DETAILED ACTION
This office action is in response to Applicant’s Amended submission filed on 5/4/2026. Applicant has amended claims 1-2, 6, 8-9, 13, and 15-16. As such, Claims 1-20 are pending and have been examined.
Response to Amendment and Arguments
35 U.S.C. 101 Rejections
Applicant’s amendment and remarks toward the rejection has been fully reconsidered, and is persuasive, therefore the rejection has been withdrawn.
35 U.S.C. 102/103 Rejections
Applicant’s arguments are moot in view of the new or modified grounds of rejection that
were necessitated by the amendments to the Claims. Applicant’s arguments are directed to material that is added by the most recent amendments to the independent Claims. Response, p. 10.
Examiners Note
Examiner noted claim 20, although in different statutory category, which previously tracks to claims 6 and 13, however, the amendment was not applied to claim 20, which appears to be typo/oversight, please double check the claim and make the necessary change if necessary.
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) 1-6, 8-13 and 15-20 are rejected under 35 U.S.C. 103 as being unpatentable over Muralidharan, in view of Dunn (US 20200394360).
Regarding Claim 1, Muralidharan discloses: 1. An apparatus comprising one or more processors and one or more storage devices storing instructions that are operable, when executed by the one or more processors, to cause the one or more processors to:
extract a feature set from a plurality of service message data objects associated with an application framework; ([0003] In accordance with one aspect, a computer-implemented method is provided. In one embodiment, the computer-implemented method comprises: determining, based on one or more natural language data fields of a software incident data object for a software application framework and using a natural language feature extraction machine learning model, a natural language feature data object for the software incident data object;) also see para 0039. [software incident data object is a type of service message data object]
apply a supervised natural language processing model to the feature set to generate a plurality of classification data objects associated with the plurality of service message data objects that classify a respective service message data object as belonging to a predefined class of a plurality of predefined classes, ([0039] the predicted incident severity level is a discrete classification output that is selected from a set of candidate classes (e.g., a set comprising a low predicted incident severity level class, a medium predicted incident severity level class, and a high predicted incident severity level class). In some embodiments, the predicted incident severity level is a continuous regression output (e.g., a value selected from the range 0-1). In some embodiments, the predicted incident severity level is a multi-valued output describing, for each candidate class of a set of candidate classes, a likelihood that the corresponding software incident data object is associated with the candidate class. [0040] the incident severity level detection machine learning model is a supervised trained machine learning model (e.g., a fully-connected supervised trained machine learning model).)
wherein the predefined class is representative of a natural language processing contextual description associated with the respective service message data object; ([0049] The natural language processing unit 112 may be configured to determine, based on one or more natural language data fields of a software incident data object for the software application framework and using a natural language processing machine learning framework, a natural language feature data object for the software incident data object. [0111] the incident analysis features describe one or more analysis features for a software incident data object, including a root cause category, a feature describing whether the incident was detected by monitoring, and/or the like. In some embodiments, the root cause category feature may take one of the following values: Change, Dependency, Scale, Architecture, Bug, Unknown.) [root cause category reads/or aligns with predefined reason class] Also see para 0038 and 0107.
initiate a rendering of a dashboard visualization via an electronic interface based at least in part on the plurality of classification data objects ([0128] At operation 504, the software monitoring data management computing device 106 performs one or more prediction-based actions based on the one or more incident signatures. For example, performing the one or more prediction-based actions may include enabling display of a prediction output user interface that displays the one or more incident signatures. As another example, performing the one or more prediction-based actions may include enabling display of a prediction output user interface that is configured to receive one or more user feedback data objects for the one or more incident signatures.) Also see figs. 6, 9 and 12 which shows data visualization based on classification data objects.
Although Muralidharan discloses incident review data object is used to retrain the natural language processing machine learning framework, however it does not explicitly disclose adjusting one or more parameters based on evaluated performance.
Dunn discloses: evaluate performance of the supervised natural language processing model using one or more performance metrics; ([0086] As described herein, multiple different neural networks can be used in the course of a conversation (e.g. multiple back and forth communications between a user and a system), and data for such communications can be used in machine learning operations to update the neural networks or other systems used for future interactions with users and operations to associate intent categories and actions with words from a user communication. Usage data by users can be used to adjust weights in a neural network to improve intent category assignments and track changes in user intent trends (e.g. final user intent results identified at the end of a user conversation with a system as compared with assigned intents based on initial user communications). Data generated by intent management engine 615 can be stored with associated message data in message data store 620, and this data can be used for various updates, including managing data for continuous real-time analysis updates or other dynamic feedback and modifications to a system,)
and adjust one or more parameters of the supervised natural language processing model based at least in part on the evaluated performance. ([0086] As described herein, multiple different neural networks can be used in the course of a conversation (e.g. multiple back and forth communications between a user and a system), and data for such communications can be used in machine learning operations to update the neural networks or other systems used for future interactions with users and operations to associate intent categories and actions with words from a user communication. Usage data by users can be used to adjust weights in a neural network to improve intent category assignments and track changes in user intent trends (e.g. final user intent results identified at the end of a user conversation with a system as compared with assigned intents based on initial user communications). Data generated by intent management engine 615 can be stored with associated message data in message data store 620, and this data can be used for various updates, including managing data for continuous real-time analysis updates or other dynamic feedback and modifications to a system,)
Muralidharan and Dunn are considered analogous art. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Muralidharan to combine the teaching of Dunn, because the technique described would improve intent category assignments and track changes in user intent trends (Dunn, [0086]).
Regarding Claim 2, Muralidharan and Dunn discloses: All the elements of claim 1,
Muralidharan further discloses: wherein the plurality of service message data objects respectively comprise at least a description data field associated with a service request by a user identifier, and wherein the one or more storage devices store instructions are operable, when executed by the one or more processors, to further cause the one or more processors to: extract the feature set from the plurality of service message data objects by extracting the description data field from the respective service message data object. ([0005] In accordance with yet another aspect, an apparatus comprising at least one processor and at least one memory including computer program code is provided. In one embodiment, the at least one memory and the computer program code may be configured to, with the processor, cause the apparatus to: determine, based on one or more natural language data fields of a software incident data object for a software application framework and using a natural language feature extraction machine learning model, a natural language feature data object for the software incident data object; determine, based on one or more structured data fields of the software incident data object and using a structured data feature extraction machine learning model, a structured data feature data object for the software incident data object;) [natural language data field functions as a description)]. Para 0115 also discloses a description data field.
Regarding Claim 3, Muralidharan and Dunn discloses: All the elements of claim 1,
Muralidharan further discloses: wherein the predefined class is representative of a reason for making a service request. ([0111] In some embodiments, the incident analysis features describe one or more analysis features for a software incident data object, including a root cause category, a feature describing whether the incident was detected by monitoring, and/or the like. In some embodiments, the root cause category feature may take one of the following values: Change, Dependency, Scale, Architecture, Bug, Unknown.) [Root cause category reads on type of predefined class that is representative of a reason for making a service request]
Regarding Claim 4, Muralidharan and Dunn discloses: All the elements of claim 1,
Muralidharan further discloses: wherein the dashboard visualization comprises at least one module, and wherein the at least one module is configured to display a predetermined format for displaying data based on the plurality of classification data objects. ([0079] FIG. 6 depicts an operational example of a software monitoring data display user interface 600 that includes user interface segments 611-615 that each describe various properties of a corresponding software incident data object. As further depicted in FIG. 6, because of the selection of the user interface element 601, the software monitoring data display user interface 600 currently displays information about “open” (i.e., unresolved) software incident data objects. User interface segments 611-615 are described in greater detail below, in relation to describing exemplary data fields of a software incident data object.)
Regarding Claim 5, Muralidharan and Dunn discloses: all of claim 1.
Dunn further discloses: wherein the dashboard visualization comprises a module configured to display a proportion of service message data objects associated with a respective classification. ([0131] FIGS. 11A-11F show examples of aspects of dashboard reports… For example, FIG. 11A can include data that illustrates summary information about a number of conversations, an average duration of a communication session, an intent score, intents by conversations, intent trends, intent durations, or other such information as data 1114 using data summary interface 1110, data metric 1108 graphics, and chart 1116 data. These analytics can become available as users contact the intent-driven contact center and their intents are ascertained. Interface elements for adding filters, selecting filters, and setting value ranges for filters can be used to select the displayed data using add filter element 1102, filter type selections 1104, filter value selections 1106 and other such user interface selections. For example, in the illustrated filter interface 1100, data summary interfaces 1110 can show different summary data types in associated data 1114 areas of each interface area, along with an associated trend arrow 1112 for each interface. Chart 1116 can show a volume of intent category assignments over time. For example, each line of chart 1116 can indicate a volume of intent assignments in a given hour of a day for a given geographic area (e.g. where intent assignments drop to near zero in the middle of the night). Each intent category can have a one or more associated data summary interface 1110 that shows statistical values about the intent category, such as a daily average volume, a weekly average volume, a weekly average trend (e.g. change) over the past year, or any other such metrics.)
Muralidharan and Dunn are considered analogous art. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Muralidharan to combine the teaching of Dunn, because examples described herein improve the operation of devices in a communication system by improving the efficiency of AI and machine based communications, reducing the processing resources used to facilitate responses to user communications, and improving the quality of machine driven communications in such a system (Dunn, [0007]).
Regarding Claim 6, Muralidharan and Dunn discloses: All the elements of claim 1,
Dunn further discloses: wherein the one or more storage devices store instructions are operable, when executed by the one or more processors, to further cause the one or more processors to: evaluate the performance of the supervised natural language processing model using one or more performance metrics at a predetermined time interval. ([0086] As described herein, multiple different neural networks can be used in the course of a conversation (e.g. multiple back and forth communications between a user and a system), and data for such communications can be used in machine learning operations to update the neural networks or other systems used for future interactions with users and operations to associate intent categories and actions with words from a user communication. Usage data by users can be used to adjust weights in a neural network to improve intent category assignments and track changes in user intent trends (e.g. final user intent results identified at the end of a user conversation with a system as compared with assigned intents based on initial user communications). Data generated by intent management engine 615 can be stored with associated message data in message data store 620, and this data can be used for various updates, including managing data for continuous real-time analysis updates or other dynamic feedback and modifications to a system,)
The rationale for the combination would be similar to the one already provided in claim 1.
Claim 8 recites a computer-implemented method claim that corresponds to the apparatus of claim 1 and is therefore rejected under the same grounds as claim 1 above.
Claims 9-11 are computer-implemented method claims that corresponds to claims 2-4 and therefore are also rejected under same grounds as claims 2-4.
Claims 12 and 13 are computer-implemented method claim with limitation similar to the limitations of Claims 5 and 6 and are rejected under similar rationale.
Regarding Claim 15, Muralidharan discloses: 15. A computer program product comprising at least one non-transitory computer readable storage medium having computer executable code portions stored therein, the computer executable code portions comprising program code instructions configured to: ([0004] In accordance with another aspect, a computer program product is provided. The computer program product may comprise at least one computer-readable storage medium having computer-readable program code portions stored therein, the computer-readable program code portions comprising executable portions configured to:)
As for the rest of the claim, they recite similar elements of claim 1. Thus, the rejection applied to claim 1 is equally applicable.
Claims 16-18 are computer program product claims that corresponds to claims 2-4 and therefore are also rejected under same grounds as claims 2-4.
Claim 19 is computer program product claim with limitation similar to the limitations of Claim 5 and is rejected under similar rationale.
Regarding Claim 20, Muralidharan and Dunn discloses: All the elements of claim 15,
Dunn further discloses: the program code instructions further configured to: evaluate the performance of the supervised natural language processing model using one or more performance metrics at a predetermined time interval; ([0086] As described herein, multiple different neural networks can be used in the course of a conversation (e.g. multiple back and forth communications between a user and a system), and data for such communications can be used in machine learning operations to update the neural networks or other systems used for future interactions with users and operations to associate intent categories and actions with words from a user communication. Usage data by users can be used to adjust weights in a neural network to improve intent category assignments and track changes in user intent trends (e.g. final user intent results identified at the end of a user conversation with a system as compared with assigned intents based on initial user communications). Data generated by intent management engine 615 can be stored with associated message data in message data store 620, and this data can be used for various updates, including managing data for continuous real-time analysis updates or other dynamic feedback and modifications to a system,)
and adjust one or more parameters of the supervised natural language processing model based on the one or more performance metrics. ([0086] As described herein, multiple different neural networks can be used in the course of a conversation (e.g. multiple back and forth communications between a user and a system), and data for such communications can be used in machine learning operations to update the neural networks or other systems used for future interactions with users and operations to associate intent categories and actions with words from a user communication. Usage data by users can be used to adjust weights in a neural network to improve intent category assignments and track changes in user intent trends (e.g. final user intent results identified at the end of a user conversation with a system as compared with assigned intents based on initial user communications). Data generated by intent management engine 615 can be stored with associated message data in message data store 620, and this data can be used for various updates, including managing data for continuous real-time analysis updates or other dynamic feedback and modifications to a system,)
The rationale for the combination would be similar to the one already provided in claim 1.
Claims 7 and 14 are rejected under 35 U.S.C. 103 as being unpatentable over Muralidharan in view of Dunn, and further in view of Ni (US 20220308952).
Regarding Claim 7, Muralidharan and Dunn discloses: all of claim 1.
Although Muralidharan teaches using a BERT model, and teaches multi-category, it does not explicitly disclose the BERT model is fine-tuned for multi-class text classification.
Ni discloses: wherein the supervised natural language processing model is a bidirectional transformer model that is fine-tuned for multi-class text classification. ([0069] A log segment representation model is built by BERT that learns a feature representation from log pattern ID sequences (e.g., each log line in an obtained log file may be translated to a log pattern ID) in a pre-training manner by a Masked Language Model (MLM). The BERT-based log segment representation is used as a feature representation for training a multi-class text classifier from labeled log segments (e.g., with each log segment labeled by an issue as its class) in a fine-tuning manner.)
Muralidharan, Dunn and Ni are considered analogous art. Therefore, it would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to modify the teachings of Muralidharan and Dunn to combine the teaching of Ni, because machine learning model can analyze the log files automatically, the efficiency of a support or monitoring and analytics platform (or technical support engineers thereof) may be significantly improved. (Ni, [0067]).
Claim 14 is computer-implemented method claim with limitation similar to the limitations of Claim 7 and are rejected under similar rationale.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure: Kalia (US 20200067798) – discloses “In accordance with the system 100, a memory 104 can store computer executable components executable by the processor 102. The ingestion component 108 can receive service or change tickets and determine automation opportunities for a service catalog. The categorization component 110 can determine domain specific clusters and categories of the service or change tickets. The categorization component employs a frequency-based approach that determines frequency of verbs and nouns associated with a subset of the service or change tickets. This frequency-based approach finds relevant verb and noun pairs by analyzing deep parsed graphs generated for each ticket, and wherein the valid verb and noun pairs represent valid actions. The categorization component can also employ a topic modeling approach that utilizes Latent Dirichlet allocation and non-negative matrix factorization to analyze a subset of the service or change tickets by generating clusters. Within these embodiments, the categorization component may also employ deep learning to analyze a subset of the service or change tickets by generating clusters." See Abstract, para 0020, 0025-0026, 0029, 0032, 0034-0035, and figs 6-7 for additional details.
Paramesh, S. P., & Shreedhara, K. S. (2018). Automated IT service desk systems using machine learning techniques. In Data analytics and learning: proceedings of dal 2018 (pp. 331-346). Singapore: Springer Singapore. – disclose service ticket routing in a service desk system, using data preprocessing, vector representation, vector dimensionality reduction, building classifier model, and outputting ticket label from unstructured input tickets. See Abstract and figs. 2 and sections 3-4 for additional details.
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 Philip H Lam whose telephone number is (571)272-1721. The examiner can normally be reached 9 AM-3 PM Pacific time.
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, Bhavesh Mehta can be reached on 571-272-7453. 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.
/PHILIP H LAM/ Examiner, Art Unit 2656
Prosecution Insights