The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
DETAILED ACTION
Status of the Claims
This is in response to the amendment filed on 6/16/26. Claims 1-30 are pending in the application.
The text of those sections of Title 35, U.S. Code not included in this action can be found in a prior Office action.
Specification
Objections to the Specification have been withdrawn.
Terminal Disclaimer
The terminal disclaimer filed on 6/16/26 disclaiming the terminal portion of any patent granted on this application which would extend beyond the expiration date of the full statutory term of any patent granted on pending reference Application Number US 12183173 B2 has been reviewed and is accepted. The terminal disclaimer has been recorded.
Claim Rejections - 35 USC § 101
Claim Rejections under 35 USC § 101 have been withdrawn.
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 Rejections - 35 USC § 103
Claims 1-3, 5, 7-11, 14-17, 19-21 and 24-30 are rejected under 35 U.S.C. 103 as being unpatentable over Porras et al. (Porras; US 20160219078) in view of Zhu et al. (Zhu; US 20190324780), further in view of Tur et al. (Tur; US 20170061316 A1).
Regarding Claim 1, Porras discloses an apparatus for providing personalized and contextualized environment security information (Abstract facilitate analysis of network activity or to respond to network security events), comprising:
a memory (Fig 8); and
a processor (Fig 8) coupled with the memory and configured to:
detect a security event in an environment based on a first portion of collected sensor data ([0028] user interaction detection device(s) 106 may include the interactive display device 104 and/or other human activity detection devices (e.g., various types of sensors, including motion sensors, kinetic sensors, proximity sensors, thermal sensors, pressure sensors, force sensors, inertial sensors, cameras, microphones, gaze tracking systems, and/or others; [0036] network activity data 140 may be generated by, e.g., one or more network sensors or passive network monitoring programs; 602 of Fig 6A monitor network traffic);
generate contextual information (for instance highlighted nodes) for the security event based on a second portion of the collected sensor data, wherein the second portion is different from the first portion ([0038]-[0039] security threats detected; [0111] 610 of Fig 6A, nodes or flows on the network visualization may be highlighted dynamically in response to the occurrence of network events or un-highlighted in response to the network events being remediated (e.g., by user interactions 120));
user interface device(s) 104 may be embodied as…a touchscreen display device…includes audio input and output devices capable of capturing and recording human conversational spoken natural language input and outputting system-generated conversational spoken natural language output (such as microphones, speakers and headphones or earbuds)),
generate a user-specific remediation conversation from a plurality of different remediation conversations ([0087], 0091]-[0092]) based on the user for instance new web server connecting), and a user-specific subset of the contextual information (Figs 6A-6B, 612 user dialog detected, 614 translate dialog to network directive, 664 respond with NL dialog; [0091] notify me when a new web server appears on the network,[ the subset is a specific scenario established by the user to notify of new web servers],), wherein each of the plurality of different remediation conversations comprises dialogue ([0092] natural language generator (NLG) module 446 generates a natural language version of the dialog output intent 444, the NL dialog output 448, which is output via, e.g., one or more speakers [0091] notification that “a new web server has connected to the network.” [0087] confirmations that the system 110 is going to execute a user-requested command (e.g., “are you sure you want me to disconnect that node from the network? OK, disconnecting the node from the network”) interaction model 414 may be defined or personalized for specific types of users; [0091]-[0092] responds to specific requests made by the user) response may include the phrase “Which <node> do you want to disconnect”; [0117] system may proceed to block 672 and respond by outputting NL dialog asking the user for further clarification of the request); and
output at least a first portion of the user-specific remediation conversation on the output device ([0092] the NL dialog output 448, which is output via, e.g., one or more speakers, displays, or other user interface and/or user interaction detection devices 104, 106…an NL response may include the phrase “Which <node> do you want to disconnect”), where <node> indicates a parameter*; notification that “a new web server has connected to the network.”; 664 of Fig 6B;). Porras does NOT specify a user profile for the conversation nor tailoring the dialogue, but does teach identifying a user ([0075] handling subsystem 122 may perform authentication processes to verify a user's identity).
In the same field of endeavor, Zhu discloses a method of receiving a user input including a partial request from a client system of a first user, analyzing the user input to generate one or more candidate hypotheses based on a personalized language model where each of the candidate hypotheses includes one or more of an intent-suggestion or a slot-suggestion, sending instructions for presenting one or more suggested auto-completions corresponding to one or more of the candidate hypotheses, respectively, to the client system, where each suggested auto-completion comprises the partial request and the corresponding candidate hypothesis, receiving an indication of a selection by the first user of a first suggested auto-completion of the suggested auto-completions from the client system, and executing one or more tasks based on the first suggested auto-completion selected by the first user via one or more agents. The social-networking system may also include suitable components such as network interfaces, security mechanisms.
Zhu teaches a user profile indicative of access rights and user interface preferences ([0004] user profile may include demographic information, communication-channel information, and information on personal interests of the user; [0006] execute tasks that are relevant to user interests and preferences based on the user profile without a user input. In particular embodiments, the assistant system may check privacy settings to ensure that accessing a user's profile or other user information and executing different tasks are permitted subject to the user's privacy settings; [0039] assistant system 140 may proactively execute pre-authorized tasks that are relevant to user interests and preferences based on the user profile).
Zhu discloses generating a remediation conversation based on the user profile, the security event, and the contextual information, wherein the remediation conversation comprises dialogue for providing security information and guiding the first user to resolve the security event (Abstract, [0006] The assistant system may generate a response for the user regarding the information or services by using natural-language generation; [0029]; [0056] processing result may be stored in the user context engine 225 as part of the user profile. The online inference service 227 may analyze the conversational data associated with the user that are received by the assistant system 140 at a current time. The analysis result may be stored in the user context engine 225 also as part of the user profile…inference service 227 may extract personalization features from the plurality of data… semantic information aggregator 230 may then send the aggregated information to the NLU module 220 to facilitate the domain classification/selection; [0061] NLG 271 may use different language models and/or language templates to generate natural language outputs…generation of natural language outputs may be also personalized for each user).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras with Zhu using a user profile in order to facilitate convenience and efficiency when creating a dialogue with a known user.
The combination doesn’t specify tailoring the dialogue. However Porras teaches some examples of applications including multi-modal user interfaces…for example, Tur et al., PCT International Application Publication No. WO 2011/028833, entitled “Method and Apparatus for Tailoring Output of an Intelligent Automated Assistant to a User” ([0083]).
In the same field of endeavor, Tur et al. (Tur; US 20170061316 A1) discloses a method for tailoring the output of an intelligent automated assistant by conducting an interaction with a human user, collecting data about the user using a multimodal set of sensors positioned in a vicinity of the user, making a set of inferences about the user in accordance with the data, and tailoring an output to be delivered to the user in accordance with the set of inferences.
Tur discloses tailoring the dialogue ([0005] apparatus for tailoring the output of an intelligent automated assistant; [0013] adjustments can be applied to all users; [0056]-[0057] Fig 4 step 412, the output selection module 204 of the interaction management system 106 formulates an output responsive to the user's intent (e.g., directions to Bart's house)…In step 414, the output is adjusted in accordance with the user's preferences…this adjustment is applied to one or more of the following system actions: the pattern of assistance (e.g., the steps used to guide the user toward fulfilling his intent), the modality of the system output (e.g., speech, text, graphics, etc.), or the words that make up the system output (e.g., less formal language for younger and/or informally dressed users). For instance, if the user appears to be rushed, the output may be abbreviated) to a user profile ([0017] system inputs may include stored user data, such as a user profile).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras and Zhu with Tur using tailored dialogue in order to provide added knowledge and the ability to adapt to user preferences to enhance user experience as suggested by Tur ([0004]).
Regarding the new limitation of a user-specific subset of the contextual information filtered or selected based on profile attributes:
Porras teaches filtering contextual information based on profile attributes ([0091] input intent 436 is “notify me when a new web server appears on the network,”…system-generated NL dialog output in the form of a notification that “a new web server has connected to the network.” [the contextual information is filtered based on the profile of the user who specifically is looking for the appearance of new web servers]).
ZHU teaches also teaches filtering responses based on user profile ([0050] proactive agent 285 may generate candidate entities associated with the proactive task based on a user profile. The generation may be based on a straightforward backend query using deterministic filters to retrieve the candidate entities from a structured data store).
Regarding Claim 2, Porras discloses the collected sensor data is received from a plurality of sensors located in the environment ([0028] user interaction detection device(s) 106 may include the interactive display device 104 and/or other human activity detection devices (e.g., various types of sensors, including motion sensors, kinetic sensors, proximity sensors, thermal sensors, pressure sensors, force sensors, inertial sensors, cameras, microphones, gaze tracking systems, and/or others; [0036] network activity data 140 may be generated by, e.g., one or more network sensors or passive network monitoring programs; 602 of Fig 6A monitor network traffic).
2>Regarding Claims 3 and 17, Porras discloses the processor is further configured to: receive, from the first user, a user response to the user-specific remediation conversation, wherein the user response requests additional information pertaining to the security event ([0113] In block 616, the computing system 100 branches in one of two directions, depending on the interaction type. If the computing system 100 interprets the user interaction as a network exploration directive (e.g., a request to manipulate the view of the visualization), the computing system 100 branches to block 622); identify a subset of the plurality of sensors in the environment that generated data used to formulate the security event (sensors that detect events are identified sensors and are a part of a subset of sensors); collect additional sensor data from the subset of the plurality of sensors or from all of the plurality of sensors (602 of Fig 6A continues to monitor system); and output a second portion of the user-specific remediation conversation (60 of Fig 6A continues to update display).
3>Regarding Claims 5 and 19, Porras discloses the user response is one of: a verbal response (Fig 4B, [0092] the NL dialog output 448, which is output via, e.g., one or more speakers), a gesture ([0021]), a physical input ([0027] touchscreen), and an expression ([0070] facial expression).
3>Regarding Claim 6, Porras discloses the processor is further configured to: receive a second user response from the first user ([0071] interaction model 414 may be defined or personalized for specific types of users); and output a third portion of the user-specific remediation conversation, wherein portions of the user-specific remediation conversation are outputted until the security event is resolved or the first user ceases providing user responses (see Figs 4B, 6A for ongoing dialog; [0117] In block 670 update the graphical elements of the visualization 114 to indicate graphically in the visualization 114 that the node is now quarantined).
Regarding Claim 7, Porras discloses the user-specific remediation conversation for the first user is different from a second user-specific remediation conversation from the plurality of different remediation conversations associated with a second user having at least one of a different access rights or a different user interface preferences ([0071] interaction model 414 may be defined or personalized for specific types of users; [0075] handling subsystem 122 may perform authentication processes to verify a user's identity).
2>Regarding Claim 8, Porras discloses the first portion of the collected sensor data comprises first sensor data from one sensor of the plurality of sensors and wherein the second portion of the collected sensor data comprises second sensor data from one or more different sensors ([0028], [0036] describe a plurality of different sensors, all sensors are used to monitor for security events).
8>Regarding Claim 9, Porras discloses the contextual information comprises a cause of the security event, and wherein the processor is further configured to generate the contextual information for the security event based on the second portion of the collected sensor data by: determining the cause of the security event by comparing the second portion of the collected sensor data with activity templates comprising sensor data during historic activities and associated activity identifiers ([0038] network activity data 140 may include historical records of network activity and/or predictive models and/or predictive models (models make predictions based a determination that a current activity is comparatively close to a historic activity); [0052] the network topology data 220 may identify nodes…infection profile data 222 includes, for example, statistical information based on historical infection data, or other information which indicates typical patterns or behaviors of known infections); and retrieving the second portion of the collected sensor data from the one or more different sensors based on the cause of the security event ([0038]-[0039] security threats detected; [0111] 610 of Fig 6A, nodes or flows on the network visualization may be highlighted dynamically in response to the occurrence of network events or un-highlighted in response to the network events being remediated (e.g., by user interactions 120), 616, 622 (continues in loop of directives and monitoring); [0028], [0036] describe a plurality of different sensors, all sensors are used to monitor for security events).
9>Regarding Claim 10, Porras discloses the user-specific remediation conversation comprises options for resolving the security event, wherein different options are provided by the user-specific remediation conversation for different contextual information (616 of Fig 6A options of network exploration or security directive, similarly with 664 of Fig 6B; Fig 6A, 622 continues to update the display and continue the dialog regarding highlighted events).
Regarding Claim 11, Porras discloses the user interface preferences comprises a preferred medium of communication, wherein a medium includes audio, video, and/or physical feedback; wherein audio preferences include at least one of: a preferred language ([0080] natural language), a preferred voice output, or a preferred speech speed; wherein video preferences include at least one of: an appearance of a user interface where the user-specific remediation conversation is generated (446, 448 of Fig 4B, 604 of Fig 6A; Fig 6B), or a video quality; and wherein physical feedback preferences include at least one of: a touchscreen sensitivity of the output device ([0027]) touchscreen display), a vibration strength of the output device, or a haptic feedback sensitivity of the output device.
2>Regarding Claim 14, Porras discloses the security event represents a summary of monitored activity over a period of time from at least one sensor of the plurality of sensors (610 of Fig 6A summarizes highlight network events).
Regarding Claim 15, Porras discloses a method for providing personalized and contextualized environment security information (Abstract facilitate analysis of network activity or to respond to network security events), comprising:
detecting a security event in an environment based on a first portion of collected sensor data ([0028] user interaction detection device(s) 106 may include the interactive display device 104 and/or other human activity detection devices (e.g., various types of sensors, including motion sensors, kinetic sensors, proximity sensors, thermal sensors, pressure sensors, force sensors, inertial sensors, cameras, microphones, gaze tracking systems, and/or others; [0036] network activity data 140 may be generated by, e.g., one or more network sensors or passive network monitoring programs; 602 of Fig 6A monitor network traffic);
generating contextual information for the security event based on a second portion of the collected sensor data, wherein the second portion is different from the first portion ([0038]-[0039] security threats detected; [0111] 610 of Fig 6A, nodes or flows on the network visualization may be highlighted dynamically in response to the occurrence of network events or un-highlighted in response to the network events being remediated (e.g., by user interactions 120));
user interface device(s) 104 may be embodied as…a touchscreen display device…includes audio input and output devices capable of capturing and recording human conversational spoken natural language input and outputting system-generated conversational spoken natural language output (such as microphones, speakers and headphones or earbuds)),
generating a user-specific remediation conversation from a plurality of different remediation conversations ([0087], [0091]-[0092]) based on the user for instance new web server connecting), and a user-specific subset of the contextual information (Figs 6A-6B, 612 user dialog detected, 614 translate dialog to network directive, 664 respond with NL dialog; [0091] notify me when a new web server appears on the network,[ the subset is a specific scenario established by the user to notify of new web servers],)), wherein each of the plurality of different remediation conversations comprises dialogue ([0092] natural language generator (NLG) module 446 generates a natural language version of the dialog output intent 444, the NL dialog output 448, which is output via, e.g., one or more speakers [0091] notification that “a new web server has connected to the network.” [0087] confirmations that the system 110 is going to execute a user-requested command (e.g., “are you sure you want me to disconnect that node from the network? OK, disconnecting the node from the network”) interaction model 414 may be defined or personalized for specific types of users; [0091]-[0092] responds to specific requests made by the user)) response may include the phrase “Which <node> do you want to disconnect”; [0117] system may proceed to block 672 and respond by outputting NL dialog asking the user for further clarification of the request); and
outputting at least a first portion of the user-specific remediation conversation on the output device ([0092] the NL dialog output 448, which is output via, e.g., one or more speakers, displays, or other user interface and/or user interaction detection devices 104, 106…an NL response may include the phrase “Which <node> do you want to disconnect”), where <node> indicates a parameter*; notification that “a new web server has connected to the network.”; 664 of Fig 6B). Porras does NOT specify a user profile for the conversation nor tailoring the dialogue, but does teach identifying a user ([0075] handling subsystem 122 may perform authentication processes to verify a user's identity).
Zhu teaches a user profile indicative of access rights and user interface preferences ([0004] user profile may include demographic information, communication-channel information, and information on personal interests of the user; [0006] execute tasks that are relevant to user interests and preferences based on the user profile without a user input. In particular embodiments, the assistant system may check privacy settings to ensure that accessing a user's profile or other user information and executing different tasks are permitted subject to the user's privacy settings; [0039] assistant system 140 may proactively execute pre-authorized tasks that are relevant to user interests and preferences based on the user profile).
Zhu discloses generating a remediation conversation based on the user profile, the security event, and the contextual information, wherein the remediation conversation comprises dialogue for providing security information and guiding the first user to resolve the security event (Abstract, [0006] The assistant system may generate a response for the user regarding the information or services by using natural-language generation; [0029]; [0056] processing result may be stored in the user context engine 225 as part of the user profile. The online inference service 227 may analyze the conversational data associated with the user that are received by the assistant system 140 at a current time. The analysis result may be stored in the user context engine 225 also as part of the user profile…inference service 227 may extract personalization features from the plurality of data… semantic information aggregator 230 may then send the aggregated information to the NLU module 220 to facilitate the domain classification/selection; [0061] NLG 271 may use different language models and/or language templates to generate natural language outputs…generation of natural language outputs may be also personalized for each user).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras with Zhu using a user profile in order to facilitate convenience and efficiency when creating a dialogue with a known user.
The combination doesn’t specify tailoring the dialogue. However Porras teaches some examples of applications including multi-modal user interfaces…for example, Tur et al., PCT International Application Publication No. WO 2011/028833, entitled “Method and Apparatus for Tailoring Output of an Intelligent Automated Assistant to a User” ([0083]).
Tur discloses tailoring the dialogue ([0005] apparatus for tailoring the output of an intelligent automated assistant; [0013] adjustments can be applied to all users; [0056]-[0057] Fig 4 step 412, the output selection module 204 of the interaction management system 106 formulates an output responsive to the user's intent (e.g., directions to Bart's house)…In step 414, the output is adjusted in accordance with the user's preferences…this adjustment is applied to one or more of the following system actions: the pattern of assistance (e.g., the steps used to guide the user toward fulfilling his intent), the modality of the system output (e.g., speech, text, graphics, etc.), or the words that make up the system output (e.g., less formal language for younger and/or informally dressed users). For instance, if the user appears to be rushed, the output may be abbreviated) to a user profile ([0017] system inputs may include stored user data, such as a user profile).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras and Zhu with Tur using tailored dialogue in order to provide added knowledge and the ability to adapt to user preferences to enhance user experience as suggested by Tur ([0004]).
Regarding the new limitation of a user-specific subset of the contextual information filtered or selected based on profile attributes:
Porras teaches filtering contextual information based on profile attributes ([0091] input intent 436 is “notify me when a new web server appears on the network,”…system-generated NL dialog output in the form of a notification that “a new web server has connected to the network.” [the contextual information is filtered based on the profile of the user who specifically is looking for the appearance of new web servers]).
ZHU teaches also teaches filtering responses based on user profile ([0050] proactive agent 285 may generate candidate entities associated with the proactive task based on a user profile. The generation may be based on a straightforward backend query using deterministic filters to retrieve the candidate entities from a structured data store).
Regarding Claim 16, Porras discloses the collected sensor data is received from a plurality of sensors located in the environment ([0028] user interaction detection device(s) 106 may include the interactive display device 104 and/or other human activity detection devices (e.g., various types of sensors, including motion sensors, kinetic sensors, proximity sensors, thermal sensors, pressure sensors, force sensors, inertial sensors, cameras, microphones, gaze tracking systems, and/or others; [0036] network activity data 140 may be generated by, e.g., one or more network sensors or passive network monitoring programs; 602 of Fig 6A monitor network traffic).
Regarding Claim 20, Porras discloses a non-transitory computer-readable medium storing instructions, executable by a processor ([0136]), for performing a method for providing personalized and contextualized environment security information (Abstract facilitate analysis of network activity or to respond to network security events)), comprising:
detecting a security event in an environment based on a first portion of collected sensor data ([0028] user interaction detection device(s) 106 may include the interactive display device 104 and/or other human activity detection devices (e.g., various types of sensors, including motion sensors, kinetic sensors, proximity sensors, thermal sensors, pressure sensors, force sensors, inertial sensors, cameras, microphones, gaze tracking systems, and/or others; [0036] network activity data 140 may be generated by, e.g., one or more network sensors or passive network monitoring programs; 602 of Fig 6A monitor network traffic);
generating contextual information for the security event based on a second portion of the collected sensor data, wherein the second portion is different from the first portion ([0038]-[0039] security threats detected; [0111] 610 of Fig 6A, nodes or flows on the network visualization may be highlighted dynamically in response to the occurrence of network events or un-highlighted in response to the network events being remediated (e.g., by user interactions 120));
user interface device(s) 104 may be embodied as…a touchscreen display device…includes audio input and output devices capable of capturing and recording human conversational spoken natural language input and outputting system-generated conversational spoken natural language output (such as microphones, speakers and headphones or earbuds)),
generating a user-specific remediation conversation from a plurality of different remediation conversations ([0087], [0091]-[0092]) based on the user profile, the security event ([0091] for instance new web server connecting), and a user-specific subset of the contextual information (Figs 6A-6B, 612 user dialog detected, 614 translate dialog to network directive, 664 respond with NL dialog; [0091] notify me when a new web server appears on the network,[ the subset is a specific scenario established by the user to notify of new web servers]), wherein each of the plurality of different remediation conversations comprises dialogue ([0092] natural language generator (NLG) module 446 generates a natural language version of the dialog output intent 444, the NL dialog output 448, which is output via, e.g., one or more speakers; [0091] notification that “a new web server has connected to the network.” [0087] confirmations that the system 110 is going to execute a user-requested command (e.g., “are you sure you want me to disconnect that node from the network? OK, disconnecting the node from the network”)) tinteraction model 414 may be defined or personalized for specific types of users; [0091]-[0092] responds to specific requests made by the user) response may include the phrase “Which <node> do you want to disconnect”; [0117] system may proceed to block 672 and respond by outputting NL dialog asking the user for further clarification of the request); and
outputting at least a first portion of the user-specific remediation conversation on the output device ([0092] the NL dialog output 448, which is output via, e.g., one or more speakers, displays, or other user interface and/or user interaction detection devices 104, 106…an NL response may include the phrase “Which <node> do you want to disconnect”), where <node> indicates a parameter*; notification that “a new web server has connected to the network.”; 664 of Fig 6B). Porras does NOT specify a user profile for the conversation nor tailoring the dialogue, but does teach identifying a user([0075] handling subsystem 122 may perform authentication processes to verify a user's identity).
Zhu teaches a user profile indicative of access rights and user interface preferences ([0004] user profile may include demographic information, communication-channel information, and information on personal interests of the user; [0006] execute tasks that are relevant to user interests and preferences based on the user profile without a user input. In particular embodiments, the assistant system may check privacy settings to ensure that accessing a user's profile or other user information and executing different tasks are permitted subject to the user's privacy settings; [0039] assistant system 140 may proactively execute pre-authorized tasks that are relevant to user interests and preferences based on the user profile).
Zhu discloses generating a remediation conversation based on the user profile, the security event, and the contextual information, wherein the remediation conversation comprises dialogue for providing security information and guiding the first user to resolve the security event (Abstract, [0006] The assistant system may generate a response for the user regarding the information or services by using natural-language generation; [0029]; [0056] processing result may be stored in the user context engine 225 as part of the user profile. The online inference service 227 may analyze the conversational data associated with the user that are received by the assistant system 140 at a current time. The analysis result may be stored in the user context engine 225 also as part of the user profile…inference service 227 may extract personalization features from the plurality of data… semantic information aggregator 230 may then send the aggregated information to the NLU module 220 to facilitate the domain classification/selection; [0061] NLG 271 may use different language models and/or language templates to generate natural language outputs…generation of natural language outputs may be also personalized for each user).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras with Zhu using a user profile in order to facilitate convenience and efficiency when creating a dialogue with a known user.
The combination doesn’t specify tailoring the dialogue. However Porras teaches some examples of applications including multi-modal user interfaces…for example, Tur et al., PCT International Application Publication No. WO 2011/028833, entitled “Method and Apparatus for Tailoring Output of an Intelligent Automated Assistant to a User” ([0083]).
Tur discloses tailoring the dialogue ([0005] apparatus for tailoring the output of an intelligent automated assistant; [0013] adjustments can be applied to all users; [0056]-[0057] Fig 4 step 412, the output selection module 204 of the interaction management system 106 formulates an output responsive to the user's intent (e.g., directions to Bart's house)…In step 414, the output is adjusted in accordance with the user's preferences…this adjustment is applied to one or more of the following system actions: the pattern of assistance (e.g., the steps used to guide the user toward fulfilling his intent), the modality of the system output (e.g., speech, text, graphics, etc.), or the words that make up the system output (e.g., less formal language for younger and/or informally dressed users). For instance, if the user appears to be rushed, the output may be abbreviated) to a user profile ([0017] system inputs may include stored user data, such as a user profile).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras and Zhu with Tur using tailored dialogue in order to provide added knowledge and the ability to adapt to user preferences to enhance user experience as suggested by Tur ([0004]).
Regarding the new limitation of a user-specific subset of the contextual information filtered or selected based on profile attributes:
Porras teaches filtering contextual information based on profile attributes ([0091] input intent 436 is “notify me when a new web server appears on the network,”…system-generated NL dialog output in the form of a notification that “a new web server has connected to the network.” [the contextual information is filtered based on the profile of the user who specifically is looking for the appearance of new web servers]).
ZHU teaches also teaches filtering responses based on user profile ([0050] proactive agent 285 may generate candidate entities associated with the proactive task based on a user profile. The generation may be based on a straightforward backend query using deterministic filters to retrieve the candidate entities from a structured data store).
Regarding Claim 21, Porras discloses a method for providing personalized and contextualized environment security information (Abstract facilitate analysis of network activity or to respond to network security events), comprising:
executing one or more machine learning models ([0071] rules, templates, and/or classifiers of the non-verbal interaction model 414 may be predefined, developed based on experimentation/observation, or learned by applying e.g., machine learning techniques to training data) that individually or in combination:
determine contextual information (including network sensors and the sensors detecting a user) for a security event ([0038] The network analytics subsystem 142 generates data indicative of a past, present, or future network context (e.g., current network context 144) and, particularly when an infection or threat is detected, one or more network event indicators 146) based on input sensor data ([0038]-[0039] security threats detected; [0111] 610 of Fig 6A, nodes or flows on the network visualization may be highlighted dynamically in response to the occurrence of network events or un-highlighted in response to the network events being remediated (e.g., by user interactions 120)), wherein the contextual information comprises a cause of the security event ([0038] analyze the network activity data 140 over time to determine network flow characteristics and node behaviors that may indicate the existence of a network infection or some other type of network threat);
identify a plurality of remediation actions for resolving the security event based on the contextual information ([0097]-[0099] security initiative 124 may comprise a high level directive corresponding to a gesture to “quarantine that node,” the network-executable actions 132 produced by the security initiative translator module 510…security initiative translator module 510 may resolve the higher-level network security directives using a pre-defined set of templates, rules, or policies, which may include, for example, “block,” “deny,” “allow,” “redirect,” “quarantine,” “undo,” “constrain,” and/or “info” directives);
output a user remediation conversation comprising dialogue ([0087], [0091]-[0092] the NL dialog output 448, which is output via, e.g., one or more speakers, displays, or other user interface and/or user interaction detection devices 104, 106…an NL response may include the phrase “Which <node> do you want to disconnect”), where <node> indicates a parameter*; notification that “a new web server has connected to the network.”;664 of Fig 6B) visualization 700 to present the user with remediation options; Figs 6A-6B, 612 user dialog detected, 614 translate dialog to network directive, 664 respond with NL dialog; [0087], [0091]-[0092]);
receive, in the user remediation conversation, for example, “block,” “deny,” “allow,” “redirect,” “quarantine,” “undo,” “constrain,” and/or “info” directives); and
execute the selection of the at least one remediation action ([0099] A “block” directive may, for example, cause the system 110 to implement a full duplex filter). Porras does NOT specify a user profile for the conversation nor tailoring the dialogue, but does teach identifying a user ([0075] handling subsystem 122 may perform authentication processes to verify a user's identity).
Zhu teaches a user profile indicative of access rights and user interface preferences ([0004] user profile may include demographic information, communication-channel information, and information on personal interests of the user; [0006] execute tasks that are relevant to user interests and preferences based on the user profile without a user input. In particular embodiments, the assistant system may check privacy settings to ensure that accessing a user's profile or other user information and executing different tasks are permitted subject to the user's privacy settings; [0039] assistant system 140 may proactively execute pre-authorized tasks that are relevant to user interests and preferences based on the user profile).
Zhu discloses generating a remediation conversation based on the user profile, the security event, and the contextual information, wherein the remediation conversation comprises dialogue for providing security information and guiding the first user to resolve the security event (Abstract, [0006] The assistant system may generate a response for the user regarding the information or services by using natural-language generation; [0029]; [0056] processing result may be stored in the user context engine 225 as part of the user profile. The online inference service 227 may analyze the conversational data associated with the user that are received by the assistant system 140 at a current time. The analysis result may be stored in the user context engine 225 also as part of the user profile…inference service 227 may extract personalization features from the plurality of data… semantic information aggregator 230 may then send the aggregated information to the NLU module 220 to facilitate the domain classification/selection; [0061] NLG 271 may use different language models and/or language templates to generate natural language outputs…generation of natural language outputs may be also personalized for each user).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras with Zhu using a user profile in order to facilitate convenience and efficiency when creating a dialogue with a known user.
The combination doesn’t specify tailoring the dialogue. However Porras teaches some examples of applications including multi-modal user interfaces…for example, Tur et al., PCT International Application Publication No. WO 2011/028833, entitled “Method and Apparatus for Tailoring Output of an Intelligent Automated Assistant to a User” ([0083]).
Tur discloses tailoring the dialogue ([0005] apparatus for tailoring the output of an intelligent automated assistant; [0013] adjustments can be applied to all users; [0056]-[0057] Fig 4 step 412, the output selection module 204 of the interaction management system 106 formulates an output responsive to the user's intent (e.g., directions to Bart's house)…In step 414, the output is adjusted in accordance with the user's preferences…this adjustment is applied to one or more of the following system actions: the pattern of assistance (e.g., the steps used to guide the user toward fulfilling his intent), the modality of the system output (e.g., speech, text, graphics, etc.), or the words that make up the system output (e.g., less formal language for younger and/or informally dressed users). For instance, if the user appears to be rushed, the output may be abbreviated) to a user profile ([0017] system inputs may include stored user data, such as a user profile).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras and Zhu with Tur using tailored dialogue in order to provide added knowledge and the ability to adapt to user preferences to enhance user experience as suggested by Tur ([0004]).
Regarding the new limitation of a user-specific subset of the contextual information filtered or selected based on profile attributes:
Porras teaches filtering contextual information based on profile attributes ([0091] input intent 436 is “notify me when a new web server appears on the network,”…system-generated NL dialog output in the form of a notification that “a new web server has connected to the network.” [the contextual information is filtered based on the profile of the user who specifically is looking for the appearance of new web servers]).
ZHU teaches also teaches filtering responses based on user profile ([0050] proactive agent 285 may generate candidate entities associated with the proactive task based on a user profile. The generation may be based on a straightforward backend query using deterministic filters to retrieve the candidate entities from a structured data store).
Regarding Claim 24, Porras discloses the security event is detected in an environment, and wherein the user remediation conversation further comprises functional calls for performing daily operational procedure associated with the environment ([0053] reputation data 224 may be updated regularly (e.g., daily)).
Zhu teaches an agenda item may comprise a recurring item such as a daily digest ([0049]).
24>Regarding Claim 25, Porras discloses the functional calls enable a user to search the contextual information by requesting one or more of: video data, map data, event data ([0029] examples of network exploration directives 118 involve querying the system 110 for specific data, for example, to request that the visualization 114 display additional details about the current behavior of a network flow or node), photos, and personnel information.
24>Regarding Claim 26, Porras discloses the functional calls enable a user to edit security settings comprising one or more of: managing alerts ([0020] generate an interactive display for a human user, based on those alerts, thereby presenting a real-time visual depiction of the network, and of the current activity, flows, and cyber-threats. Components of the system 110 are designed to conduct conversational natural language dialog with a human user, including to receive natural language requests for one or more desired courses of action to remediate network threats), managing alarms, and generating emergency protocols ([0097]-[0099] security initiative 124 may comprise a high level directive corresponding to a gesture to “quarantine that node,” the network-executable actions 132 produced by the security initiative translator module 510…security initiative translator module 510 may resolve the higher-level network security directives using a pre-defined set of templates, rules, or policies, which may include, for example, “block,” “deny,” “allow,” “redirect,” “quarantine,” “undo,” “constrain,” and/or “info” directives).
Regarding Claim 27, Porras discloses at least one other remediation action from the plurality of remediation actions is presented in different dialogue visualization 700 to present the user with remediation options; Figs 6A-6B, 612 user dialog detected, 614 translate dialog to network directive, 664 respond with NL dialog, [0071] interaction model 414 may be defined or personalized for specific types of users; [0075] handling subsystem 122 may perform authentication processes to verify a user's identity).
Zhu discloses remediation action from remediation actions in different dialogue for a different input user profile ([0056] processing result may be stored in the user context engine 225 as part of the user profile. The online inference service 227 may analyze the conversational data associated with the user that are received by the assistant system 140 at a current time. The analysis result may be stored in the user context engine 225 also as part of the user profile).
Tur discloses tailoring the dialogue ([0005] apparatus for tailoring the output of an intelligent automated assistant).
Regarding Claim 28, Porras discloses the user remediation conversation is output in an audio-format ([0027] user interface device(s) 104 includes audio input and output devices capable of capturing and recording human conversational spoken natural language input and outputting system-generated conversational spoken natural language output).
Regarding Claim 29, Porras discloses user responses in the user remediation conversation are provided using one or more of: voice commands ([0099]), text, or gestures ([0070] text, gestures).
Regarding Claim 30, Porras discloses the dialogue is tailored based on the user interface preferences comprising a preferred medium of communication, wherein a medium includes audio, video, and/or physical feedback; wherein audio preferences include at least one of: a preferred language ([0080] natural language), a preferred voice output, or a preferred speech speed; wherein video preferences include at least one of: an appearance of a user interface where the user remediation conversation is generated (446, 448 of Fig 4B, 604 of Fig 6A), or a video quality; and wherein physical feedback preferences include at least one of: a touchscreen sensitivity of an output device ([0027]) touchscreen display), a vibration strength of the output device, or a haptic feedback sensitivity of the output device (XZ).
Tur discloses tailoring the dialogue ([0005] apparatus for tailoring the output of an intelligent automated assistant).
Claims 4 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Porras, Zhu and Tur, further in view of Bar-Nahum et al. (Bar-Nahum; US 20200162489 A1).
3>Regarding Claims 4 and 18, Porras doesn’t teach tracking an object, but does teach tracking capabilities ([0028]).
In the same field of endeavor, Bar-Nahum discloses a security system wherein once an indication of a detected security event is received, one or more sensors are selected based on the detected security event. The selected sensors are used to detect additional information associated with a protected airspace associated with the detected security event.
Bar-Nahum discloses the user response further requests performing object tracking on an object in the environment, wherein the processor is further configured to output the second portion of the user-specific remediation conversation by: identifying the object using the additional sensor data; collecting the additional sensor data until the object is no longer detected; and generating an image of the object for display ([0063]).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras with Bar-Nahum using object tracking in order to provide a swift response to the actual location of the security event and prevent or reduce the amount of property damage, loss of human life, or prevent criminal activity, as suggested by Bar-Nahum ([0002]).
Claim 12 is rejected under 35 U.S.C. 103 as being unpatentable over Porras, Zhu and Tur, further in view of Levy et al. (Levy; US 20190319987 A1).
Regarding Claim 12, Porras discloses the security event is indicative of a deviation from a known trend in the environment, wherein the processor is further configured to: execute a machine learning algorithm to classify whether input sensor data comprises the deviation ([0052] infection profile data 222 includes, for example, statistical information based on historical infection data, or other information which indicates typical patterns (trend) or behaviors of known infections (deviations from trends)), but doesn’t specify comparing vectors.
In the same field of endeavor, Levy discloses an interface for a threat management facility of an enterprise network supports the use of third-party security products within the enterprise network by providing access to relevant internal instrumentation and a programmatic interface for direct or indirect access to local security agents on compute instances within the enterprise network .
Levy discloses generating a feature vector representing the known trend in historic sensor data; comparing the feature vector against an input feature vector of the input sensor data; and classifying the input sensor data as comprising the deviation in response to determining, based on the comparing, that a difference between the feature vector and the input feature vector is greater than a threshold difference ([0193] the entity model may characterize a baseline of expected events derived from on events detected from the entity over an historical window, and may be expressed, e.g., as a vector in an event vector space or any other suitable representation for making comparisons to new event vectors in the event stream 1604; Fig 16).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras with Levy using vectors in order to improve accuracy and efficiency for security in an enterprise network, as suggested by Levy (Abstract, [0003]).
Claim 13 is rejected under 35 U.S.C. 103 as being unpatentable over Porras, Zhu and Tur, further in view of MALHOTRA et al. (Malhotra; US 20160189509 A1).
2>Regarding Claim 13, Porras discloses the security event is indicative of an inconsistency ([0052] infection profile data 222 includes, for example, statistical information based on historical infection data, or other information which indicates typical patterns or behaviors of known infections (inconsistency)), but doesn’t teach overriding a sensor.
In the same field of endeavor, Malhotra discloses a system for learned overrides for home security. A sensor of a security system may be armed. A trip signal may be received indicating a tripping of the sensor. It may be determined that the trip signal can be automatically overridden based on matching an identity of the sensor and a state of the security system with a pattern in a model. The pattern may represent a state of the security system in which automatically overriding the trip signal from the sensor is permitted. The trip signal from the sensor may be automatically overridden without input from a user.
Malhotra discloses between at least two sensors, wherein the processor is further configured to: receive a first sensor output from a first sensor in the environment, and a second sensor output from a second sensor in the environment; and determine, based on historical sensor data, that the first sensor should not output the first sensor output when the second sensor outputs the second sensor output ([0024] trip signal may be received indicating a tripping of the sensor. It may be determined that the trip signal can be automatically overridden based on matching an identity of the sensor and a state of the security system with a pattern in a model).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras with Malhotra using an override in order to create a convenient tailored security system to the preferences of a user based on gathered patterns, as suggested by Malhotra ([0001]).
Claims 22-23 are rejected under 35 U.S.C. 103 as being unpatentable over Porras and Zhu further in view of Levy.
Regarding Claim 22, Porras discloses the one or more machine learning models ([0071] interaction model 414 may be defined or personalized for specific types of users) are trained to determine the contextual information using a training dataset ([0071] rules, templates, and/or classifiers of the non-verbal interaction model 414 may be predefined, developed based on experimentation/observation, or learned by applying e.g., machine learning techniques to training data), but doesn’t teach using vectors.
Levy discloses a plurality of input vectors each comprising an activity template and a pre-determined cause ([0193] the entity model may characterize a baseline of expected events derived from on events detected from the entity over an historical window, and may be expressed, e.g., as a vector in an event vector space or any other suitable representation for making comparisons to new event vectors in the event stream 1604; Fig 16), wherein the activity template comprises sensor values from a plurality of sensors ([0177] first entity model may be a model characterizing a pattern of events expected from the number of sensors in a vector space).
Therefore, it would have been obvious to a person having ordinary skill in the art before the effective filing date of the claimed invention to modify Porras with Levy using vectors in order to improve accuracy and efficiency for security in an enterprise network, as suggested by Levy (Abstract, [0003]).
22>Regarding Claim 23, Porras discloses the plurality of sensors include different types of sensors ([0028] user interaction detection device(s) 106 may include the interactive display device 104 and/or other human activity detection devices (e.g., various types of sensors, including motion sensors, kinetic sensors, proximity sensors, thermal sensors, pressure sensors, force sensors, inertial sensors, cameras, microphones, gaze tracking systems, and/or others; [0036] network activity data 140 may be generated by, e.g., one or more network sensors or passive network monitoring programs; 602 of Fig 6A monitor network traffic).
Levy discloses different sensors ([0177] first entity model may be a model characterizing a pattern of events expected from the number of sensors in a vector space).
Response to Arguments
Applicant's arguments filed 6/16/26 have been fully considered but they are not persuasive for the following reasons:
Arguments:
A. Applicant argues that the Office Action relies on Porras's natural language dialog subsystem for this limitation, citing Figs. 6A-6B and paragraphs [0071] and [0092]. Office Action at pp. 7-8. However, Porras discloses a system that dynamically generates a single dialog output using a natural language generator (NLG) module 446, which maps dialog output intents 444 to "one or more predefined NL response rules or templates" and outputs the result via speakers or displays. Porras at [0092]. Porras does not disclose, teach, or suggest a plurality of different remediation conversations, each already comprising tailored dialogue for a different user profile, from which a user-specific conversation is selected.
It is respectfully submitted that Porras teaches generating a user-specific remediation conversation from a plurality of different remediation conversations ([0087], [0091]-[0092]) based on the user for instance new web server connecting), and a user-specific subset of the contextual information (Figs 6A-6B, 612 user dialog detected, 614 translate dialog to network directive, 664 respond with NL dialog; [0091] notify me when a new web server appears on the network,[ the subset is a specific scenario established by the user to notify of new web servers],), wherein each of the plurality of different remediation conversations comprises dialogue ([0092] natural language generator (NLG) module 446 generates a natural language version of the dialog output intent 444, the NL dialog output 448, which is output via, e.g., one or more speakers [0091] notification that “a new web server has connected to the network.” [0087] confirmations that the system 110 is going to execute a user-requested command (e.g., “are you sure you want me to disconnect that node from the network? OK, disconnecting the node from the network”) interaction model 414 may be defined or personalized for specific types of users; [0091]-[0092] responds to specific requests made by the user) response may include the phrase “Which <node> do you want to disconnect”; [0117] system may proceed to block 672 and respond by outputting NL dialog asking the user for further clarification of the request)
Porras teaches a plurality of different remediation conversations to choose from ([0087], [0091]-[0092]). Tur teaches tailoring responses ([0005]).
B. Applicant argues that Porras's system generates a single conversational exchange in real time using NL dialog output intents in combination with course-of-action reasoning. Porras at [0088]-[0092]. The Office Action's reliance on paragraph [0071], which states the "non-verbal interaction model 414 may be defined or personalized for specific types of users," is misplaced because paragraph [0071] pertains to the interpretation of non-verbal user inputs (gestures, combined interaction data), not to the generation of remediation dialog output. Porras at [0071]. There is no disclosure in Porras that different pre-constructed (IN CLAIM??) remediation conversations exist, let alone that the system selects among them based on a user profile. The Examiner appears to acknowledge this deficiency by inserting "XXX?" in the Office Action adjacent to this limitation. Office Action at p. 8.
It is respectfully submitted that the claim language doesn’t preclude the conversation exchange of Porras from reading on the limitations.
In response to applicant's argument that the references fail to show certain features of applicant’s invention, it is noted that the features upon which applicant relies (i.e., preconstructed) are not recited in the rejected claim(s). Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993).
Regarding arguments against para [0071], the section is used to demonstrate user influence on the system, before generating responses.
C. Applicant argues that Zhu does not remedy this deficiency. Zhu discloses a contextual auto-completion system for an assistant that generates suggested auto-completions based on a personalized language model. Zhu at Abstract. Zhu does not disclose a plurality of different remediation conversations, each tailored to a different user profile, from which the system selects. Zhu at [0039], [0056]. Tur similarly fails to cure this gap. Tur teaches adjusting output format (e.g., modality, word choice, pattern of assistance) based on inferred user characteristics. Tur at [0057]. But tailoring the manner of presentation is not the same as maintaining and selecting from a plurality of substantively different conversations, each directed to a different user profile.
It is respectfully submitted that Zhu teaches a user profile indicative of access rights and user interface preferences ([0004] user profile may include demographic information, communication-channel information, and information on personal interests of the user; [0006] execute tasks that are relevant to user interests and preferences based on the user profile without a user input. In particular embodiments, the assistant system may check privacy settings to ensure that accessing a user's profile or other user information and executing different tasks are permitted subject to the user's privacy settings; [0039] assistant system 140 may proactively execute pre-authorized tasks that are relevant to user interests and preferences based on the user profile).
Zhu discloses generating a remediation conversation based on the user profile, the security event, and the contextual information, wherein the remediation conversation comprises dialogue for providing security information and guiding the first user to resolve the security event (Abstract, [0006] The assistant system may generate a response for the user regarding the information or services by using natural-language generation; [0029]; [0056] processing result may be stored in the user context engine 225 as part of the user profile. The online inference service 227 may analyze the conversational data associated with the user that are received by the assistant system 140 at a current time. The analysis result may be stored in the user context engine 225 also as part of the user profile…inference service 227 may extract personalization features from the plurality of data… semantic information aggregator 230 may then send the aggregated information to the NLU module 220 to facilitate the domain classification/selection; [0061] NLG 271 may use different language models and/or language templates to generate natural language outputs…generation of natural language outputs may be also personalized for each user).
In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
D. Applicant argues that none of the cited references disclose or suggest generating the conversation based on "a user-specific subset of the contextual information." The claims require that contextual information is filtered on a per-user basis (i.e., different users receive different subsets of contextual information in their respective conversations). Porras provides the same network context 144 and visualization 114 to all users. Porras at [0026], [0044]-[0045]. Zhu stores a user profile comprising personalization features, but does not disclose selecting a subset of security- event contextual information based on a user profile. Tur tailors the output's delivery form but does not filter underlying contextual content by user. Tur at [0056]-[0057].
It is respectfully submitted that Porras teaches at least three different responses which depend on the what an individual user is looking for, not a single generic response for all users.
In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
E. Applicant argues that The Office Action relies on Porras paragraph [0071], citing the statement that "rules, templates, and/or classifiers of the non-verbal interaction model 414 may be . . . learned by applying e.g., machine learning techniques to training data." Office Action at p. 22. However, Porras's machine learning disclosure is limited exclusively to the non-verbal interaction model 414, which is used to interpret user input gestures and combined interaction data. Porras at [0071]. Porras does not disclose machine learning models that determine a cause of a security event, identify remediation actions, output tailored dialogue, receive a user selection, or execute a remediation action (HENCE THE 103??). These are separate and distinct functions that the claims require to be performed by the ML models individually or in combination.
It is respectfully submitted that Porras teaches determining the intended goal or objective of a user's natural language dialog can involve the application of artificial-intelligence based automated reasoning methods and systems. ([0081]). And the COA (course of action) reasoning module 440 may determine by executing various task flows, analyzing the intent history, and/or conducting other automated (e.g., artificial intelligence-based) reasoning activities, that an appropriate dialog output intent 442 is system-generated NL dialog output in the form of a notification that “a new web server has connected to the network.” [0091]. Artifical-intelligence employs machine learning to learn inputs and determine responses, therefore, Porras reads on the claim language.
F. Applicant argues that The Office Action conflates Porras's use of ML for gesture interpretation with the comprehensive ML pipeline recited in claim 21. In Porras, the identification of remediation actions (e.g., "block," "quarantine") is performed by the security initiative translator module 510 using "a pre-defined set of templates, rules, or policies." Porras at [0099]. The dialog output is generated by the NLG module 446 using NL response rules or templates. Porras at [0092]. The execution of directives is performed by sending flow rules to network switches. Porras at [0030]. None of these operations involve machine learning models. Porras teaches a pipeline of distinct rule-based modules, not one or more ML models performing the claimed functions individually or in combination.
It is respectfully submitted that Porras teaches determining the intended goal or objective of a user's natural language dialog can involve the application of artificial-intelligence based automated reasoning methods and systems. ([0081]). And the COA (course of action) reasoning module 440 may determine by executing various task flows, analyzing the intent history, and/or conducting other automated (e.g., artificial intelligence-based) reasoning activities, that an appropriate dialog output intent 442 is system-generated NL dialog output in the form of a notification that “a new web server has connected to the network.” [0091]. Artifical-intelligence employs machine learning to learn inputs and determine responses, therefore, Porras reads on the claim language.
G. Applicant argues that Neither Zhu nor Tur cures this deficiency. Zhu discloses a personalized language model based on a recurrent neural network for generating auto-completion suggestions. Zhu at [0069]. This model predicts text entries for user convenience; it does not determine security event causes, identify remediation actions, or execute selected actions. Tur discloses using classifiers to determine user characteristics and affective state for tailoring output. Tur at [0046]-[0047]. Tur's classifiers do not determine security event causes, identify remediation actions, or execute selected remediation actions.
In response to applicant's arguments against the references individually, one cannot show nonobviousness by attacking references individually where the rejections are based on combinations of references. See In re Keller, 642 F.2d 413, 208 USPQ 871 (CCPA 1981); In re Merck & Co., 800 F.2d 1091, 231 USPQ 375 (Fed. Cir. 1986).
H. Applicant argues the Office Action maps "access rights" to Zhu's privacy settings and user preferences. Office Action at pp. 23-24. However, Zhu's privacy settings control whether the assistant system may access or store a user's data; they are data-privacy controls, not functional access rights that determine which security remediation actions a user is authorized to view or execute. Zhu at [0006] ("check privacy settings to ensure that accessing a user's profile or other user information and executing different tasks are permitted subject to the user's privacy settings"). Claim 21 requires that the dialogue presents at least one remediation action based on the access rights – meaning different users see different remediation options depending on their authorization level. This is a fundamentally different concept from data-privacy gating.
It is respectfully submitted that Zhu is used to teach a user profile, but does include access rights as well based on privacy settings [0006].
I. Applicant argues Porras discloses presenting remediation options (e.g., shun, re-route, quarantine) to a user via visualization 700. Porras at [0118], Fig. 7A. However, Porras does not disclose filtering which remediation actions are presented based on a user's access rights. Porras authenticates users (Porras at [0075]) but does not teach that different users are presented with different remediation options based on their authorization level.
Because the combination fails to teach the claimed ML-model-driven pipeline and the access-rights-based filtering of remediation actions, the rejection of claim 21 should be withdrawn. All claims depending from claim 21 are patentable for at least the same reasons.
In response to applicant's argument that the references fail to show certain features of applicant’s invention, it is noted that the features upon which applicant relies (i.e., filtering) are not recited in rejected claim 21. Although the claims are interpreted in light of the specification, limitations from the specification are not read into the claims. See In re Van Geuns, 988 F.2d 1181, 26 USPQ2d 1057 (Fed. Cir. 1993).
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 extension fee 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 date of this final action.
Any inquiry concerning this communication or earlier communications from the examiner should be directed to MARK S RUSHING whose telephone number is (571)270-5876. The examiner can normally be reached on 10-6pm.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Davetta Goins can be reached at 571-272-2957. 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.
/MARK S RUSHING/Primary Examiner, Art Unit 2689
Prosecution Insights