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 .
Claims 1-18 are presented for examination.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claim(s) 1-9 and 18 is/are rejected under 35 U.S.C. 103 as being unpatentable over Aubineau (US 2022/0253607 A1) in view of Leeds (US 11,431,660 B1), and further in view of Vibbert (US 2016/0042735 A1).
Regarding Claim 1, Aubineau (US 2022/0253607 A1) teach
A computer-implemented method for creating an application, comprising:
obtaining, by at least one processor, configuration information of a first processing entity, the first processing entity being associated with the application, the application being associated with a plurality of processing entities comprising the first processing entity, the configuration information indicating an action to be performed by the first processing entity in response to a determination that the first processing entity is not suitable for processing a request; (Para. [0039]-[0041], “In step 302, the user annotates the Web API specification(s) for analysis by the system. For each Web API to be used by the chatbot, a human annotates it with a description of the APIs function. The user also annotates each of the needed inputs of Web API with a question that defines how to retrieve this input…In step 303, the system generates an action catalog from the annotated Web API specification…In step 304, the system receives the annotated Web API specification and automatically generates a rule set…each rule in the rule set corresponds to a Web API or a portion of the Web API and each rule condition corresponds to an input to the Web API. These rules can be processed at runtime by a goal-based reasoning engine.”) Examiner Comments: Aubineau teach obtaining configuration information (the user-annotated Web API specification, action catalog, and rule set) for a chatbot processing entity, wherein that configuration information defines the action rules the entity follows when processing requests, and further teaches in Para. [0046] that a defined action (displaying a fallback message or requesting clarification) is performed when the entity determines it cannot process the user request, reading on the claimed configuration information indicating an action to be performed in response to a determination that the first processing entity is not suitable for processing the request under the broadest reasonable interpretation.
Aubineau did not specifically teach
creating the application based on the configuration information; and the application being associated with a plurality of processing entities.
However, Leeds (US 11,431,660 B1) teach
the application being associated with a plurality of processing entities comprising the first processing entity, and creating the application based on the configuration information; (Col. 1, lines 34-53, “CCAIs are modular, distributed, scalable, and capable of incremental improvement and evolutionary extension, via growth, updates, and replacement, and through utilizing self-play for honing, including, through the present invention, selecting, aggregating, and pruning sets of asymmetrical chatbots, and optimizing configurations for achieving collaborative goals. In addition, CCAIs provide for reusable components (for example, chatbot subminds trained on specialized content, trained for specialized interactions, or capable of interacting with specialized devices).”) Examiner Comments: Leeds teach a Collaborative Conversational AI (CCAI) application that is created from and associated with a plurality of independent processing entities (subminds), each being a specialized chatbot or AI agent, reading on the claimed application being associated with a plurality of processing entities and created based on the configuration information.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Leeds’s multi-submind architecture into Aubineau’s configuration-based chatbot creation system in order to improve coverage and reliability by distributing processing across multiple independently configured agents. Both references are in the same field of conversational AI system design, and the advantages of using a plurality of specialized processing entities to handle a broader range of user requests would have provided a reasonable expectation of success.
Aubineau and Leeds did not specifically teach
the plurality of processing entities having jump relationships between respective processing entities represented by a data structure having a start processing entity and at least one further processing entity.
However, Vibbert (US 2016/0042735 A1) teach
the plurality of processing entities having jump relationships between respective processing entities represented by a data structure having a start processing entity and at least one further processing entity; (Para. [0057], “Task specifications 312, 314, and 316 describing task specific behavior may be modeled with tree diagrams of task subcomponents…Dialog task specifications 314, 324, and 334 may comprise dialog agents and dialog agencies, which may each be independent program subroutines”; Para. [0061], “Root node 402 may be the topmost node in the task specification layer…the root node may control the execution of children nodes 410, 412, and 414…Each node of the task tree 400 may be a dialog agency or a dialog agent. Terminating nodes 414, 420, 422, 430, and 432 of task tree 400 may be dialog agents and non-terminating nodes 410, 412, and 424 may be dialog agencies”; Para. [0081], “the user can also take initiative and shift the focus of the conversation to another part of the dialog task tree 500.”) Examiner Comments: Vibbert teach that a plurality of processing entities (dialog agents and dialog agencies, each an independent program subroutine) are represented as the nodes of a tree diagram data structure having a root node (the claimed start processing entity) and children nodes (the claimed at least one further processing entity), and that the conversation shifts focus, that is, jumps, from one node to another part of the tree, reading on the claimed plurality of processing entities having jump relationships between respective processing entities represented by a data structure having a start processing entity and at least one further processing entity.
It would further have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to represent the plurality of processing entities and their inter-entity relationships of the Aubineau and Leeds system using the hierarchical tree data structure of Vibbert, in which a root node (start processing entity) and children nodes (further processing entities) correspond to dialog agents and dialog agencies. One of ordinary skill would have been motivated to do so because Vibbert teaches that organizing the plurality of dialog agents and agencies as nodes of a task tree provides an orderly, navigable representation of the permissible transitions among the entities and allows the dialog engine to traverse the tree and shift the focus of the conversation from one entity to another part of the tree (Vibbert, Para. [0080]-[0081]), thereby yielding the predictable benefit of structured and controllable routing of a request from a starting entity to further entities in a multi-entity conversational system.
Regarding Claim 2, Aubineau, Leeds, and Vibbert teach
The method of Claim 1.
Leeds further teaches
wherein the action indicates jumping to a second processing entity of the plurality of processing entities to process the request, the second processing entity being determined based on an entity jump history of the application (Col. 4, lines 1-14, “Bots may be evolutionary capable, cyclical or stable, wherein a bot may enable incremental changes to its decision mechanism (i.e., exhibiting persuadability) via incremental or extensible neural networks, extensible grammars or parsers, recording of experience and decision data, decision tree creation and extension, or adjustment of response mechanism sliders”; Col. 1, lines 53-60, “The disclosed invention also provides for a non-reductive representation of decision making and experience, and reconveyance of responses based on context.”) Examiner Comments: Leeds teach that when a current submind (first processing entity) is unable to adequately process a request, the forum routes the request to another submind (second processing entity) based on a recorded history of decisions and experience data capturing the prior routing sequence (entity jump history), reading on the claimed action of jumping to a second processing entity determined based on an entity jump history.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Leeds's multi-submind architecture into Aubineau's configuration-based chatbot creation system in order to improve coverage and reliability by distributing processing across multiple independently configured agents. Both references are in the same field of conversational AI system design, and the advantages of using a plurality of specialized processing entities to handle a broader range of user requests.
Regarding Claim 3, Aubineau, Leeds, and Vibbert teach
The method of Claim 1.
Vibbert further teaches
wherein the action indicates jumping to the starting processing entity of the plurality of processing entities to process the request (Para. [0062], “Root node 402 may be the topmost node in the task specification layer…the root node may control the execution of children nodes 410, 412, and 414”; Para. [0081], “the user can also take initiative and shift the focus of the conversation to another part of the dialog task tree 500.”) Examiner Comments: Vibbert teach that the conversation may shift focus among the nodes of the task tree, including to the root node (the starting processing entity) that is the topmost node controlling the children nodes, reading under the broadest reasonable interpretation on the claimed action of jumping to the starting processing entity of the plurality of processing entities.
It would further have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to represent the plurality of processing entities and their inter-entity relationships of the Aubineau and Leeds system using the hierarchical tree data structure of Vibbert, in which a root node (start processing entity) and children nodes (further processing entities) correspond to dialog agents and dialog agencies. One of ordinary skill would have been motivated to do so because Vibbert teaches that organizing the plurality of dialog agents and agencies as nodes of a task tree provides an orderly, navigable representation of the permissible transitions among the entities and allows the dialog engine to traverse the tree and shift the focus of the conversation from one entity to another part of the tree (Vibbert, Para. [0080]-[0081]), thereby yielding the predictable benefit of structured and controllable routing of a request from a starting entity to further entities in a multi-entity conversational system.
Regarding Claim 4, Aubineau, Leeds, and Vibbert teach
The method of Claim 1.
Leeds further teaches
wherein the action indicates jumping to a third processing entity of the plurality of processing entities to process the request, the third processing entity being determined from the at least one further processing entity of the plurality of processing entities based on the request (Col. 2, lines 54-60, “A submind has basic chat capabilities, a communication conduit to a forum, and a means for processing natural language that normally includes the ability to assess proposed responses in the context of the conversation to enable collaboration.” Fig. 25 (Cycle Detail, Phase 1 Propose)) Examiner Comments: Leeds teach that each submind proposes a response and is evaluated based on its natural language understanding of the current request, and the forum selects the submind whose proposal best fits the request, that is, determines a further entity from among the plurality based on the request, reading on the claimed action of jumping to a third processing entity determined from the at least one further processing entity based on the request.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Leeds's multi-submind architecture into Aubineau's configuration-based chatbot creation system in order to improve coverage and reliability by distributing processing across multiple independently configured agents. Both references are in the same field of conversational AI system design, and the advantages of using a plurality of specialized processing entities to handle a broader range of user requests.
Regarding Claim 5, Aubineau, Leeds, and Vibbert teach
The method of Claim 1.
Aubineau further teaches
wherein the action indicates processing the request by the first processing entity (Para. [0046], “if the action is not identified, the chatbot displays a message indicating that no action can be found that can do the user request. For example: ‘Sorry, but the request cannot be performed.’ Alternatively, the chatbot may request that the user rephrase the question or to provide more information which might help the system to find an appropriate action in the action catalog.”) Examiner Comments: Aubineau teach that when the first processing entity (chatbot) cannot identify a suitable action for the request, it still responds to the user directly, either with a fallback message or a clarification request, meaning the first processing entity itself processes the request, reading on the claimed action of the first processing entity processing the request.
Regarding Claim 6, Aubineau, Leeds, and Vibbert teach
The method of Claim 1.
Leeds further teach
wherein the first processing entity comprises an agent or a referenced application (Col. 2, lines 49-65, “A Submind is a forum ‘lead’ role, able to contribute to both the content and the decision-making of the conversation…A submind may be human, artificial, or a combination, including a CCAI entity based on the present invention. A Collaboratized Chatbot may serve as a submind.”) Examiner Comments: Leeds teach that a processing entity (submind) can be any collaboratized AI agent or another CCAI instance (a referenced application), which under the broadest reasonable interpretation reads on the claimed first processing entity comprising an agent or a referenced application.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Leeds's multi-submind architecture into Aubineau's configuration-based chatbot creation system in order to improve coverage and reliability by distributing processing across multiple independently configured agents. Both references are in the same field of conversational AI system design, and the advantages of using a plurality of specialized processing entities to handle a broader range of user requests.
Regarding Claim 7, Aubineau, Leeds, and Vibbert teach
The method of Claim 1.
Aubineau further teach
wherein the first processing entity is determined as not suitable to process the request if at least one of the following conditions is met: a level of matching between prompt information of the first processing entity and the request is lower than a predetermined level; prompt information of the first processing entity indicates that the first processing entity is incapable of processing the request; or the first processing entity fails to be configured with a capability to process the request, the capability comprising a knowledge capability or a plug-in capability (Para. [0046], “In step 403, the response of the question answering system is analyzed to identify the action and whether the action is identified. Confidence scores can be computed at this point to decide if the candidate response(s) is good enough to be considered for further action, i.e., executing a Web API call…In step 404, if the action is not identified, the chatbot displays a message indicating that no action can be found that can do the user request.”) Examiner Comments: Aubineau teach that unsuitability is determined when the confidence score (level of matching between the entity’s action catalog, its prompt information, and the request) falls below a threshold, corresponding to the first claimed condition, and Leeds further teaches that subminds assess whether they have the natural language means (knowledge or capability) to handle a request (Col. 2, lines 49-67), corresponding to the second and third claimed conditions, such that the combination reads on all three claimed conditions for determining that the first processing entity is not suitable.
Regarding Claim 8, Aubineau, Leeds, and Vibbert teach
The method of Claim 1.
Aubineau further teaches
wherein the application is configured to: create a plug-in corresponding to the action for the first processing entity based on the configuration information, such that the first processing entity is capable of performing the action by invoking the plug-in (Para. [0040]-[0041], “In step 303, the system generates an action catalog from the annotated Web API specification…the action catalog is a set of sentences where each sentence describes a Web API based on its annotated description. In step 304, the system receives the annotated Web API specification and automatically generates a rule set. In embodiments of the invention, each rule in the rule set corresponds to a Web API or a portion of the Web API and each rule condition corresponds to an input to the Web API. These rules can be processed at runtime by a goal-based reasoning engine.”) Examiner Comments: Aubineau teach that the application automatically generates actionable computational components (action catalog entries and executable rules) from the configuration information, wherein each component corresponds to a specific action and is invoked by the processing entity at runtime to execute that action, reading under the broadest reasonable interpretation on the claimed plug-in corresponding to the action that is created based on configuration information and invoked by the first processing entity to perform the action.
Regarding Claim 9, Aubineau, Leeds, and Vibbert teach
The method of Claim 8.
Leeds further teaches
wherein the application is further configured to: update plug-in description information of the plug-in based on entity description information of at least one processing entity corresponding to the action (Col. 4, lines 1-10, “Bots may be evolutionary capable, cyclical or stable, wherein a bot may enable incremental changes to its decision mechanism (i.e., exhibiting persuadability) via incremental or extensible neural networks, extensible grammars or parsers, recording of experience and decision data, decision tree creation and extension, or adjustment of response mechanism sliders”; Col. 1, lines 34-53, “CCAIs are modular, distributed, scalable, and capable of incremental improvement and evolutionary extension, via growth, updates, and replacement.”) Examiner Comments: Leeds teach that the description and decision mechanisms of processing entities (subminds) are dynamically updated based on recorded experience and collaboration data from other participating entities, reading on the claimed updating of plug-in description information of the plug-in based on entity description information of at least one processing entity corresponding to the action.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Leeds's multi-submind architecture into Aubineau's configuration-based chatbot creation system in order to improve coverage and reliability by distributing processing across multiple independently configured agents. Both references are in the same field of conversational AI system design, and the advantages of using a plurality of specialized processing entities to handle a broader range of user requests.
Regarding Claim 18, is a device claim corresponding to the method claim above (Claim 1) and, therefore, is rejected for the same reasons set forth in the rejection of claim 1.
Claim(s) 10-17 is/are rejected under 35 U.S.C. 103 as being unpatentable over Leeds (US 11,431,660 B1) in view of Aubineau (US 2022/0253607 A1), and further in view of Vibbert (US 2016/0042735 A1).
Regarding Claim 10, Leeds (US 11,431,660 B1) teach
A computer-implemented method for processing a request, comprising:
receiving, by at least one processor, request for an application; (Col. 1, lines 34-53, “Further, CCAIs are modular, distributed, scalable, and capable of incremental improvement and evolutionary extension, via growth, updates, and replacement, and through utilizing self-play for honing, including, through the present invention, selecting, aggregating, and pruning sets of asymmetrical chatbots, and optimizing configurations for achieving collaborative goals.” Fig. 35 (Proctor Flowchart: “Acquire & announce prompt to bots”)) Examiner Comments: Leeds teach that the CCAI application receives a user conversational prompt (request) via the Proctor and forum mechanism, as illustrated in the Proctor Flowchart (Fig. 35) where the Proctor acquires and announces the prompt to the subminds, reading on the claimed step of receiving a request for an application.
in response to a determination that a first processing entity associated with the application is not suitable for processing the request, determining a second processing entity from a plurality of processing entities associated with the application comprising the first processing entity based on an entity jump history of the application, the entity jump history indicating a historical jump process from the starting processing entity of the plurality of processing entities to the first processing entity; (Col. 2, lines 54-68, “A submind has basic chat capabilities, a communication conduit to a forum, and a means for processing natural language that normally includes the ability to assess proposed responses in the context of the conversation to enable collaboration”; Col. 4, lines 1-15, “Bots may be evolutionary capable, cyclical or stable, wherein a bot may enable incremental changes to its decision mechanism (i.e., exhibiting persuadability) via incremental or extensible neural networks, extensible grammars or parsers, recording of experience and decision data, decision tree creation and extension, or adjustment of response mechanism sliders. The disclosed invention also provides for a non-reductive representation of decision making and experience, and reconveyance of responses based on context.”) Examiner Comments: Leeds teach that when a current submind (first processing entity) fails to provide an adequate response (is not suitable), the Proctor and forum determine the next submind (second processing entity) from the plurality of subminds based on a recorded history of decisions and experience data that tracks which subminds have been active from the starting submind to the current one, reading on the claimed determining of a second processing entity from a plurality based on an entity jump history indicating the historical jump process from the starting processing entity to the first processing entity.
switching, by at least one processor, to the second processing entity to process the request. (Col. 1, lines 54-65, “A submind is one of a set of independent, collaborating, intelligent entities that, functioning together on a forum, present themselves as a single AI. For example, a submind can be viewed as a forum of collaborators itself controlling a single participant in a higher forum.” Fig. 35 (Proctor Flowchart: “Announce & convey winning response”)) Examiner Comments: Leeds teach that the forum, via the Proctor, switches processing to the selected next submind (second processing entity) to handle the request, reading on the claimed step of switching to the second processing entity to process the request.
Leeds did not specifically teach.
Leeds did not specifically teach
the entity jump history as an explicitly maintained sequential record of the historical jump process from the starting processing entity to the first processing entity.
However, Aubineau teach
the entity jump history indicating a historical jump process from the starting processing entity of the plurality of processing entities to the first processing entity (explicit sequential tracking of the processing path) (Para. [0047], “In preferred embodiments, the reasoning engine then starts a backward chaining reasoning to retrieve answers to all the questions associated with goal to get the input values needed for execution. Backward chaining (or backward reasoning) is an inference method which ‘works backward’ from a defined goal.”) Examiner Comments: Aubineau teach an inference engine that explicitly maintains an ordered chain of sequential processing steps (the backward-chaining chain from goal to preceding conditions), corresponding to the claimed entity jump history that records the historical jump process from a starting processing entity to the first processing entity, providing the technical mechanism for identifying the prior entity in a routing sequence.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Aubineau’s explicit sequential-tracking mechanism into Leeds’s multi-submind CCAI system in order to maintain an entity jump history that records the ordered chain of subminds from the starting submind to the current one, because tracking the full routing chain enables the system to avoid circular routing, implement intelligent backtracking, and make context-aware routing decisions based on prior entity transitions, which are recognized goals in multi-agent conversational system design, and both references are in the same field with a reasonable expectation of success.
Leeds and Aubineau did not specifically teach
the plurality of processing entities having jump relationships between respective processing entities represented by a data structure having a start processing entity and at least one further processing entity.
However, Vibbert (US 2016/0042735 A1) teach
the plurality of processing entities having jump relationships between respective processing entities represented by a data structure having a start processing entity and at least one further processing entity; (Para. [0057], “Task specifications 312, 314, and 316 describing task specific behavior may be modeled with tree diagrams of task subcomponents…Dialog task specifications 314, 324, and 334 may comprise dialog agents and dialog agencies, which may each be independent program subroutines”; Para. [0061], “Root node 402 may be the topmost node in the task specification layer…the root node may control the execution of children nodes 410, 412, and 414…Each node of the task tree 400 may be a dialog agency or a dialog agent”; Para. [0081], “the user can also take initiative and shift the focus of the conversation to another part of the dialog task tree 500.”) Examiner Comments: Vibbert teach that a plurality of processing entities (dialog agents and dialog agencies) are represented as the nodes of a tree diagram data structure having a root node (the claimed start processing entity) and children nodes (the claimed at least one further processing entity), and that the conversation shifts focus, that is, jumps, from one node to another part of the tree, reading on the claimed plurality of processing entities having jump relationships between respective processing entities represented by a data structure having a start processing entity and at least one further processing entity.
It would further have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to represent the plurality of processing entities and their inter-entity relationships of the Leeds and Aubineau system using the hierarchical tree data structure of Vibbert, in which a root node (start processing entity) and children nodes (further processing entities) correspond to dialog agents and dialog agencies, because Vibbert teaches that organizing the plurality of dialog entities as nodes of a task tree provides an orderly, navigable representation of the permissible transitions among the entities and allows the dialog engine to traverse the tree and shift the focus of the conversation from one entity to another part of the tree (Vibbert, Para. [0080]-[0081]), yielding the predictable benefit of structured and controllable routing of a request from a starting entity to further entities.
Regarding Claim 11, Leeds, Aubineau, and Vibbert teach
The method of Claim 10.
Leeds further teach
based on the application switching to the second processing entity, updating the entity jump history (Col. 4, lines 1-10, “Bots may be evolutionary capable, cyclical or stable, wherein a bot may enable incremental changes to its decision mechanism (i.e., exhibiting persuadability) via incremental or extensible neural networks, extensible grammars or parsers, recording of experience and decision data, decision tree creation and extension, or adjustment of response mechanism sliders.”) Examiner Comments: Leeds teach that when the forum switches processing to a different submind (second processing entity), the experience and decision data, which constitutes the entity jump history, are recorded and updated in the system, reading on the claimed step of updating the entity jump history based on the application switching to the second processing entity.
Regarding Claim 12, Leeds, Aubineau, and Vibbert teach
The method of Claim 11.
Leeds further teach
in response to the application responding to the request using the second processing entity, determining whether the historical jump process from the starting processing entity to the first processing entity comprises the second processing entity; and in response to determining that the historical jump process fails to comprise the second processing entity, adding the second processing entity to the historical jump process to update the entity jump history (Col. 4, lines 1-10, “Bots may be evolutionary capable, cyclical or stable, wherein a bot may enable incremental changes to its decision mechanism (i.e., exhibiting persuadability) via incremental or extensible neural networks, extensible grammars or parsers, recording of experience and decision data, decision tree creation and extension, or adjustment of response mechanism sliders.” Fig. 35 (Proctor Flowchart)) Examiner Comments: Leeds teach that the system checks its recorded experience and decision data each conversation cycle to avoid duplication in the routing records, and when a new submind responds to a request it is added to the routing history, which combined with Aubineau’s sequential-tracking mechanism teaches checking whether the second processing entity is already in the jump history and adding it only if absent, reading on the claimed limitation of updating the entity jump history.
Regarding Claim 13, Leeds, Aubineau, and Vibbert teach
The method of Claim 11.
Leeds and Aubineau further teach
in response to determining that the historical jump process comprises the second processing entity, determining a part from the historical jump process, the part corresponding to a jump process from the starting processing entity to the second processing entity; and updating the entity jump history based on the part (Leeds, Col. 1, lines 53-60, “The disclosed invention also provides for a non-reductive representation of decision making and experience, and reconveyance of responses based on context”; Aubineau, Para. [0047], “In preferred embodiments, the reasoning engine then starts a backward chaining reasoning to retrieve answers to all the questions associated with goal to get the input values needed for execution.”) Examiner Comments: Leeds’s non-reductive experience tracking combined with Aubineau’s backward-chaining reasoning teach identifying the relevant segment of the routing history from the starting entity to the identified second entity (the part), and updating the entity jump history based on that extracted part in order to eliminate a circular routing segment, reading on the claimed determining of a part from the historical jump process and updating the entity jump history based on that part.
Regarding Claim 14, Leeds, Aubineau, and Vibbert teach
The method of Claim 11.
Leeds further teach
in response to determining that the application responds to the request using a third processing entity, updating the entity jump history to indicate jumping from the first processing entity to the third processing entity to respond to the request, the third processing entity being different from the second processing entity (Col. 1, lines 34-60, “CCAIs are modular, distributed, scalable, and capable of incremental improvement and evolutionary extension…The disclosed invention also provides for a non-reductive representation of decision making and experience”; Col. 4, lines 1-14, “Bots may be evolutionary capable, cyclical or stable, wherein a bot may enable incremental changes to its decision mechanism (i.e., exhibiting persuadability) via incremental or extensible neural networks, extensible grammars or parsers, recording of experience and decision data.”) Examiner Comments: Leeds teach that when a third submind ultimately responds to a request (different from the previously determined second submind), the system records this outcome as new experience and decision data in the routing history, reading on the claimed updating of the entity jump history to indicate jumping from the first processing entity to the third processing entity.
Regarding Claim 15, Leeds, Aubineau, and Vibbert teach
The method of Claim 10.
Leeds further teach
wherein the historical jump process indicates that the second processing entity jumped to the first processing entity to respond to a particular request during operation (Col. 4, lines 1-14, “Bots may be evolutionary capable, cyclical or stable, wherein a bot may enable incremental changes to its decision mechanism (i.e., exhibiting persuadability) via incremental or extensible neural networks, extensible grammars or parsers, recording of experience and decision data, decision tree creation and extension.” Fig. 25 (Cycle Detail, Phase 3 Decide)) Examiner Comments: Leeds teach that the recorded experience and decision data captures which subminds previously handed off (jumped) to other subminds for particular requests during operation, reading on the claimed historical jump process indicating that the second processing entity previously jumped to the first processing entity to respond to a particular request.
Regarding Claim 16, Leeds, Aubineau, and Vibbert teach
The method of Claim 10.
Aubineau further teaches
wherein determining a second processing entity from a plurality of processing entities associated with the application based on an entity jump history of the application comprises: determining a number of processing entities jumped during a processing of the request; and in response to the number of processing entities being less than a threshold number, determining a second processing entity from a plurality of processing entities associated with the application based on an entity jump history of the application (Para. [0046], “Confidence scores can be computed at this point to decide if the candidate response(s) is good enough to be considered for further action, i.e., executing a Web API call.”) Examiner Comments: Aubineau teach evaluating a threshold-based criterion (confidence score) at each processing step to decide whether to proceed with routing to further processing actions or to stop, which under the broadest reasonable interpretation reads on the claimed step of determining whether the number of processing entities jumped is below a threshold number before continuing with entity jump history-based routing to a second processing entity.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Aubineau’s threshold-based stopping criterion into Leeds’s multi-submind routing system to cap the maximum number of entity transitions permitted during a single request, preventing routing loops, excessive latency, and resource exhaustion, all well-recognized concerns in multi-agent conversational system design, with a reasonable expectation of success.
Regarding Claim 17, Leeds, Aubineau, and Vibbert teach
The method of Claim 16.
Aubineau further teach
in response to the number of processing entities reaching the threshold number, responding to the request using the first processing entity (Para. [0046], “if the action is not identified, the chatbot displays a message indicating that no action can be found that can do the user request. For example: ‘Sorry, but the request cannot be performed.’ … Alternatively, the chatbot may request that the user rephrase the question or to provide more information which might help the system to find an appropriate action in the action catalog.”) Examiner Comments: Aubineau teach that when the routing or reasoning process reaches a terminal condition (exceeding processing steps or no action found meeting the threshold), the first processing entity (chatbot) itself responds to the user with a fallback message, reading on the claimed step of responding to the request using the first processing entity when the threshold number of entity jumps is reached.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention to incorporate Aubineau’s threshold-based stopping criterion into Leeds’s multi-submind routing system to cap the maximum number of entity transitions permitted during a single request, preventing routing loops, excessive latency, and resource exhaustion, all well-recognized concerns in multi-agent conversational system design, with a reasonable expectation of success.
Response to Arguments
Applicant’s arguments with respect to claims 1-18 have been considered but are moot because the arguments do not apply to the previous cited sections of the references used in the previous office action. The current office action is now citing additional references to address the newly added claimed limitations.
Conclusion
THIS ACTION IS MADE FINAL. 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 AMIR SOLTANZADEH whose telephone number is (571)272-3451. The examiner can normally be reached M-F, 9am - 5pm ET.
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, Wei Mui can be reached at (571) 272-3708. 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.
/AMIR SOLTANZADEH/Examiner, Art Unit 2191 /Ted T. Vo/Primary Examiner, Art Unit 2191
Prosecution Insights