DETAILED ACTION
This examination is in response to the communication filed on 05/05/2026. Claims 1-15 are currently pending, wherein 8 has been canceled and claims 1, 4, 7, 9, 11-15 have been amended, and new claims 16-20 have been added.
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 .
Response to Amendment/argument
Applicant’s amendment/arguments with respect to the rejection of claims 1-15 under §101 have been fully considered and are persuasive. The rejection of claims 1-15 under §101 has been withdrawn.
Applicant’s arguments with respect to claims 1-15 under §§102 and 103 have been considered but are moot because the new ground of rejection does not rely on any reference applied in the prior rejection of record for any teaching or matter specifically challenged in the argument.
Claim Rejections - 35 USC § 103
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Claim 1, 2 and 6-13 are rejected under 35 U.S.C. 103 as being unpatentable over Vukovic et al. (US 10,650,356 B2; herein “Vukovic”) cited in Applicant IDS filed on 04/05/2024 further in view of Wu et al. (US 2021/0374201 A1; herein “Wu”).
Regarding claim 1, Vukovic teaches a method for providing assistance for a user of a processing machine, comprising the following steps performed by a server computer (Fig. 9, Computer System/Server 912):
receiving a request text, representing the request for assistance from the user (Fig. 2A, step 205; Fig. 8, step 810 and col. 17, lines 58-62 teaches “…an operation 80 of in response to receiving computer system service data…” );
receiving access information, representing the user who is requesting assistance and a type of machine for which the assistance is being requested (Col. 8, lines 32-39 teaches “…a complete feature set (e.g., set of all feature related to a computer system issue, computer system, or user skill level) is utilized by statistical analyzer 350…” and col. 9, lines 21-23 teaches “…natural language processing system 412 analyzes a received unstructured textual report relating to the user skill level” the User skill level is interpreted as information representing who is requesting the assistance);
processing the request text through a natural language processing engine, trained to extract values for a plurality of predetermined parameters, so as to generate an input vector containing the values thus extracted from the request text (Col. 9, lines 9-20 teaches “…the natural language processing system 412 responds to electronic document submissions sent by client application 408…analyzes a received unstructured textual report (e.g., unstructured textual data 305, error reports 312, user input 309, emails 307, text messages 315…catalogs 340, computer system information 345, etc.) to identify a feature or feature set…and one or more suggestions (e.g., how to resolve the service issue)”; Fig. 8, step 820 and col. 17, lines 58-62 teaches ““…an operation 80 of in response to receiving computer system service data, identifying, by a second computer system, a computer system service category among a plurality of computer system categories…” );
accessing a database containing a plurality of solution texts, each solution text constituting a predetermined reply to a possible request for assistance or to a type of request for assistance (Col. 9, lines 9-20 teaches “…the natural language processing system 412…analyzes a received unstructured textual report (e.g., …catalogs 340, computer system information 345, etc.) to identify a feature or feature set…and one or more suggestions (e.g., how to resolve the service issue)”; Fig. 8, step 820 and col. 17, lines 58-62 teaches ““…an operation 80 of in response to receiving computer system service data, identifying, by a second computer system, a computer system service category among a plurality of computer system categories…Fig. 8 steps 830 and 840 and col. 17, line 63 to col. 18, line 7 “…an operation 820 of identifying, by the second computer system, one or more computer system service tasks…an operation 830 of selecting, by the second computer system, a catalog among a plurality of catalogs…an operation 840 of generating…one or more suggestions based on the catalog and the one or more computer system service tasks…” );
selecting a solution text from the plurality of solution texts, by processing the input vector, to make the selected solution text available to the user (Fig. 8, step 850 and Col. 18, lines 5-7 teaches “…displaying by the second computer system, the one or more suggestion on a display logically coupled to the computer system”).
Vukovic fails to disclose wherein the access information and the input vector generated from the requested text are used by the server computer to generate an enriched input vector, the enriched input vector including both the values for the predetermined parameters and the access information as recited in amended claim 1.
Wu teaches an action recommendation engine which includes functionality to send pre-signup data 110 and action descriptions 118 to a recommendation engine which then provides a recommended action based on a comparison between the pre-signup data vectors and action description vectors. More specifically, Wu, ¶¶[0022]-[0023] teach “The pre-signup data (110) may represented a goal of the user to perform one or more specific actions using the MA (108)…include search terms (112), webpage descriptions (114), and/or button descriptions (116)…a button may correspond to a specific product or service offered by the MA (108)…pre-signup data (110) may correspond to an identifier (e.g., a unique identifier) of the user…“and ¶[0025] teaches “the pre-signup-data vector (120) is a combination (e.g., a concatenation) of vectors corresponding to search terms (112), webpage descriptions (114), and/or button descriptions (116)”. Accordingly, Wu teaches recommending an action based on the processing of the pre-signup data vector, which is a concatenation of vectors, and action description vectors. The pre-signup data vector is interpreted as a combination of vectors corresponding to the claimed enriched input vector. Thus Wu teaches the access information and the input vector generated from the requested text are used by the server computer to generate an enriched input vector, the enriched input vector including both the values for the predetermined parameters and the access information
Vukovic differs from the claimed invention in that Vukovic fails to disclose generating an enriched vector by concatenating/combine vectors including user/access information and topic information. Combining/concatenating vectors to produce enhanced vectors is known in the art as evidenced by Wu. Therefore, it would have been obvious to one having ordinary skill in the art to have modified the system taught by Vukovic to include enhancing the input vector with user identification information as taught by Wu as it merely constitutes that combination of known processes to achieve the predictable result of using user identification to determine which responses are more appropriate for the specific user goals/desires.
Regarding claim 2, the combination of Vukovic and Wu teaches all of the elements of claim 1 (see detailed element listing above). In addition, Vukovic further teaches the database is provided with a reference dataset including, for each solution text, a respective solution vector, made up of values of the plurality of predetermined parameters extracted from that solution text by the NLP engine, the step of selecting being carried out dependently also on the reference dataset (Fig. 2A and col. 4, lines 28-48 and col. 6, lines 48-52 teaches “…the intelligent self-service delivery advisory further extracts relevant features and then, using a machine learning model trained using support vector machine (SVM), predicts the category and task associated with the change request” i.e., generates the feature set vectors and col. 8, lines 35-53 teaches “… a complete feature set…is utilized by statistical analyzer 350, using the method described herein (e.g., kMeans clustering), to determine correlations between features…of a particular computer system issue…and possible solutions…k-means clustering aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest means, serving as a protype of the cluster…statistical analyzer 350 generates suggestions 355…”).
Regarding claim 6, the combination of Vukovic and Wu teaches all of the elements of claim 1 (see detailed element listing above). In addition, Vukovic further teaches the plurality of predetermined parameters includes semantic metadata (Col. 9, lines 36-39 teaches “In an embodiment, natural language processor 414 performs various method and techniques for analyzing electronic documents (e.g., syntactic analysis, semantic analysis, etc.)…These modules includes, but are not limited to…a semantic relationship identifier 420…”).
Regarding claim 7, the combination of Vukovic and Wu teaches all of the elements of claim 1 (see detailed element listing above). In addition, Vukovic further teaches wherein the step of selecting a solution text is carried out also dependently on the access information (Col. 8, lines 32-39 teaches “…a complete feature set (e.g., set of all feature related to a computer system issue, computer system, or user skill level) is utilized by statistical analyzer 350…” and col. 9, lines 21-23 teaches “…natural language processing system 412 analyzes a received unstructured textual report relating to the user skill level” the User skill level is interpreted as information representing who is requesting the assistance ).
Regarding claim 9, Vukovic teaches a maintenance system for providing assistance for a user of a processing machine, comprising a server computer (Fig. 9, Computer System/Server 912) and a database (Col. 10, lines 5-6 teaches “…the output of natural language processing system 412 populates a text index, a triple store, or a relational database…”) containing a plurality of solution texts, each solution text constituting a predetermined reply to a possible request for assistance or to a type of request for assistance (Col. 10, lines 60-66 teaches “…information corpus 426 is a storage mechanism that houses a standardized, consistent, clean and integrated list of features…information corpus 42 also stores, for each feature, a list of associated suggestions…” ), wherein the server computer:
is configured to receive, through a communication system, a request text representing the request for assistance from the user, and access information representing the user who is requesting assistance and a type of machine for which the assistance is being requested for (Fig. 2A, step 205; Fig. 8, step 810 and col. 17, lines 58-62 teaches “…an operation 80 of in response to receiving computer system service data…” and Col. 8, lines 32-39 teaches “…a complete feature set (e.g., set of all feature related to a computer system issue, computer system, or user skill level) is utilized by statistical analyzer 350…”),
is programmed to process the request text through a natural language processing engine, trained to extract values for a plurality of predetermined parameters, so as to generate an input vector containing the values thus extracted from the request text (Col. 9, lines 9-20 teaches “…the natural language processing system 412 responds to electronic document submissions sent by client application 408…analyzes a received unstructured textual report (e.g., unstructured textual data 305, error reports 312, user input 309, emails 307, text messages 315…catalogs 340, computer system information 345, etc.) to identify a feature or feature set…and one or more suggestions (e.g., how to resolve the service issue)”; Fig. 8, step 820 and col. 17, lines 58-62 teaches ““…an operation 80 of in response to receiving computer system service data, identifying, by a second computer system, a computer system service category among a plurality of computer system categories…”),
includes a selector programmed to select a solution text from the plurality of solution texts by processing the input vector and on the access information (Fig. 8, step 850 and Col. 18, lines 5-7 teaches “…displaying by the second computer system, the one or more suggestion on a display logically coupled to the computer system”);
is configured to make the selected solution text available to the user (Fig. 8, step 850 and Col. 18, lines 5-7 teaches “…displaying by the second computer system, the one or more suggestion on a display logically coupled to the computer system”).
Vukovic fails to disclose wherein the access information and the input vector generated from the requested text are used by the server computer to generate an enriched input vector, the enriched input vector including both the values for the predetermined parameters and the access information as recited in amended claim 1.
Wu teaches an action recommendation engine which includes functionality to send pre-signup data 110 and action descriptions 118 to a recommendation engine which then provides a recommended action based on a comparison between the pre-signup data vectors and action description vectors. More specifically, Wu, ¶¶[0022]-[0023] teach “The pre-signup data (110) may represented a goal of the user to perform one or more specific actions using the MA (108)…include search terms (112), webpage descriptions (114), and/or button descriptions (116)…a button may correspond to a specific product or service offered by the MA (108)…pre-signup data (110) may correspond to an identifier (e.g., a unique identifier) of the user…“and ¶[0025] teaches “the pre-signup-data vector (120) is a combination (e.g., a concatenation) of vectors corresponding to search terms (112), webpage descriptions (114), and/or button descriptions (116)”. Accordingly, Wu teaches recommending an action based on the processing of the pre-signup data vector, which is a concatenation of vectors, and action description vectors. The pre-signup data vector is interpreted as a combination of vectors corresponding to the claimed enriched input vector. Thus Wu teaches the access information and the input vector generated from the requested text are used by the server computer to generate an enriched input vector, the enriched input vector including both the values for the predetermined parameters and the access information
Vukovic differs from the claimed invention in that Vukovic fails to disclose generating an enriched vector by concatenating/combine vectors including user/access information and topic information. Combining/concatenating vectors to produce enhanced vectors is known in the art as evidenced by Wu. Therefore, it would have been obvious to one having ordinary skill in the art to have modified the system taught by Vukovic to include enhancing the input vector with user identification information as taught by Wu as it merely constitutes that combination of known processes to achieve the predictable result of using user identification to determine which responses are more appropriate for the specific user goals/desires.
Regarding claim 10, the combination of Vukovic and Wu teaches all of the elements of claim 9 (see detailed element listing above). In addition, Vukovic further teaches the database is provided with a reference dataset including, for each solution text, a respective solution vector, made up of values of the plurality of predetermined parameters extracted from that solution text by the NLP engine, and wherein the selector is programmed to process the input vector based on the reference dataset (Fig. 2A and col. 4, lines 28-48 and col. 6, lines 48-52 teaches “…the intelligent self-service delivery advisory further extracts relevant features and then, using a machine learning model trained using support vector machine (SVM), predicts the category and task associated with the change request” i.e., generates the feature set vectors and col. 8, lines 35-53 teaches “… a complete feature set…is utilized by statistical analyzer 350, using the method described herein (e.g., kMeans clustering), to determine correlations between features…of a particular computer system issue…and possible solutions…k-means clustering aims to partition n observations into k clusters in which each observation belongs to the cluster with the nearest means, serving as a protype of the cluster…statistical analyzer 350 generates suggestions 355…”).
Regarding claim 11, Vukovic teaches all of the elements of claim 9 (see detailed element listing above). In addition, Vukovic further teaches the database includes a list of predetermined categories of possible requests from the user (Fig. 8, step 810 and col. 17, lines 58-62 teaches “…an operation 810 of … identifying, by a second computer system, a computer system service category among a plurality of computer system categories” See also Fig. 2A, step 210 “Identify Category (SVM));
the server computer is programmed to perform a step of categorizing, in the server computer which it processes the request text to derive therefrom a degree of matching the predetermined categories (Fig. 2A, step 210 “Identify Category (SVM)),
the selector is programmed to assign a confidence level to the selection (Col. 2, lines 56-62 teaches “…the groups of suggestions lists change categories…for each suggestion…and a confidence values from a column 140…for each suggestion”),
the server computer is also programmed to generate a message for the user, dependently on the result of the step of categorizing, if the confidence level on the selection text is less than a predetermined threshold value (Col. 2, lines 63-65 teaches “…the advisory offers suggestions, with certain confidence…” and col. 3, lines 1-10 teaches “…each suggestion will assigned a confidence value to show the hierarchical nature of the suggestions…a confidence value from column 140 is the likelihood that a suggestion will resolve the issue” ).
Regarding claim 12, the combination of Vukovic and Wu teaches all of the elements of claim 9 (see detailed element listing above). In addition, Vukovic further teaches the server computer is configured to receive, through the communication system, access information representing the user who is requesting assistance and the type of machine for which the assistance is being requested, and wherein the selector is programmed to select the solution text also dependently on the access information (Col. 8, lines 32-39 teaches “…a complete feature set (e.g., set of all feature related to a computer system issue, computer system, or user skill level) is utilized by statistical analyzer 350…” and col. 9, lines 21-23 teaches “…natural language processing system 412 analyzes a received unstructured textual report relating to the user skill level” the User skill level is interpreted as information representing who is requesting the assistance).
Regarding claim 13, the combination of Vukovic and Wu teaches all of the elements of claim 9 (see detailed element listing above). In addition, Vukovic further teaches the server computer is programmed to generate a reaction request and is configured to receive a reaction signal representing an evaluation by the user of the selected solution text (Col. 6, lines 7-11 teaches “…the intelligent self-service delivery advisory further includes an operation 260 of receiving context and history analysis, wherein the history of the user’s selections and input for previous computer system issues are analyzed to provide input for the initial suggestions”);
the selector includes a machine-learned model programmed to self-learn based on the reaction signal (Col. 6 , lines 3-6 teaches “…the intelligent self-service deliver advisory can learn from the interaction with the user and user choices to update state and personalize recommendation for the future (and improve recommendation to other similar tasks and user roles)”).
Regarding claim 16, the combination of Vukovic and Wu teaches all of the elements of claim 1 (See detailed element mapping above). In addition, Vukovic further teaches the enriched input vector is further based on a typology of the machine (Under a broadest reasonable interpretation, “typology of the machine” is interpreted as system or machine configuration information; Col. 8, lines 32-39 teaches “…a complete feature set (e.g., set of all features related to a computer system issue, computer system, or user skill level) is utilized by statistical analyzer 350” all features related to a computer system is interpreted as including typology information).
Claims 3-5 are rejected under 35 U.S.C. 103 as being unpatentable over the combination of Vukovic and Wu as applied to claim 2 above, and further in view of Srivastava et al. (US 2018/0260760 A1; herein “Srivastava”) cited in Applicant IDS filed 04/05/2024.
Regarding claim 3, the combination of Vukovic and Wu teaches all of the elements of claim 2 (see detailed element listing above). In addition, Vukovic further teaches the step of selecting is carried out by a selector including a machine-learned model, the method further comprising
a step of learning by the machine-learned model responsive to a reaction signal received by the server computer (Col. 6, lines 7-11 teaches “…the intelligent self-service delivery advisory further includes an operation 260 of receiving context and history analysis, wherein the history of the user’s selections and input for previous computer system issues are analyzed to provide input for the initial suggestions”),
Vukovic fails to explicitly disclose the reaction signal representing an evaluation by the user of the selected solution text.
Srivastava teaches an automated ticket resolution system and method that includes, inter alia, classifying, using a data model, the ticket data into a ticket type, generating, using the data model and based on the ticket type, a set of recommended resolutions for resolving the issue associated with the projection, and selecting from the set of recommended resolutions, a particular resolution based on a set of selection criteria (Srivastava, Abstract). In addition, Srivastava further teaches the reaction signal representing an evaluation by the user of the selected solution text (¶[0019] teaches “…the cloud platform may monitor the project and/or receive feedback about the particular resolution to determine an effectiveness level of the resolution…This may allow subsequent requests for historical ticket data (e.g., to update the ticket data model) to access the ticket and the particular resolution associated with the ticket”).
The combination of Vukovic and Wu differs from the claimed invention, as defined in claim 3, in that the combination fails to disclose incorporating user feedback regarding previously selected solutions as model fine-tuning parameter. Fine-tuning machine-learning models based on user feedback representing the effectiveness of past model performance is known in the art as evidenced by Srivastava. Therefore, it would have been obvious to one having ordinary skill in the art, before the effective filing date of the invention, to have modified the intelligent self-service delivery advisor system taught by the combination of Vukovic and Wu to include user feedback regarding effectiveness of past model suggestions as taught by Srivastava as it merely constitutes the combination of known processes to achieve the predictable result of fine-tuning the suggestion selection machine learning model based on the effectiveness of past performance.
Regarding claim 4, the combination of Vukovic, Wu and Srivastava teaches all of the elements of claim 3 (see detailed element listing above). In addition, Vukovic further teaches the database includes a list of predetermined categories of possible requests from the user (Fig. 8, step 810 and col. 17, lines 58-62 teaches “…an operation 810 of … identifying, by a second computer system, a computer system service category among a plurality of computer system categories” See also Fig. 2A, step 210 “Identify Category (SVM));
the server computer performs a step of categorizing, in which the server computer processes the request text to derive therefrom a degree of matching the predetermined categories, the machine-learned model assigns a confidence level to the selection, if the confidence level on the selection text is less than a predetermined threshold value or if the reaction signal is negative, the server computer performs a step of generating a message for the user dependently on the result of the step of categorizing (Col. 2, lines 56-62 teaches “…the groups of suggestions lists change categories…for each suggestion…and a confidence values from a column 140…for each suggestion” and Col. 2, lines 63-65 teaches “…the advisory offers suggestions, with certain confidence…” and col. 3, lines 1-10 teaches “…each suggestion will assigned a confidence value to show the hierarchical nature of the suggestions…a confidence value from column 140 is the likelihood that a suggestion will resolve the issue”).
Regarding claim 5, the combination of Vukovic, Wu and Srivastava teaches all of the elements of claim 4 (see detailed element listing above). In addition, Srivastava further teaches categorization is performed by the NLP engine, the NLP engine being also trained to derive from it the degree of matching the predetermined categories of the request texts (“¶[0038] teaches “In some implementations, automatic ticket classification module 270 may apply a text similarity technique”).
The combination of Vukovic and Wu differs from the claimed invention, as defined in claim 5, in that the combination fails to disclose categorizing the service request text utilizing a natural language processing engine that utilizes a degree of matching algorithm. Categorizing service requests based on a degree of matching algorithm, e.g., text similarity technique is known in the art as evidenced by Srivastava. Therefore, it would have been obvious to one having ordinary skill in the art, before the effective filing date of the invention, to have modified the intelligent self-service delivery advisor system taught by Vukovic to include utilize a text similarity technique as taught by Srivastava as it merely constitutes the substitution of known processes to achieve the predictable result of classifying the text of the service requests based it similarity with predefined categories.
Claims 14 and 15 are rejected under 35 U.S.C. 103 as being unpatentable over f Vukovic, in view of Wu, further in view of Mackie et al. (US 2023/0006907 A1; herein “Mackie”).
Regarding claim 14, Vukovic teaches a method for obtaining assistance for a processing machine, the method comprising the following steps, performed by a client computer (Fig. 4, element 400) accessible by a user of the machine and configured to exchange information with a server computer (Fig. 4, element 412) through a communication system (Fig. 4, element 415) :
providing access information representing the user who is requesting assistance and the type of machine for which the assistance is being requested (Col. 8, lines 32-39 teaches “…a complete feature set (e.g., set of all feature related to a computer system issue, computer system, or user skill level) is utilized by statistical analyzer 350…” and col. 9, lines 21-23 teaches “…natural language processing system 412 analyzes a received unstructured textual report relating to the user skill level” the User skill level is interpreted as information representing who is requesting the assistance);
enabling the user to enter a request text, representing the request for assistance from the user (Col. 3, lines 28-38 teaches “…the dynamic user interface 100 will provide information on why a selection is not available…The advisor then prompts the user for additional information…” and Fig. 2A, step 205; Fig. 8, step 810 and col. 17, lines 58-62 teaches “…an operation 80 of in response to receiving computer system service data…” );
via the client computer (the “or” makes this limitation optional) or the server computer, processing the request text through a natural language processing engine, trained to extract values for a plurality of predetermined parameters, so as to generate an input vector containing the values thus extracted from the request text (Col. 9, lines 9-20 teaches “…the natural language processing system 412 responds to electronic document submissions sent by client application 408…analyzes a received unstructured textual report (e.g., unstructured textual data 305, error reports 312, user input 309, emails 307, text messages 315…catalogs 340, computer system information 345, etc.) to identify a feature or feature set…and one or more suggestions (e.g., how to resolve the service issue)”; Fig. 8, step 820 and col. 17, lines 58-62 teaches ““…an operation 80 of in response to receiving computer system service data, identifying, by a second computer system, a computer system service category among a plurality of computer system categories…”);
via the client computer (the “or” makes this limitation optional) or the server computer, accessing a database containing a plurality of solution texts, each solution text constituting a predetermined reply to a possible request for assistance or to a type of request for assistance (Col. 9, lines 9-20 teaches “…the natural language processing system 412…analyzes a received unstructured textual report (e.g., …catalogs 340, computer system information 345, etc.) to identify a feature or feature set…and one or more suggestions (e.g., how to resolve the service issue)”; Fig. 8, step 820 and col. 17, lines 58-62 teaches ““…an operation 80 of in response to receiving computer system service data, identifying, by a second computer system, a computer system service category among a plurality of computer system categories…Fig. 8 steps 830 and 840 and col. 17, line 63 to col. 18, line 7 “…an operation 820 of identifying, by the second computer system, one or more computer system service tasks…an operation 830 of selecting, by the second computer system, a catalog among a plurality of catalogs…an operation 840 of generating…one or more suggestions based on the catalog and the one or more computer system service tasks…”);
via the client computer (the “or” makes this limitation optional) or the server computer, selecting a solution text from the plurality of solution texts by processing the the input vector and on the access information (col. 17, line 63 to col. 18, line 7 “…an operation 840 of generating…one or more suggestions based on the catalog and the one or more computer system service tasks…” );
making the selected solution text available to the user (Fig. 8, step 850 and Col. 18, lines 5-7 teaches “…displaying by the second computer system, the one or more suggestion on a display logically coupled to the computer system”).
Vukovic fails to disclose wherein the access information and the input vector generated from the requested text are used by the server computer to generate an enriched input vector, the enriched input vector including both the values for the predetermined parameters and the access information as recited in amended claim 14.
Wu teaches an action recommendation engine which includes functionality to send pre-signup data 110 and action descriptions 118 to a recommendation engine which then provides a recommended action based on a comparison between the pre-signup data vectors and action description vectors. More specifically, Wu, ¶¶[0022]-[0023] teach “The pre-signup data (110) may represented a goal of the user to perform one or more specific actions using the MA (108)…include search terms (112), webpage descriptions (114), and/or button descriptions (116)…a button may correspond to a specific product or service offered by the MA (108)…pre-signup data (110) may correspond to an identifier (e.g., a unique identifier) of the user…“and ¶[0025] teaches “the pre-signup-data vector (120) is a combination (e.g., a concatenation) of vectors corresponding to search terms (112), webpage descriptions (114), and/or button descriptions (116)”. Accordingly, Wu teaches recommending an action based on the processing of the pre-signup data vector, which is a concatenation of vectors, and action description vectors. The pre-signup data vector is interpreted as a combination of vectors corresponding to the claimed enriched input vector. Thus Wu teaches the access information and the input vector generated from the requested text are used by the server computer to generate an enriched input vector, the enriched input vector including both the values for the predetermined parameters and the access information
The combination of Vukovic and Wu differs from the claimed invention, as defined in claim 14 in that Vukovic fails to disclose generating an enriched vector by concatenating/combine vectors including user/access information and topic information. Combining/concatenating vectors to produce enhanced vectors is known in the art as evidenced by Wu. Therefore, it would have been obvious to one having ordinary skill in the art to have modified the system taught by Vukovic to include enhancing the input vector with user identification information as taught by Wu as it merely constitutes that combination of known processes to achieve the predictable result of using user identification to determine which responses are more appropriate for the specific user goals/desires.
The combination of Vukovic and Wu fails to disclose receiving and validating login credentials entered by the user or that the step of providing access information and enabling the user to enter a request test are responsive to a result of validating the login credentials.
Mackie disclose a user interface which enables “a user to submit user authentication credentials for verification, and if successfully verified, began an authenticated session associated with a particular application” (Mackie, ¶[0068]). Therefore, Mackie teaches receiving and validating login credentials entered by the user and responsive to a result of validating the login credentials allowing access to an application.
The combination of Vukovic and Wu differs from the claimed invention, as defined by claim 14, in that the combination fails to disclose that the user interface includes means for receiving and validating login credentials prior at allowing user access to the intelligent self-service delivery advisor. User interfaces which provide means for receiving and validating login credentials prior to allowing access to an application are known in the art as evidenced by Mackie. Therefore, it would have been obvious to one having ordinary skill in the art, before the effective filing date of the invention, to have modified the dynamic user interface taught by the combination of Vukovic and Wu to include means for validating user login credentials prior to allowing access to the intelligent self-service delivery advisor as it merely constitutes the combination of known processes to achieve the predictable result to preventing unauthorized access to the intelligent self-service delivery advisor.
Regarding claim 15, Vukovic teaches a method for providing assistance for a user of a processing machine, the method comprising the following steps:
, providing access information representing the user who is requesting assistance and the type of machine which the assistance is being requested for (Col. 8, lines 32-39 teaches “…a complete feature set (e.g., set of all feature related to a computer system issue, computer system, or user skill level) is utilized by statistical analyzer 350…” and col. 9, lines 21-23 teaches “…natural language processing system 412 analyzes a received unstructured textual report relating to the user skill level” the User skill level is interpreted as information representing who is requesting the assistance);
enabling the user to enter a request text, representing the request for assistance from the user (Col. 3, lines 28-38 teaches “…the dynamic user interface 100 will provide information on why a selection is not available…The advisor then prompts the user for additional information…” and Fig. 2A, step 205; Fig. 8, step 810 and col. 17, lines 58-62 teaches “…an operation 80 of in response to receiving computer system service data…”);
via a server computer (Fig. 6, host device 621 and Fig. 9, Server 912), receiving the access information and the request text (Col. 8, lines 32-39 teaches “…a complete feature set (e.g., set of all feature related to a computer system issue, computer system, or user skill level) is utilized by statistical analyzer 350…” and col. 9, lines 21-23 teaches “…natural language processing system 412 analyzes a received unstructured textual report relating to the user skill level” the User skill level is interpreted as information representing who is requesting the assistance);
processing the request text through a natural language processing engine, trained to extract values for a plurality of predetermined parameters, so as to generate an input vector containing the values thus extracted from the request text (Col. 9, lines 9-20 teaches “…the natural language processing system 412 responds to electronic document submissions sent by client application 408…analyzes a received unstructured textual report (e.g., unstructured textual data 305, error reports 312, user input 309, emails 307, text messages 315…catalogs 340, computer system information 345, etc.) to identify a feature or feature set…and one or more suggestions (e.g., how to resolve the service issue)”; Fig. 8, step 820 and col. 17, lines 58-62 teaches ““…an operation 80 of in response to receiving computer system service data, identifying, by a second computer system, a computer system service category among a plurality of computer system categories…”);
accessing a database containing a plurality of solution texts, each solution text constituting a predetermined reply to a possible request for assistance or to a type of request for assistance (Col. 9, lines 9-20 teaches “…the natural language processing system 412…analyzes a received unstructured textual report (e.g., …catalogs 340, computer system information 345, etc.) to identify a feature or feature set…and one or more suggestions (e.g., how to resolve the service issue)”; Fig. 8, step 820 and col. 17, lines 58-62 teaches ““…an operation 80 of in response to receiving computer system service data, identifying, by a second computer system, a computer system service category among a plurality of computer system categories…Fig. 8 steps 830 and 840 and col. 17, line 63 to col. 18, line 7 “…an operation 820 of identifying, by the second computer system, one or more computer system service tasks…an operation 830 of selecting, by the second computer system, a catalog among a plurality of catalogs…an operation 840 of generating…one or more suggestions based on the catalog and the one or more computer system service tasks…”);
selecting a solution text from the plurality of solution texts, dependently on the input vector and on the access information, to make the selected solution text available to the user (col. 17, line 63 to col. 18, line 7 “…an operation 840 of generating…one or more suggestions based on the catalog and the one or more computer system service tasks…” ).
Vukovic fails to disclose wherein the access information and the input vector generated from the requested text are used by the server computer to generate an enriched input vector, the enriched input vector including both the values for the predetermined parameters and the access information as recited in amended claim 15.
Wu teaches an action recommendation engine which includes functionality to send pre-signup data 110 and action descriptions 118 to a recommendation engine which then provides a recommended action based on a comparison between the pre-signup data vectors and action description vectors. More specifically, Wu, ¶¶[0022]-[0023] teach “The pre-signup data (110) may represented a goal of the user to perform one or more specific actions using the MA (108)…include search terms (112), webpage descriptions (114), and/or button descriptions (116)…a button may correspond to a specific product or service offered by the MA (108)…pre-signup data (110) may correspond to an identifier (e.g., a unique identifier) of the user…“and ¶[0025] teaches “the pre-signup-data vector (120) is a combination (e.g., a concatenation) of vectors corresponding to search terms (112), webpage descriptions (114), and/or button descriptions (116)”. Accordingly, Wu teaches recommending an action based on the processing of the pre-signup data vector, which is a concatenation of vectors, and action description vectors. The pre-signup data vector is interpreted as a combination of vectors corresponding to the claimed enriched input vector. Thus Wu teaches the access information and the input vector generated from the requested text are used by the server computer to generate an enriched input vector, the enriched input vector including both the values for the predetermined parameters and the access information
The combination of Vukovic and Wu differs from the claimed invention, as defined in claim 15 in that Vukovic fails to disclose generating an enriched vector by concatenating/combine vectors including user/access information and topic information. Combining/concatenating vectors to produce enhanced vectors is known in the art as evidenced by Wu. Therefore, it would have been obvious to one having ordinary skill in the art to have modified the system taught by Vukovic to include enhancing the input vector with user identification information as taught by Wu as it merely constitutes that combination of known processes to achieve the predictable result of using user identification to determine which responses are more appropriate for the specific user goals/desires.
The combination of Vukovic and Wu fails to disclose via a client computer accessible to a user of the machine, receiving and validating login credentials entered by the user or that the step of providing access information and enabling the user to enter a request test are responsive to a result of validating the login credentials.
Mackie disclose a user interface which enables “a user to submit user authentication credentials for verification, and if successfully verified, began an authenticated session associated with a particular application” (Mackie, ¶[0068]). Therefore, Mackie teaches receiving and validating login credentials entered by the user and responsive to a result of validating the login credentials allowing access to an application.
The combination of Vukovic and Wu differs from the claimed invention, as defined by claim 15, in that the combination fails to disclose that the user interface includes means for receiving and validating login credentials prior at allowing user access to the intelligent self-service delivery advisor. User interfaces which provide means for receiving and validating login credentials prior to allowing access to an application are known in the art as evidenced by Mackie. Therefore, it would have been obvious to one having ordinary skill in the art, before the effective filing date of the invention, to have modified the dynamic user interface taught by the combination of Vukovic and Wu to include means for validating user login credentials prior to allowing access to the intelligent self-service delivery advisor as it merely constitutes the combination of known processes to achieve the predictable result to preventing unauthorized access to the intelligent self-service delivery advisor.
Claims 17-19 are rejected under 35 U.S.C. 103 as being unpatentable over the combination of Vukovic, Wu and Srivastava as applied to claim 3 above, and further in view of Kumar et al. (US 2022/0075843 A1; herein “Kumar”).
Regarding claim 17, the combination of Vukovic, Wu and Srivastava teaches all of the elements of claim 3 (See detailed element mapping above). In addition, Vukovic further teaches the selector is configured to select the solution text based on an algorithm based on a (Col. 8, lines 32-52 teaches “…a complete feature set…is utilized by statistical analyzer 350, using the methods described herein (e.g., kMeans clustering), to determine correlations between features…and possible solutions…Particular features within the suggestions 355 are weighted” kMeans clustering is based on distances).
However, the combination of Vukovic, Wu and Srivastava fails to specifically recite that the kMeans is based on weighted distances.
Kumar teaches a vector clustering algorithm/methods that utilizes a weighted K-means clustering algorithm, wherein element weights are defined by Euclidian norm which is a distance measure (Kumar, ¶[0096]).
The combination of Vukovic, Wu and Srivastava differs from the claimed invention, as defined by claim 17, in that the combination fails to disclose that the K-means clustering is based on a weighted distance. Weighted K-means clustering is known in the art as evidenced by Kumar. Therefore, it would have been obvious to one having ordinary skill in the art, before the effective filing date of the invention, to have modified the dynamic user interface taught by the combination of Vukovic, Wu and Srivastava to include weight K-means clustering as it merely constitutes the combination of known processes to achieve the predictable result of allowing control over much individual features contribute the vector embedding clustering.
Regarding claim 18, the combination of Vukovic, Wu, Srivastava, and Kumar teaches all of the elements of claim 17 (See detailed element mapping above). In addition, Kumar further teaches the server computer modifies weights of the algorithm based on the distance (¶[0096] teaches “The weighted K-means clustering technique, using element weights defined by … Euclidian norm, may be incorporated into the compression method disclosed herein. For example, the weighted K-means clustering may be used at step 406 of the method 400. Thus, the disclosed weighted K-means clustering technique enables non-uniform quantization, which may enable greater focus on different vectors, based upon their relative importance (as represented by their relative element weights). In particular, the use of the Euclidian norm to define element weights enables the disclosed weighted K-means clustering technique to be used independently of the frequency of the words in a corpus.”).
The combination of Vukovic, Wu and Srivastava differs from the claimed invention, as defined by claim 17, in that the combination fails to disclose that the K-means clustering is based on a weighted distance. Weighted K-means clustering is known in the art as evidenced by Kumar. Therefore, it would have been obvious to one having ordinary skill in the art, before the effective filing date of the invention, to have modified the dynamic user interface taught by the combination of Vukovic, Wu and Srivastava to include weight K-means clustering as it merely constitutes the combination of known processes to achieve the predictable result of allowing control over much individual features contribute the vector embedding clustering.
Regarding claim 19, the combination of Vukovic, Wu, Srivastava and Kumar teaches all of the elements of claim 18 (See detailed element mapping above). In addition, Vukovic further teaches the selected solution text is transmitted by the server computer and wherein the server computer receives from the user a reaction signal representing an evaluation by the user of the selected solution text, and wherein the server computer modifies the weights of the algorithm further in dependence of the reaction signal from the user (¶[0029] teaches “…feedback from actual user actions may be used to train the classifier…”).
Claim 20 is rejected under 35 U.S.C. 103 as being unpatentable over the combination of Vukovic and Wu as applied to claim 1, further in view of Lim et al. (US 2021/0374201 A1; herein “Lim”).
Regarding claim 20, the combination of Vukovic and Wu teaches all of the elements of claim 1 (See detailed element mapping above). In addition, Wu further teaches the selected solution text is transmitted by the server computer and wherein the server computer receives from the user a reaction signal representing an evaluation by the user of the selected solution text and wherein the method further comprising a step of training the NLP engine responsive to the reaction signal received at the server computer (¶[0029] teaches “…feedback from actual user actions may be used to train the classifier…” ).
The combination of Vukovic and Wu fails to specifically disclose wherein the reaction signal is either positive or negative.
Lim teaches a system and method for providing recommendations which includes, inter alia, using positive and negative feedback with respect to the recommendations for training of the models. (Lim ¶[0078).
The combination of Vukovic and Wu differs from the claimed invention, as defined by claim 20, in that the combination fails to disclose using positive and negative feedback data to training the model(s). Using positive and negative feedback for training models is well known in the art as evidenced by Lim. Therefore, it would have been obvious to one having ordinary skill in the art, before the effective filing date of the invention, to have modified the dynamic user interface taught by the combination of Vukovic and Wu to include the feedback including positive or negative reactions to train the model as taught by Lim as it merely constitutes the combination of known processes to achieve the predictable result of utilizing user feedback/reactions when training the recommendation/solution response model in order to improve model performance.
Conclusion
Applicant's amendment necessitated the new ground(s) of rejection presented in this Office action. Accordingly, THIS ACTION IS MADE FINAL. See MPEP § 706.07(a). Applicant is reminded of the extension of time policy as set forth in 37 CFR 1.136(a).
A shortened statutory period for reply to this final action is set to expire THREE MONTHS from the mailing date of this action. In the event a first reply is filed within TWO MONTHS of the mailing date of this final action and the advisory action is not mailed until after the end of the THREE-MONTH shortened statutory period, then the shortened statutory period will expire on the date the advisory action is mailed, and any nonprovisional extension fee (37 CFR 1.17(a)) pursuant to 37 CFR 1.136(a) will be calculated from the mailing date of the advisory action. In no event, however, will the statutory period for reply expire later than SIX MONTHS from the mailing date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to PENNY L CAUDLE whose telephone number is (703)756-1432. The examiner can normally be reached M-Th 8:00 am to 5:00 pm eastern.
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, Daniel Washburn can be reached at 571-272-5551. 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.
/PENNY L CAUDLE/Examiner, Art Unit 2657
/DANIEL C WASHBURN/Supervisory Patent Examiner, Art Unit 2657
Prosecution Insights