query intents generated from the query-URL click graph for the target entity are allocated into target time intervals based on a corresponding query intent time for each target query intent. The target time intervals in the time series are separated by target milestones in the time series. A prediction of a target performance result for the target entity for an interval after a target milestone in the time series is generated by inputting to the trained machine learning model target query intents allocated to the target entity in a target time interval before the target milestone. . . . “) , and wherein the policy information indicates at least one of: a first policy in which a longest time period in which only the first intent operation is executed is selected as the target time period (see Fig. 4, ¶ [0045] “ . . . FIG. 4 illustrates a detailed schematic of a query intent generator 402 of an example performance prediction computing system 400. As previously discussed, in a training stage, a training query intent time series and time-dependent performance results for multiple training entities are input to a time-factored aggregator 406 and allocated into intervals relative to milestones in order to train a machine learning model 408. In a predicting stage, the query intent time series for the target entity is input to the time-factored aggregator 406, allocated into intervals relative to milestones, and input to the (trained) machine learning model 408 to predict whether the target entity will satisfy a success condition based on its query intent time series. FIG. 4 provides a more detailed description of a query intent generator 402 than provided in FIG. 3. . . .”) and a network device that executes only the first intent operation in the target time period is selected as the target network device(see Fig. 4 ¶ ¶ [0049-0050] “ . . . During a training stage, a time-factored aggregator 406 receives the query intent time series 452 and time-dependent performance results for the training entities and allocates them in training intervals in the time series relative to a training milestone. For each training milestone, aggregated pre-milestone intents and aggregated post-milestone performance results are submitted to the machine learning model 408 to train the machine learning model 408. . . (see ¶ [0050] “ . . . During a prediction stage, the time-factored aggregator 406 receives the query intent time series 452 for the target entity and allocates the query intents to target intervals in the time series relative to a target milestone. For each target milestone, aggregated pre-milestone intents are submitted to the machine learning model 408 to predict whether the target entity will satisfy a success condition based on its query intent time series. Accordingly, in the prediction stage, the machine learning model is being used to determine labels (e.g., “performance results” or the likelihood of achieving performance results that satisfy a success condition) for the pre-milestone observations (e.g., target query intents) of the target entity. The machine learning model 408 outputs a performance confidence score 418 representing the likelihood that the performance result is accurate or the likelihood of achieving performance results that satisfy a success condition . . .”;); or a second policy in which a time period in which there is a maximum quantity of network devices that execute only the first intent operation is selected as the target time period (see ¶ [0069] “ . . . wherein the generating operation includes generating a different prediction of a different target performance result for the target entity for a different target time interval after a different target milestone in the time series by inputting to the trained machine learning model target query intents for the target entity allocated to a different time interval before the different target milestone . . .”), and a network device that executes only the first intent operation in the target time period is selected as the target network device (see (see ¶ [0070] “ . . . An example computing device for predicting performance of a target entity includes a time-factored aggregator configured to allocate training query intents generated from a query-URL click graph for multiple training entities into training time intervals in a time series based on a corresponding query intent time for each training query intent. The training time intervals in the time series are separated by training milestones in the time series. The time-factored aggregator is also configured to allocate training performance results for the multiple training entities into the training time intervals in the time series based on a corresponding performance time of each training performance result. A machine learning model is configured to receive the allocated training query intents and allocated training performance results. The machine learning model is trained for one or more of the training milestones of the time series based on the training query intents allocated to a training time interval prior to the training milestone and the training performance results allocated to a training time interval after the training milestone. The machine learning model is further configured to generate a prediction of a target performance result for the target entity for an interval after a target milestone in the time series by inputting to the trained machine learning model target query intents allocated to the target entity in a target time interval before the target milestone . . .”) ;
The motivation to combine Shrivastava with the combination of Geddes and R is described for the motivation of claim 1 and is incorporated herein.
The combination of Geddes, R, and Shrivastava fails to explicitly teach However Adelkader teaches
determining, by the knowledge consumption network element, performance evaluation information (e.g. satisfaction) based on the policy information (see Abdelkader Fig. 3 ¶ [0064] “ . . . Upon receiving the intent fulfilment notification, the consumer 120 sends a data collection request to check the state of the intent objective (step S309). After gathering the relevant data, e.g. KPIs, parameters, etc. (step S311), the consumer 120 calculates its satisfaction level for the fulfillment of the intent by the IDNMS 110 (step S313). According to at least some example embodiments, this satisfaction calculation is performed based on a utility function which is at least one of consumer, intent and service specific, and takes into account relevant KPIs and aspects important for that specific consumer, intent, or service. . . “), wherein the determining the performance evaluation information based on the policy information comprises determining the at least one target network device (e.g. consumer entity) and the target time period (e.g. intent execution time)based on the policy information (see Abdelkader Fig. 3 ¶ [0063] “ . . . As seen in FIG. 3, consumer 120 of an intent service submits an intent request to IDNMS 110 through intent specification platform 111 and standardized interfaces 121 (step S301). The IDNMS 110 proceeds to fulfil the submitted intent through intent logic execution in intent fulfillment system 112 (step S303) and performs a list of actions affecting specific network resources 130 to achieve the objective of the intent (step S305). After the intent execution, the IDNMS 110 sends an intent fulfilment notification to the consumer 120 informing it that the requested intent has been fully executed (S307), or in other cases failed to execute, was partially executed, etc. . . .”)
wherein the performance evaluation information(e.g. satisfaction) (see Abdelkader ¶ [0026] “ . . . a range of the intent fulfilment satisfaction indicator indicates levels of satisfaction from the lowest to the highest. . .”) indicates a change value of a performance indicator of at least one target network device(see Abdelkader Fig. 1 consumer 120) in a target time period (see Abdelkader ¶ [0066] “ . . consumer 120 requests in step S301 “Increase capacity of X slice by 50%”. Utility function U=f(α.Math.T, β.Math.C, θ.Math.T_SI), where T is the total intent execution time, C is the relative increase in capacity after intent fulfillment, T_SI is the service interruption time, and α+β+θ=1, are weights representing interest of the consumer 120. The intent execution time and service interruption time are evaluated relative to a reference time (of e.g. 1 second), for example, to allow for comparability of the three components . . “) and executing the first intent operation and executing no other intent operation other than the first intent operation during the target time period (see Abdelkader Fig. 3 ¶ [0063] “ . . . As seen in FIG. 3, consumer 120 of an intent service submits an intent request to IDNMS 110 through intent specification platform 111 and standardized interfaces 121 (step S301). The IDNMS 110 proceeds to fulfil the submitted intent through intent logic execution in intent fulfillment system 112 (step S303) and performs a list of actions affecting specific network resources 130 to achieve the objective of the intent (step S305). After the intent execution, the IDNMS 110 sends an intent fulfilment notification to the consumer 120 informing it that the requested intent has been fully executed (S307), or in other cases failed to execute, was partially executed, etc. . . “); and
sending, by the knowledge consumption network element, the performance evaluation information to the knowledge provision network element (see Abdelkader ¶¶ [0041-0048] “ . . . a feedback report is evaluated. The feedback report comprises at least one of the following information: a measurement that indicates a level of satisfaction with a fulfillment of an intent which has been submitted to an intent-driven network management system, an identification of a consumer entity (e.g. consumer 120 of FIG. 1) that has submitted the intent, an identification of the intent for which the first feedback report is being provided, an identification of a service related to the intent, an indication of first network control operations to achieve a higher level of satisfaction with the fulfillment of the intent, a utility function used for calculating the level of satisfaction indicated by the measurement, a result of an evaluation of second network control operations as to whether these will achieve a higher level of satisfaction with the fulfillment of the intent. . . .”).
The motivation to combine Abdelkader with the combination of Geddes, R, and Shrivastava is described for the rejection of claim 3 and is incorporated herein.
In regard to claim 16, the combination of Geddes, R, Shrivastava and Abdelkader teaches wherein the policy information further indicates a manner of determining the at least one target network device and the target time period (see R¶ [0028], ¶ [0049] as described for the rejection of claim 3 and is incorporated herein) ; and
the obtaining, by the knowledge provision network element, performance evaluation information comprises: determining, by the knowledge provision network element, the performance evaluation information (e.g. satisfaction) based on the policy information (e.g. network control operations) (see Abdelkader ¶ [0039] as described for the rejection of claim 3 and is incorporated herein) ; or
sending, by the knowledge provision network element, the policy information (e.g. network control operations) to a knowledge consumption network element (see Abdelkader ¶ [0027]” . . . when the satisfaction is inadequate then the IDNMS 110 requests for further information and hints on how to improve the intent fulfillment. For example, the IDNMS 110 asks the consumer 120 to provide a list of pseudo-operations that are normally performed (e.g. legacy network control operations, which are also referred to as “first network control operation” in the following) to achieve the needed outcome and learns therefore to improve future outputs. . . .”) and receiving, by the knowledge consumption network element, the policy information from the knowledge provision network element (see Abdelkader ¶ [0031]” . . . , the IDNMS 110 provides information about the intent fulfillment with a request to the consumer 120 to provide guidance on how the fulfillment can be improved. For example, the IFS 112 shares a list of pseudo-operations (which are also referred to as “second network control operation” in the following) performed and asks the consumer 120 to review it and highlight possible wrong/non-optimal actions. . . “) ;
determining, by the knowledge consumption network element, the performance evaluation information (e.g. satisfaction) based on the policy information (see Abdelkader Fig. 3 ¶ [0064] as described for the rejection of claim 11 and is incorporated herein) ;
sending, by the knowledge consumption network element, the performance evaluation information to the knowledge provision network element (see Abdelkader ¶¶ [0041-0048] as described for the rejection of claim 11 and is incorporated herein) ; and
receiving, by the knowledge provision network element, the performance evaluation information from the knowledge consumption network element (see Abdelkader ¶ ¶ [0054 -0057] “. . . evaluating a second feedback report which comprises at least one of the following information: the indication of first network control operations to achieve the higher level of satisfaction with the fulfillment of the intent, the utility function used for calculating the level of satisfaction indicated by the measurement, the result of the evaluation of second network control operations as to whether these will achieve the higher level of satisfaction with the fulfillment of the intent. . . .”).
The motivation to combine Abdelkader with the combination of Geddes, R and Shrivastava is described for the rejection of claim 3 and is incorporated herein.
Claims 4, 9, 12, and 17 are rejected under 35 U.S.C. 103 as being un-patentable over Geddes (U.S. 2004/0205182 A1; herein referred to as Geddes) in view of R et al. (U.S. 2021/0028980 A1; herein referred to as R) in further view of Shrivastava et al. (U.S. 2019/0378048 A1; herein referred to Shrivastava) in further view of Abdelkader et al. (U.S. 2024/0214287 A1; herein referred to as Adelkader) as applied to claims 3, 11, and 16 in further view of Desai et al. (U.S. 2020/0162325 A1; herein referred to as Desai).
In regard to claim 4, the combination of Geddes, R, Shrivastava, Adelkader, and Desai teaches wherein the policy information comprises at least one of the following constraint conditions: a quantity of target network devices is greater than or equal to a first threshold; and a length of the target time period is greater than or equal to a second threshold (see Desai ¶ [0136] “ . . . Such network health factor(s) include, but are not limited to, (1) a percentage (which can be a configurable parameter determined based on experiments and/or empirical studies) of client association rejections for client devices (e.g., wireless end points 130A-F) that have attempted to associate with a network node/site over a given time frame (which can be a configurable parameter determined based on experiments and/or empirical studies); (2) a percentage (which can be a configurable parameter determined based on experiments and/or empirical studies) of access points (APs) down in a given site over a given period of time (which can be a configurable parameter determined based on experiments and/or empirical studies); (3) a number (which can be a configurable parameter determined based on experiments and/or empirical studies) of Dynamic Frequency Selection (DFS) Hits over a given time frame (which can be a configurable parameter determined based on experiments and/or empirical studies); (4) a number (which can be a configurable parameter determined based on experiments and/or empirical studies) of coverage holes over a given period of time (which can be a configurable parameter determined based on experiments and/or empirical studies); (5) a percentage (which can be a configurable parameter determined based on experiments and/or empirical studies) of capacity reduction over a given time period (which can be a configurable parameter determined based on experiments and/or empirical studies); (6) a percentage (which can be a configurable parameter determined based on experiments and/or empirical studies) of APs with Quality of Service Enhanced Basic Service Set (QBSS) higher than a threshold (which can be expressed as a percentage and is a configurable parameter determined based on experiments and/or empirical studies); (7) WAN link failures on a given site or list of sites over a period of time (which can be a configurable parameter determined based on experiments and/or empirical studies), etc. . . .”).
It would have been obvious to one with ordinary skill in the art before the effective filing date of the applicant’s application to incorporate systems and methods for automatic provisioning of network components through the use of intent -based networking that measure performance of the network based on the policies implemented, as taught by Desai, into systems and methods for providing intelligent decision support in network management by collecting data and incorporating the data into a knowledge base that can be used to infer the intent for a network device, using a controller device that maintains and updates knowledge information on the configuration of the network device and previous device-level intent configuration information, when deployed in 5G networks, e.g. radio access networks, and other (e.g. future) generations of wireless/mobile networks and, specifically, the use of intents in managing networks, and performance over a time interval as taught by the combination of Geddes, R, Shrivastava and Abdelkader. Such incorporation provides setting limits on the number of devices and time of implementation for gathering data for the knowledge base.
In regard to claim 9, the combination of Geddes, R, Shrivastava, Adelkader, and Desai teaches wherein the policy information comprises at least one of the following constraint conditions: a quantity of target network devices is greater than or equal to a first threshold; and a length of a target time period is greater than or equal to a second threshold (see Desai ¶ [0136] as described for the rejection of claim 4 and is incorporated herein).
The motivation to combine the references is described for the rejection of claim 4 and is incorporated herein.
In regard to claim 12, the combination of Geddes, R, Shrivastava, Adelkader, and Desai teaches wherein the policy information comprises at least one of the following constraint conditions: a quantity of target network devices is greater than or equal to a first threshold; and a length of the target time period is greater than or equal to a second threshold (see Desai ¶ [0136] as described for the rejection of claim 4 and is incorporated herein).
The motivation to combine the references is described for the rejection of claim 4 and is incorporated herein.
In regard to claim 17, the combination of Geddes, R, Shrivastava, Adelkader, and Desai teaches wherein the policy information comprises at least one of the following constraint conditions: a quantity of target network devices is greater than or equal to a first threshold; and a length of the target time period is greater than or equal to a second threshold (see Desai ¶ [0136] as described for the rejection of claim 4 and is incorporated herein).
The motivation to combine the references is described for the rejection of claim 4 and is incorporated herein.
Claims 5 – 7 10, 13, and 18 – 20 are rejected under 35 U.S.C. 103 as being un-patentable over Geddes (U.S. 2004/0205182 A1; herein referred to as Geddes) in view of R et al. (U.S. 2021/0028980 A1; herein referred to as R) in further view of Shrivastava et al. (U.S. 2019/0378048 A1; herein referred to Shrivastava) in further view of Abdelkader et al. (U.S. 2024/0214287 A1; herein referred to as Adelkader) as applied to claims 2 -3, 11, and 15 – 16 in further view of Wang (U.S. 2023/0245651 A1; herein referred to as Wang).
In regard to claim 5, the combination of Geddes, R, Shrivastava, Adelkader, and Wang teaches further comprising:
obtaining, by the knowledge provision network element, confidence information indicating a trustworthiness level of the performance evaluation information (see Wang ¶ [0399] “ . . . The AI system evaluates the user’s input against the existing intents and objectives in the OKB and assigns a confidence score to each potential intent and objective. This confidence score represents the AI system’s level of certainty that a particular intent or objective is the most relevant one for the user’s input. . . “) ; and
the updating, by the knowledge provision network element, the intent knowledge based on the policy information comprises: updating the intent knowledge based on the policy information and the confidence information (see Wang ¶ [0401] “ . . . The AI system then selects the intent and objective with the highest confidence score as the most relevant one and generates a response to the user. The confidence score can also be used to improve the accuracy of the AI system’s future predictions. . . .”; see Wang ¶ [0266] “ . . . The conversational AI agent’s objective is to maximize cumulative reward over time by learning an optimal policy, which is a mapping of states to actions that yield the highest expected reward. The reward system acts as a measure of the desirability of the agent’s actions and helps guide the agent’s learning process. . . “).
It would have been obvious to one with ordinary skill in the art before the effective filing date of the applicant’s application to incorporate systems and methods for using an AI system for enabling contextually relevant communications based on determinizing the most relevant intent for a device, as taught by Wang, into systems and methods for providing intelligent decision support in network management by collecting data and incorporating the data into a knowledge base that can be used to infer the intent for a network device, using a controller device that maintains and updates knowledge information on the configuration of the network device and previous device-level intent configuration information, when deployed in 5G networks, e.g. radio access networks, and other (e.g. future) generations of wireless/mobile networks and, specifically, the use of intents in managing networks, and performance over a time interval as taught by the combination of Geddes, R, Shrivastava and Abdelkader. Such incorporation Such incorporation enables a confidence score to be determined for the knowledge data gathered.
In regard to claim 6, the combination of Geddes, R, Shrivastava, Abdelkader, and Wang teaches wherein the obtaining, by the knowledge prov1s1on network element, confidence information comprises: receiving, by the knowledge provision network element, the confidence information from the knowledge consumption network element (see Wang ¶ [0404] “ . . The AI systems can use various types of contextual information to generate confidence scores, such as user profile data, historical behavior, location, time of day, weather, and device type. User profile data can include age, gender, occupation, and interests, while historical behavior can include search and purchase history. Location can be determined by GPS or IP address, while weather data can also be used. Time of day and device type can also be considered . . .”).; and
the obtaining, by the knowledge provision network element, performance evaluation information comprises: receiving, by the knowledge provision network element, the performance evaluation information from the knowledge consumption network element when the trustworthiness level is greater than or equal to a third threshold (e.g. objective function) ( see Wang ¶¶ [0268-0269] “ . . An objective function 219 in the AI system 200 is a mathematical representation of the AI system’s goal or the desired outcome it aims to achieve. The objective function quantifies the performance of the AI system by assigning a numerical value to its current state, taking into consideration various factors such as accuracy, efficiency, and other performance metrics. In the context of ML and optimization, the objective function plays an important role in guiding the AI system during training or decision-making processes. The AI system’s goal is to either minimize or maximize the objective function, depending on the specific problem being addressed. For example, in a supervised learning task like regression or classification, the objective function is often a loss function that measures the difference between the predicted output and the actual target values. The AI system’s goal would be to minimize this loss function, thereby improving the accuracy of its predictions . . .”)
The motivation to combine Wang with the combination of Geddes, R, Shrivastava and Abdelkader is described for the rejection of claim 5 and is incorporated herein. Additionally, Wang enables confidence and performance parameters for determining the accuracy of the knowledge being gathered.
In regard to claim 7, the combination of Geddes, R, Shrivastava Abdelkader, and Wang teaches further comprising: receiving, by the knowledge provision network element, first information (e.g. contextual information) from the network management network element (see Wang ¶ [0005] “ . . . The most relevant contextual information to the user is predicted by the AI system. The AI system then transforms the most relevant contextual information into textual form and predicts a set of intents and objectives for user-centered interaction. . . .”) , wherein the first information indicates a manner of determining the confidence information (see Wang ¶ [0416] “ . . . the AI system determines whether any additional information is needed 1904. If the AI system determines that the available contextual information is insufficient or the AI system is unable to determine the user’s intent and objective with a reasonable level of confidence, it may request additional information again or provide alternative options for the user to choose from. . . .”) ; and
the obtaining, by the knowledge provision network element, confidence information comprises: determining, by the knowledge provision network element, the confidence information based on the first information (see Wang ¶ [0432] “ . . . the AI agent would use contextual information to generate a confidence score for the user’s intent and objective. . .”) ; or
sending, by the knowledge prov1s1on network element, the first information to the knowledge consumption network element and receiving, by the knowledge provision network element, the confidence information from the knowledge consumption network element (see Wang ¶ [0406] “ . . . the AI system is using contextual information to make an educated guess about the user’s intent or objective, which can be represented as a confidence score to guide the system’s actions. . . “).
The motivation to combine Wang with the combination of Geddes, R, Shrivastava and Abdelkader is described for the rejection of claim 5 and is incorporated herein. Additionally, Wang uses confidence scores to determine whether to update a knowledge base.
In regard to claim 10, the combination of Geddes, R, Shrivastava ,Abdelkader, and Wang teaches further comprising:
sending, by the network management network element (see Wang ¶ [0005] “ . . . The most relevant contextual information to the user is predicted by the AI system. The AI system then transforms the most relevant contextual information into textual form and predicts a set of intents and objectives for user-centered interaction. . . .”), first information to the knowledge provision network element, wherein the first information (e.g. contextual information) indicates a manner of determining confidence information (see Wang ¶ [0416] “ . . . the AI system determines whether any additional information is needed 1904. If the AI system determines that the available contextual information is insufficient or the AI system is unable to determine the user’s intent and objective with a reasonable level of confidence, it may request additional information again or provide alternative options for the user to choose from. . . .”) indicating a trustworthiness level of performance evaluation information (see Wang ¶ [0399] “ . . . The AI system evaluates the user’s input against the existing intents and objectives in the OKB and assigns a confidence score to each potential intent and objective. This confidence score represents the AI system’s level of certainty that a particular intent or objective is the most relevant one for the user’s input. . . “) , the performance evaluation information indicating a change value of a performance indicator of at least one target network device in the target time period (see R ¶ [0028] “ . . . Administrator 12 uses controller device 10 to configure elements 14 to specify certain operational characteristics that further the objectives of administrator 12. For example, administrator 12 may specify for an element 14 a particular operational policy regarding security, device accessibility, traffic engineering, quality of service (QoS), network address translation (NAT), packet filtering, packet forwarding, rate limiting, or other policies. Controller device 10 uses one or more network management protocols designed for management of configuration data within managed network elements 14, such as the SNMP protocol or the Network Configuration Protocol (NETCONF) protocol or a derivative thereof, such as the Juniper Device Management Interface, to perform the configuration. In general, NETCONF provides mechanisms for configuring network devices and uses an Extensible Markup Language (XML)-based data encoding for configuration data, which may include policy data . . .”; see R ¶ [0049] “ . . . at time t2 containing one or more OOB configuration changes (“device config” in the expression for v.sub.2 in FIG. 3). In one or more aspects, the duration of time between each of time t.sub.0, t.sub.1, and t.sub.2 may be the same (or substantially the same) (e.g., 30 seconds, 5 minutes, 30 minutes, 2 hours, or any other interval of time). . . .”) and executing the first intent operation and executing no other intent operation other than the first intent operation during the target time period (see Abdelkader Fig. 3 ¶ [0063] “ . . . As seen in FIG. 3, consumer 120 of an intent service submits an intent request to IDNMS 110 through intent specification platform 111 and standardized interfaces 121 (step S301). The IDNMS 110 proceeds to fulfil the submitted intent through intent logic execution in intent fulfillment system 112 (step S303) and performs a list of actions affecting specific network resources 130 to achieve the objective of the intent (step S305). After the intent execution, the IDNMS 110 sends an intent fulfilment notification to the consumer 120 informing it that the requested intent has been fully executed (S307), or in other cases failed to execute, was partially executed, etc. . . “).
The motivation to combine the references is described for the rejections of claim 2 and claim 5 and is incorporated herein.
In regard to claim 13, the combination of Geddes, R, Shrivastava, Abdelkader, and Wang teaches further comprising:
receiving, by the knowledge consumption network element, first information from the knowledge provision network element, wherein the first information (e.g. contextual information) indicates a manner of determining confidence information (see Wang ¶ [0416] “ . . . the AI system determines whether any additional information is needed 1904. If the AI system determines that the available contextual information is insufficient or the AI system is unable to determine the user’s intent and objective with a reasonable level of confidence, it may request additional information again or provide alternative options for the user to choose from. . . .”) indicating a trustworthiness level of the performance evaluation information (see Wang ¶ [0399] “ . . . The AI system evaluates the user’s input against the existing intents and objectives in the OKB and assigns a confidence score to each potential intent and objective. This confidence score represents the AI system’s level of certainty that a particular intent or objective is the most relevant one for the user’s input. . . “);
determining, by the knowledge consumption network element, the confidence information based on the first information; (see Wang ¶ [0432] “ . . . the AI agent would use contextual information to generate a confidence score for the user’s intent and objective. . .”) and
sending, by the knowledge consumption network element, the confidence information to the knowledge provision network element (see Wang ¶ [0406] “ . . . the AI system is using contextual information to make an educated guess about the user’s intent or objective, which can be represented as a confidence score to guide the system’s actions. . . “).
The motivation to combine the references is described for the rejection of claim 5 and is incorporated herein.
In regard to claim 18, the combination of Geddes, R, Shrivastava, Adelkader, and Wang teaches further comprising:
obtaining, by the knowledge provision network element, confidence information indicating a trustworthiness level of the performance evaluation information (see Wang ¶ [0399] as described for the rejection of claim 5 and is incorporated herein); and
the updating, by the knowledge provision network element, the intent knowledge based on the policy information comprises: updating, by the knowledge provision network element, the intent knowledge based on the policy information and the confidence information (see Wang ¶ [0401], ¶ [0266] as described for the rejection of claim 5 and is incorporated herein).
The motivation to combine the references is described for the rejection of claim 5 and is incorporated herein.
In regard to claim 19, the combination of Geddes, R, Shrivastava, Abdelkader, and Wang teaches wherein the obtaining, by the knowledge provision network element, confidence information comprises:
sending, by the knowledge consumption network element, the confidence information to the knowledge provision network element and receiving, by the knowledge provision network element, the confidence information from the knowledge consumption network element (see Wang ¶ [0404] as described for the rejection of claim 6 and is incorporated herein) ; and
the obtaining, by the knowledge provision network element, performance evaluation information comprises: receiving, by the knowledge prov1s1on network element, the performance evaluation information from the knowledge consumption network element when the trustworthiness level is greater than or equal to a third threshold (e.g. objective function) ( see Wang ¶¶ [0268-0269] as described for the rejection of claim 6 and is incorporated herein).
The motivation to combine the references is described for the rejection of claim 6 and is incorporated herein.
In regard to claim 20, the combination of Geddes, R, Shrivastava, Abdelkader, and Wang teaches further comprising: sending, by the network management network element, first information(e.g. contextual information) to the knowledge provision network element and receiving, by the knowledge provision network element, the first information from the network management network element (see Wang ¶ [0005] as described for the rejection of claim 7 and is incorporated herein) , wherein the first information indicates a manner of determining the confidence information (see Wang ¶ [0416] as described for the rejection of claim 7 and is incorporated herein); and
the obtaining, by the knowledge provision network element, confidence information comprises: determining, by the knowledge provision network element, the confidence information based on the first information (see Wang ¶ [0432] “ . . . the AI agent would use contextual information to generate a confidence score for the user’s intent and objective. . .”) ; or
sending, by the knowledge prov1s1on network element, the first information (e.g. contextual information) to the knowledge consumption network element (see Wang ¶ [0005] as described for the rejection of claim 7 and is incorporated herein) ;
determining, by the knowledge consumption network element, the confidence information based on the first information (see Wang ¶ [0404] “ . . . The AI systems can use various types of contextual information to generate confidence scores, such as user profile data, historical behavior, location, time of day, weather, and device type. User profile data can include age, gender, occupation, and interests, while historical behavior can include search and purchase history. Location can be determined by GPS or IP address, while weather data can also be used. Time of day and device type can also be considered . . .”) ;
sending, by the knowledge consumption network element, the confidence information to the knowledge provision network element (see Wang ¶ [0401] “ . . . The AI system then selects the intent and objective with the highest confidence score as the most relevant one and generates a response to the user. The confidence score can also be used to improve the accuracy of the AI system’s future predictions. . . .”) ; and
receiving, by the knowledge provision network element, the confidence information from the knowledge consumption network element (see Wang ¶ [0406] “ . . . the AI system is using contextual information to make an educated guess about the user’s intent or objective, which can be represented as a confidence score to guide the system’s actions. . . “).
The motivation to combine the references is described for the rejection of claim 7 and is incorporated herein.
Conclusion
There are prior art made of record which are not relied upon but are considered pertinent to applicant’s disclosure. They are listed on the PTO-892 accompanying this action
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.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JAMES N FIORILLO whose telephone number is (571)272-9909. The examiner can normally be reached on 7:30 - 5 PM Mon - Fri..
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, John A. Follansbee can be reached on 571-272-3964. 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). If you would like assistance from a USPTO Customer Service Representative or access to the automated information system, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/JAMES N FIORILLO/Primary Examiner, Art Unit 2444
Prosecution Insights