CN113901192A - Conversation method, device, equipment and medium for conversation node parameter pre-filling - Google Patents
Conversation method, device, equipment and medium for conversation node parameter pre-filling Download PDFInfo
- Publication number
- CN113901192A CN113901192A CN202111269845.7A CN202111269845A CN113901192A CN 113901192 A CN113901192 A CN 113901192A CN 202111269845 A CN202111269845 A CN 202111269845A CN 113901192 A CN113901192 A CN 113901192A
- Authority
- CN
- China
- Prior art keywords
- conversation
- node
- intention
- nodes
- session
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 74
- 230000002452 interceptive effect Effects 0.000 claims abstract description 40
- 230000008569 process Effects 0.000 claims abstract description 37
- 230000003993 interaction Effects 0.000 claims abstract description 16
- 238000013528 artificial neural network Methods 0.000 claims description 21
- 238000004590 computer program Methods 0.000 claims description 20
- 238000012549 training Methods 0.000 claims description 13
- 230000004044 response Effects 0.000 claims description 12
- 238000013473 artificial intelligence Methods 0.000 abstract description 2
- 230000004048 modification Effects 0.000 description 5
- 238000012986 modification Methods 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 238000012790 confirmation Methods 0.000 description 3
- 230000009191 jumping Effects 0.000 description 2
- 238000007726 management method Methods 0.000 description 2
- 238000012545 processing Methods 0.000 description 2
- 238000012550 audit Methods 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 230000001960 triggered effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
- 230000001755 vocal effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3329—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
- G06F40/35—Discourse or dialogue representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Human Computer Interaction (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a conversation method, a device, equipment and a medium for conversation node parameter pre-filling, which relate to the field of artificial intelligence, and the method comprises the following steps: the interactive equipment starts a conversation, outputs an inquiry statement corresponding to a current conversation node and acquires user input information corresponding to the inquiry statement; comparing the user input information with intention parameters of all conversation nodes prestored in the interaction equipment, judging whether conversation nodes to be filled except the current conversation node matched with the user input information exist according to a comparison result, if so, loading the intention parameters of the corresponding conversation nodes to be filled into the current conversation node, and associating the intention parameters with corresponding sentences in the input information; and after the question and answer of the current conversation node are finished, conversation node skipping is carried out, whether the skipped conversation node is the conversation node to be filled is judged, if yes, the inquiry of the conversation node to be filled is skipped, and the next conversation node is entered until the conversation process is finished.
Description
Technical Field
The present invention relates to the field of artificial intelligence, and in particular, to a conversation method, apparatus, device, and medium for conversation node parameter pre-population.
Background
The existing conversation robots are basically configured through a flow template, and during the conversation process, the robots carry out conversation according to the dialogs configured by the template and recognize the intention of a client, so that the next what to say is found on the template. Therefore, the whole dialog process is carried out on the template in sequence.
However, the existing template dialog robot has a drawback. If the client says some useful information in advance, the information can be identified not on the current template node, the robot does not record the useful information, and when the robot runs to the lower node, the robot still asks the client again to obtain the necessary information. The robot lacks the memory to user's input information, and the conversation flow is too solid, lacks the flexibility, leads to user experience poor.
Disclosure of Invention
In view of the problems in the prior art, the present invention provides a conversation method, device, apparatus and medium for conversation node parameter pre-population, which mainly solve the problem that the existing solution has poor user experience due to lack of flexibility in the process of intent conversation through a conversation template.
In order to achieve the above and other objects, the present invention adopts the following technical solutions.
A conversation method of conversation node parameter pre-population, comprising:
the interactive equipment starts a conversation, outputs an inquiry statement corresponding to a current conversation node and acquires user input information corresponding to the inquiry statement;
comparing the user input information with intention parameters of all conversation nodes prestored in the interaction equipment, judging whether conversation nodes to be filled except the current conversation node matched with the user input information exist according to a comparison result, if so, loading the intention parameters of the corresponding conversation nodes to be filled into the current conversation node, and associating the intention parameters with corresponding sentences in the input information;
and after the question and answer of the current conversation node are finished, conversation node skipping is carried out, whether the skipped conversation node is the conversation node to be filled is judged, if yes, the inquiry of the conversation node to be filled is skipped, and the next conversation node is entered until the conversation process is finished.
Optionally, the interactive device initiates a dialog, including:
acquiring a biological identification feature of a user, acquiring identity information of the user according to the biological identification feature, and starting a conversation according to the identity information; and/or the presence of a gas in the gas,
acquiring service requirement information input by a user, matching a corresponding service session flow according to the service requirement information, and starting a session according to the service session flow.
Optionally, before comparing the user input information with the intention parameters of all the session nodes pre-stored in the interaction device, the method includes:
acquiring intention parameters contained in all conversation nodes of the interactive equipment, and constructing an intention parameter table of corresponding services according to the intention parameters, wherein the intention parameters comprise intentions of the conversation nodes and identification information of the conversation nodes;
inputting the user input information acquired by the current conversation node into a preset intention recognition model to obtain the user intention;
and acquiring matched intention parameters from the intention parameter table according to the user intention.
Optionally, the manner of obtaining the intention recognition model includes:
obtaining synonyms or synonyms of each dialogue intention to construct a training sample set;
and constructing a neural network framework, inputting the training sample set into the neural network framework, adjusting the neural network parameters, and obtaining the intention recognition model.
Optionally, judging whether a to-be-filled session node other than the current session node matched with the user input information exists according to the comparison result, if so, loading the intention parameter of the corresponding to-be-filled session node to the current session node, and associating the intention parameter with the corresponding statement in the input information, including:
judging whether a conversation node to be filled is matched with user input information in the current conversation node or not, if so, skipping to the previous conversation node, and acquiring an intention parameter loaded by the previous conversation node;
judging whether the intention parameter loaded by the previous dialogue node corresponds to the dialogue node between the previous dialogue node and the current dialogue node, recording the intention parameter as an intermediate node, skipping the inquiry statement of the intermediate node, and executing the node skipping from the previous dialogue node to the current dialogue node.
Optionally, after obtaining the user input information corresponding to the query statement, the method further includes:
acquiring and storing user input sentences associated with the intention parameters, acquiring corresponding inquiry sentences of the dialogue nodes to be filled according to the intention parameters of the dialogue nodes to be filled, and outputting the inquiry sentences and the associated corresponding user input sentences to a preset interactive interface for information display;
and after the conversation process of the corresponding service is completed, acquiring the service requirement of the user according to the user input information contained in each jumped conversation node and the user input statement related to the intention parameter loaded by each conversation node.
Optionally, obtaining a service requirement output of the user includes:
matching a preset business rule according to the conversation node and the jump logic of the loaded intention parameter;
and acquiring corresponding service response data from a service database according to the service rule.
A conversation device with conversation node parameter pre-population, comprising:
the system comprises a wake-up module, a session starting module and a session starting module, wherein the wake-up module is used for starting a session by interactive equipment, outputting an inquiry statement corresponding to a current session node and acquiring user input information corresponding to the inquiry statement;
the parameter pre-filling module is used for comparing the user input information with intention parameters of all conversation nodes pre-stored by the interactive device, judging whether conversation nodes to be filled except the current conversation node matched with the user input information exist according to a comparison result, and if so, loading the intention parameters of the corresponding conversation nodes to be filled into the current conversation node and associating the intention parameters with corresponding sentences in the input information;
and the node skipping module is used for skipping the conversation nodes after the question and answer of the current conversation node are finished, judging whether the skipped conversation nodes are the conversation nodes to be filled, if so, skipping the inquiry of the conversation nodes to be filled, and entering the next conversation node until the conversation process is finished.
A computer device, comprising: a memory, a processor and a computer program stored on the memory and executable on the processor, the processor implementing the steps of the session method with pre-populated session node parameters when executing the computer program.
A computer-readable storage medium, on which a computer program is stored which, when being executed by a processor, carries out the steps of the session method of session node parameter pre-population.
As described above, the session method, apparatus, device and medium for session node parameter pre-population according to the present invention have the following advantages.
The method comprises the steps of comparing inquiry sentences of a user with intention parameters of all conversation nodes, preloading the intention parameters of other conversation nodes corresponding to the inquiry sentences to the current conversation node, associating the preloaded intention parameters with corresponding sentences in the inquiry sentences, recording user input information, and skipping the inquiry process of the preloaded conversation nodes to be filled when the conversation nodes skip, so that a conversation process is simplified, and user experience is enhanced.
Drawings
Fig. 1 is a flow chart illustrating a session method for pre-filling session node parameters according to an embodiment of the present invention.
Fig. 2 is a schematic flow chart of a session initiated by an interactive device according to an embodiment of the present invention.
Fig. 3 is a schematic flow chart illustrating the process of matching the intention parameters by the session node according to an embodiment of the present invention.
FIG. 4 is a block diagram of a session apparatus for session node parameter pre-population in accordance with one embodiment of the present invention.
Fig. 5 is a schematic structural diagram of a computer device according to an embodiment of the present invention.
Detailed Description
The embodiments of the present invention are described below with reference to specific embodiments, and other advantages and effects of the present invention will be easily understood by those skilled in the art from the disclosure of the present specification. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
It should be noted that the drawings provided in the following embodiments are only for illustrating the basic idea of the present invention, and the components related to the present invention are only shown in the drawings rather than drawn according to the number, shape and size of the components in actual implementation, and the type, quantity and proportion of the components in actual implementation may be changed freely, and the layout of the components may be more complicated.
The inventor has found that when a dialog is run to the second node for a business robot, such as an audit lending robot, the robot asks the customer: "ask you for a room under your name? "customer answer: "I do not have a room, but I have a car. The phrase "the customer actually expresses two intentions," i do not have a room, "and" i have a car. However, although the robot recognizes these 2 intentions, for this node, the robot will only pick "i do not have room" for processing and will ignore the intention "i have car", and then when the dialogue goes to the third node, according to the template configuration, the robot will ask the client again: "do you ask you for a car under name? This can lead to confusion during the client session due to repeated queries. It is thus concluded that not the bank staff, but the robot is speaking to him. This lack of memory interaction is not humanoid and not very friendly.
Referring to fig. 1, to solve the problem in the robot session process, the present invention provides a session method, apparatus, device and medium for pre-filling session node parameters, including the following steps:
s1: the interactive equipment starts a conversation, outputs an inquiry statement corresponding to a current conversation node and acquires user input information corresponding to the inquiry statement;
s2: comparing the user input information with intention parameters of all conversation nodes prestored in the interaction equipment, judging whether conversation nodes to be filled except the current conversation node matched with the user input information exist or not, if so, loading the corresponding intention parameters of the conversation nodes to be filled into the current conversation node, and associating the intention parameters with corresponding sentences in the input information;
s3: and after the question and answer of the current conversation node are finished, conversation node skipping is carried out, whether the skipped conversation node is the conversation node to be filled is judged, if yes, the inquiry of the corresponding conversation node is skipped, and the next conversation node is entered until the conversation process is finished.
The session method for pre-filling session node parameters provided by the present invention is described below with reference to specific embodiments.
In step S1, the interactive device starts a dialog, outputs an inquiry statement corresponding to the current dialog node, and acquires user input information corresponding to the inquiry statement.
Referring to fig. 2, in an embodiment, the interactive device initiates a session, including the following steps:
step S101, acquiring a biological identification feature of a user, acquiring identity information of the user according to the biological identification feature, and starting a conversation according to the identity information; and/or the presence of a gas in the gas,
step S102, acquiring service requirement information input by a user, matching a corresponding service session flow according to the service requirement information, and starting a session according to the service session flow.
Specifically, the interactive device may include a question and answer robot, a portable interactive terminal, or a terminal device with a display interface, and the like. Taking the question-answering robot as an example, the question-answering robot can be provided with a collecting device for collecting the biological identification characteristics of the user. Wherein the biometric features include: fingerprint characteristics, human face characteristics, iris characteristics, voice characteristics and the like, wherein the question answering robot can select one or more biological identification characteristics to identify the identity of the user. Illustratively, the acquisition device can adopt a face recognition module, the question answering robot acquires a face image through a camera, inputs the face image into the face recognition module, acquires user identity information corresponding to the face by recognizing the face image, and then starts a conversation according to the identity information. If the face recognition user is "Wang XX", the question and answer robot may output "Mr. Wang, you good! Ask what can help you? ", the output of the query statement for the specific object is made based on the identity information.
And the user identity is not distinguished, and the conversation of the interactive equipment is directly awakened according to the service requirement information input by the user. Specifically, the service session flows corresponding to different services may be preset for different services, the session node of a specific service may be determined, and the corresponding service session flow may be represented based on the service category. The user can input the service requirement information through voice or text, after the interactive equipment identifies the service requirement information in the voice or text, the interactive equipment matches the corresponding service type according to the service requirement information, and obtains the service conversation process corresponding to the service type. The similarity between the characteristics corresponding to the service requirement information and the service category can be calculated by adopting the normal form distance or the Euclidean distance, and whether the characteristics are matched with the service category or not is judged according to the similarity threshold. In another embodiment, the basic session of the interactive device may also be woken up by biometric features, such as asking the user for the need, "Mr. xx asking what can help you? ". And acquiring user service requirement information based on the basic session, and then giving the service requirement information to match with a corresponding service session process. The specific applications and interaction means are not limited herein.
And whether the user has the inquiry authority for requesting the service can be judged by collecting the biological identification characteristics of the user, and then the user authority starts the service session process of the corresponding service. The selection and setting of the specific biometric features may be adjusted according to the actual application requirements, and is not limited herein.
In step S2, the user input information is compared with the intention parameters of all the dialogue nodes pre-stored in the interaction device, and it is determined whether there is a dialogue node to be filled other than the current dialogue node that matches the user input information according to the comparison result, and if there is a dialogue node to be filled, the intention parameter of the corresponding dialogue node to be filled is loaded to the current dialogue node and associated with the corresponding sentence in the input information.
The business conversation process can comprise a plurality of conversation nodes, and connection or skip relation between the conversation nodes can be established according to conversation intention. Specifically, taking a robot conversation as an example, the conversation intents may include the following three categories:
(1) global general purpose intent: i.e. objectional intentions such as: wait for the next moment.
(2) General purpose: a common intention would correspond to only one problem, i.e. a common intention would only occur once in one robot. For example: i have a car.
(3) The general purpose of the process is as follows: the control flow direction corresponds to a plurality of problems. For example: yes, not.
Each dialog node may correspond to multiple session intents, and illustratively, a dialog node may include one general intent, multiple flow general intents, and one global general intent. In the conversation process, when a user modifies the input information of the previous conversation node, the global general intention parameter of the current conversation node can be triggered through global general intents such as 'wait once' and 'slightly wait', the modified information is input, and the conversation node which inputs the error information in the previous conversation node is jumped to. And determining which connected lower conversation node the current conversation node needs to jump to through the general purpose of the process. And the general intention corresponds to a reply sentence of the inquiry sentence corresponding to the conversation node.
Referring to fig. 3, in an embodiment, before comparing the user input information with the intention parameters of all session nodes pre-stored in the interactive device, the method includes the following steps:
step S103, obtaining intention parameters contained in all conversation nodes of the interactive equipment, and constructing an intention parameter table corresponding to the service according to the intention parameters, wherein the intention parameters comprise intentions of the conversation nodes and identification information of the conversation nodes;
step S104, inputting the user input information acquired by the current conversation node into a preset intention recognition model to obtain the user intention;
step S105, acquiring matched intention parameters from the intention parameter table according to the user intention.
Specifically, after the interactive device is awakened to determine the service session flow, the session nodes corresponding to the service session flow can be obtained, the intention parameters included in each session node are further obtained, and an intention parameter table including all the intention parameters in the service session flow is generated. Each intention parameter contains conversation intentions of the conversation nodes, such as global general intentions, process general intentions, common intentions and the like, different conversation nodes can be distinguished through unique identification codes, and the unique identification codes of the corresponding conversation nodes are stored in each intention parameter. After the intention parameter is determined, the conversation node where the intention parameter is located can be determined through the unique identification code.
In one embodiment, the hash value of each intention parameter can be calculated, and the hash value is stored in the intention parameter table, so that the intention data is not tampered, and the data security is improved. When comparing the intention parameters, whether the data are consistent or not is judged directly through hash value comparison, and data comparison efficiency can also be improved.
In one embodiment, the manner of obtaining the intent recognition model includes:
obtaining synonyms or synonyms of each dialogue intention to construct a training sample set;
and constructing a neural network framework, inputting the training sample set into the neural network framework, adjusting the neural network parameters, and obtaining the intention recognition model.
Specifically, the neural network framework can adopt a network model architecture of a deep learning neural network and a recurrent neural network to obtain an initial neural network framework. Aiming at the intention parameters contained in different business conversation processes, sample data corresponding to each intention parameter is sorted, and the sample data comprises synonyms, homonyms and the like of the conversation intention. Illustratively, the common intent is "i have a car", and the corresponding sample data is: "i has a car", "i has an audi", etc. And (3) adjusting parameters of the neural network by constructing a sample training neural network frame to obtain the corresponding relation between the sample data and the intention parameters. The user input information can be directly mapped into corresponding intention parameters through the intention recognition model obtained through training. The specific model training process is not described herein. After the intention parameters corresponding to the input information of the current session node user are obtained through the intention identification model, the hash value of the obtained intention parameters can be calculated, and the corresponding intention parameters and the unique identification codes of the corresponding session nodes are obtained from the intention parameter table according to the hash value. And acquiring the user intention of the current conversation node through the intention recognition model, matching the intention parameters in the intention parameter table according to the user intention, and determining whether the input information of the current conversation node corresponds to the intention parameters of other conversation nodes so as to load the corresponding intention parameters of other nodes into the current conversation node.
In an embodiment, determining whether there is a to-be-filled session node other than the current session node matched with the user input information according to the comparison result, if there is a to-be-filled session node, loading an intention parameter of the corresponding to-be-filled session node to the current session node, and associating the intention parameter with a corresponding statement in the input information, includes:
judging whether a conversation node to be filled is matched with user input information in the current conversation node or not, if so, skipping to the previous conversation node, and acquiring an intention parameter loaded by the previous conversation node;
judging whether the intention parameter loaded by the previous dialogue node corresponds to the dialogue node between the previous dialogue node and the current dialogue node, recording the intention parameter as an intermediate node, skipping the inquiry statement of the intermediate node, and executing the node skipping from the previous dialogue node to the current dialogue node.
Specifically, when the user input information of the current conversation node is identified to be matched with the intention parameters of other conversation nodes according to the steps, the conversation node to be filled and the corresponding intention parameters are determined according to the input information of the user at the current conversation node, and the intention parameters are pre-filled in the current conversation node. If the current dialog node interaction device outputs the query sentence "do you have a room? "i have a room and a car" and "having a car" corresponds to the intention parameter of another session node, and the unique identification code of the session node corresponding to "having a car" and the location of the session node can be determined by matching with the intention parameter table. And loading the matched intention parameters to the current conversation node, storing the 'presence of a vehicle' in the parameter values of the matched intention parameters, and associating the intention parameters. A pre-populated log table may be provided through which the intent parameters pre-populated for each conversation node are recorded. After the whole conversation process is completed, the pre-filling record table can be directly inquired to determine which conversation nodes have pre-filled intention parameters.
In step S3, after the question and answer of the current session node are completed, a session node skip is performed, and it is determined whether the skipped session node is the session node to be filled, if so, the inquiry of the session node to be filled is skipped, and the next session node is entered until the session process is completed.
Specifically, after the question and answer of the current conversation node are completed, the next conversation node can be skipped to according to the skip relation of the conversation node, whether the intention parameter of the next conversation node is pre-filled into the previous conversation node can be judged, specifically, the pre-filling record table can be inquired to determine the corresponding pre-filling intention parameter, and then whether the intention parameter in the skipped conversation node is pre-filled into the previous conversation node can be judged. If the pre-filling is carried out, the inquiry statement of the corresponding conversation node is skipped, the conversation node of the next jump is determined based on the book input statement related to the pre-filling intention parameter, the next conversation node is directly jumped to, whether the corresponding intention parameter is pre-filled is judged, if the pre-filling is not carried out, the inquiry statement of the conversation node is directly output to obtain the input information of the user, and the intention parameter is pre-filled based on the input information of the user.
When the user enters a global general intent, the intent parameters of the preceding conversation node may be pre-populated. If the input information of the user at the current conversation node is 'waiting', i have no room at local place A but have a set of rooms at place B, and the intention parameter corresponding to 'having room' corresponds to the conversation node before the current conversation node, the current conversation node can jump to the corresponding previous conversation node. The method comprises the steps of determining a jump relationship of conversation nodes according to a user input statement 'a suite is arranged at a B place' associated with pre-filled intention parameters, acquiring the intention parameters pre-filled by a previous conversation node according to previous user input information, determining the intention parameters of the conversation nodes between the previous conversation node and a current conversation node according to the pre-filled intention parameters of the previous conversation node, marking the conversation nodes pre-filled by the previous conversation node as intermediate nodes, skipping the intermediate nodes, and simplifying a jump back path of the current conversation node of the previous conversation node.
In an embodiment, after obtaining the user input information corresponding to the query statement, the method further includes:
acquiring and storing user input sentences associated with the intention parameters, acquiring corresponding inquiry sentences of the dialogue nodes to be filled according to the intention parameters of the dialogue nodes to be filled, and outputting the inquiry sentences and the associated corresponding user input sentences to a preset interactive interface for information display;
and after the conversation process of the corresponding service is completed, acquiring the service requirement of the user according to the user input information contained in each jumped conversation node and the user input statement related to the intention parameter loaded by each conversation node.
Specifically, query sentences of skipped conversation nodes and user input sentences associated with intention parameters of the skipped conversation nodes can be displayed through the interactive interface, so that a user can intuitively know information required by each conversation node and grasp requirement information of corresponding services. When the user uses the system next time, the input content of the conversation node can be adjusted through the learned service demand information, the interaction efficiency is improved, and the user experience is further enhanced.
In one embodiment, obtaining the service requirement output of the user includes:
matching a preset business rule according to the conversation node and the jump logic of the loaded intention parameter;
and acquiring corresponding service response data from a service database according to the service rule.
And the converted conversation node corresponds to the service response data. For example, the service response data a is passed through the session nodes 1, 3, 4, and the service response data B is passed through the session nodes 1, 5, 6. The specific business rules can be set according to the actual application requirements, and are not limited herein.
Taking the robot as an example, assume that the robot has five nodes:
the first node asks whether the person is the first node;
the second node asks the client whether there is a room;
the third node asks for the age of the client;
the fourth node asks the client whether the client has a car;
and the fifth node judges whether the loan condition is met or not according to the inquiry and informs the client of the loan amount which can be applied.
Configuring routing intents "i do not have room" and "i have room", jumping to the second node, and configuring a confirmation dialog.
Case flow and phone show
The robot comprises: do you good, think you confirm something about the XX loan, ask you be your own mr. do you ask?
Customer: is.
The robot comprises: do you ask you for a room?
Customer: i do not have a room, but I have a car
(NLP returns 2 intents "I do not have room" and "I have car", DM traverses PreFillElementMap for the "I have car" intent, finds that there is this intent in the intentAndParam set in the fourth node, so we do the following operations: PreFillElementMap.get ("04"). setValue ("I have car"))
The robot comprises: ask you for the years?
Customer: age 30.
(after the value of age is 30, the program finds PreFillElementMap. get ("04"). getValue () has a value, so this value "I have car" is treated as the intention of the fourth node's customer reply, and the fourth node is skipped to the verbal broadcast)
The robot comprises: you are good, and the loan amount of you is 2 ten thousand yuan according to the auditing conditions.
Customer: not good meaning, i wrongly say before, i have a room in the Shanghai.
(NLP returns the intention of 'I have room', DM triggers the route intention to jump, and broadcasts confirmation talk)
The robot comprises: good, ask you for sure there is a room?
Customer: is.
(after customer confirmation, according to the original logic, go to the third node to continue inquiring, but with the intention of pre-filling, PreFillElementMap.get ("03"). getValue () and PreFillElementMap.get ("04"). getValue () all have values, so skip in turn and go to the fifth node to broadcast)
The robot comprises: good, your amount is 5 ten thousand yuan according to the auditing conditions.
In one embodiment, as shown in fig. 4, there is provided a session device for session node parameter pre-population, the device comprising: the wake-up module 10 is configured to start a dialog by an interactive device, output an inquiry statement corresponding to a current dialog node, and acquire user input information corresponding to the inquiry statement; the parameter pre-filling module 11 is configured to compare the user input information with intention parameters of all conversation nodes pre-stored in the interaction device, determine whether a conversation node to be filled other than the current conversation node that matches the user input information exists according to a comparison result, and if the conversation node to be filled exists, load the intention parameter of the corresponding conversation node to be filled into the current conversation node and associate the intention parameter with a corresponding statement in the input information; and the node skipping module 12 is configured to skip a session node after the question and answer of the current session node are completed, determine whether the skipped session node is the session node to be filled, if so, skip the inquiry of the session node to be filled, and enter the next session node until the session process is completed.
In one embodiment, the wake-up module 10 includes: the identity authentication unit is used for acquiring the biological identification features of the user, acquiring the identity information of the user according to the biological identification features and starting a conversation according to the identity information; and the service matching unit is used for acquiring service requirement information input by a user, matching a corresponding service session flow according to the service requirement information, and starting a session according to the service session flow.
In an embodiment, the apparatus further comprises: a parameter table obtaining module, configured to obtain intention parameters included in all session nodes of the interaction device, and construct an intention parameter table corresponding to a service according to the intention parameters, where the intention parameters include intentions of the session nodes and identification information of the session nodes; the intention identification module is used for inputting the user input information acquired by the current conversation node into a preset intention identification model to obtain the user intention; and the parameter acquisition module is used for acquiring the matched intention parameters from the intention parameter table according to the intention of the user.
In one embodiment, the parameter pre-fill module 11 comprises: the parameter loading unit is used for judging whether the dialogue node to be filled is matched with the user input information in the current dialogue node or not is a previous dialogue node, if so, jumping to the previous dialogue node, and acquiring an intention parameter loaded by the previous dialogue node; and the node recording unit is used for judging whether the intention parameter loaded by the previous dialogue node corresponds to the dialogue node between the previous dialogue node and the current dialogue node, recording the intention parameter as an intermediate node, skipping the inquiry statement of the intermediate node, and executing the node jump from the previous dialogue node to the current dialogue node.
In an embodiment, the apparatus further comprises: the dialogue information display module is used for acquiring and storing the user input sentences associated with the intention parameters, acquiring corresponding inquiry sentences of the dialogue nodes to be filled according to the intention parameters of the dialogue nodes to be filled, and outputting the inquiry sentences and the associated corresponding user input sentences to a preset interactive interface for information display; and the node data response module is used for acquiring the service requirement of the user according to the user input information contained in each skipped session node and the user input statement associated with the intention parameter loaded by each session node after completing the session process of the corresponding service.
In one embodiment, the node data response module includes: the service rule matching unit is used for matching a preset service rule according to the conversation node and the jump logic of the loaded intention parameter; and the response output unit is used for acquiring corresponding service response data from the service database according to the service rule.
In one embodiment, the tool matching unit includes: the card acquisition component is used for displaying the acquired card according to the card matched with the business requirement and through the visual interactive interface, wherein the card comprises a name and description information of a corresponding business management tool; and the selection output component is used for acquiring the corresponding business management tool according to the card selected by the user.
The above-described dialog node parameter pre-populated dialog device may be implemented in the form of a computer program which may be run on a computer apparatus as shown in fig. 5. A computer device, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor.
The various modules in the dialog device that the above-described dialog node parameters are pre-populated may be implemented in whole or in part by software, hardware, and combinations thereof. The modules can be embedded in a memory of the terminal in a hardware form or independent from the memory of the terminal, and can also be stored in the memory of the terminal in a software form, so that the processor can call and execute the corresponding operations of the modules. The processor can be a Central Processing Unit (CPU), a microprocessor, a singlechip and the like.
Fig. 5 is a schematic diagram of an internal structure of the computer device in one embodiment. There is provided a computer device comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, the processor implementing the following steps when executing the computer program: the interactive equipment starts a conversation, outputs an inquiry statement corresponding to a current conversation node and acquires user input information corresponding to the inquiry statement; comparing the user input information with intention parameters of all conversation nodes prestored in the interaction equipment, judging whether conversation nodes to be filled except the current conversation node matched with the user input information exist according to a comparison result, if so, loading the intention parameters of the corresponding conversation nodes to be filled into the current conversation node, and associating the intention parameters with corresponding sentences in the input information; and after the question and answer of the current conversation node are finished, conversation node skipping is carried out, whether the skipped conversation node is the conversation node to be filled is judged, if yes, the inquiry of the conversation node to be filled is skipped, and the next conversation node is entered until the conversation process is finished.
In an embodiment, when the processor executes, the interactive device that is implemented starts a session, including: acquiring a biological identification feature of a user, acquiring identity information of the user according to the biological identification feature, and starting a conversation according to the identity information; and/or acquiring service requirement information input by a user, matching a corresponding service session flow according to the service requirement information, and starting a session according to the service session flow.
In an embodiment, before comparing the user input information with the intention parameters of all session nodes pre-stored in the interaction device, the above processor includes: acquiring intention parameters contained in all conversation nodes of the interactive equipment, and constructing an intention parameter table of corresponding services according to the intention parameters, wherein the intention parameters comprise intentions of the conversation nodes and identification information of the conversation nodes; inputting the user input information acquired by the current conversation node into a preset intention recognition model to obtain the user intention; and acquiring matched intention parameters from the intention parameter table according to the user intention.
In an embodiment, when the processor executes the above method, the manner of obtaining the intention recognition model includes: obtaining synonyms or synonyms of each dialogue intention to construct a training sample set; and constructing a neural network framework, inputting the training sample set into the neural network framework, adjusting the neural network parameters, and obtaining the intention recognition model.
In an embodiment, the determining, when the processor executes, according to a comparison result, whether a to-be-filled session node other than the current session node matching the user input information exists, and if the to-be-filled session node exists, loading an intention parameter of the corresponding to-be-filled session node to the current session node, and associating the intention parameter with a corresponding statement in the input information includes: judging whether a conversation node to be filled is matched with user input information in the current conversation node or not, if so, skipping to the previous conversation node, and acquiring an intention parameter loaded by the previous conversation node; judging whether the intention parameter loaded by the previous dialogue node corresponds to the dialogue node between the previous dialogue node and the current dialogue node, recording the intention parameter as an intermediate node, skipping the inquiry statement of the intermediate node, and executing the node skipping from the previous dialogue node to the current dialogue node.
In an embodiment, when the processor executes the above method, after obtaining the user input information corresponding to the query statement, the method further includes: acquiring and storing user input sentences associated with the intention parameters, acquiring corresponding inquiry sentences of the dialogue nodes to be filled according to the intention parameters of the dialogue nodes to be filled, and outputting the inquiry sentences and the associated corresponding user input sentences to a preset interactive interface for information display; and after the conversation process of the corresponding service is completed, acquiring the service requirement of the user according to the user input information contained in each jumped conversation node and the user input statement related to the intention parameter loaded by each conversation node.
In an embodiment, the obtaining of the service requirement output of the user, which is implemented when the processor executes, includes: matching a preset business rule according to the conversation node and the jump logic of the loaded intention parameter; and acquiring corresponding service response data from a service database according to the service rule.
In one embodiment, the computer device may be used as a server, including but not limited to a stand-alone physical server or a server cluster formed by a plurality of physical servers, and may also be used as a terminal, including but not limited to a mobile phone, a tablet computer, a personal digital assistant or a smart device. As shown in fig. 5, the computer apparatus includes a processor, a nonvolatile storage medium, an internal memory, a display screen, and a network interface, which are connected by a system bus.
Wherein, the processor of the computer device is used for providing calculation and control capability and supporting the operation of the whole computer device. A non-volatile storage medium of the computer device stores an operating system and a computer program. The computer program is executable by a processor for implementing the session method for session node parameter pre-population provided in the above embodiments. The internal memory in the computer device provides a cached execution environment for the operating system and computer programs in the non-volatile storage medium. The display interface can display data through the display screen. The display screen may be a touch screen, such as a capacitive screen or an electronic screen, and the corresponding instruction may be generated by receiving a click operation applied to a control displayed on the touch screen.
Those skilled in the art will appreciate that the configuration of the computer device shown in fig. 5 is a block diagram of only a portion of the configuration associated with the present application and does not constitute a limitation of the computer device to which the present application applies, and that a particular computer device may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer-readable storage medium is provided, having stored thereon a computer program which, when executed by a processor, performs the steps of: the interactive equipment starts a conversation, outputs an inquiry statement corresponding to a current conversation node and acquires user input information corresponding to the inquiry statement; comparing the user input information with intention parameters of all conversation nodes prestored in the interaction equipment, judging whether conversation nodes to be filled except the current conversation node matched with the user input information exist according to a comparison result, if so, loading the intention parameters of the corresponding conversation nodes to be filled into the current conversation node, and associating the intention parameters with corresponding sentences in the input information; and after the question and answer of the current conversation node are finished, conversation node skipping is carried out, whether the skipped conversation node is the conversation node to be filled is judged, if yes, the inquiry of the conversation node to be filled is skipped, and the next conversation node is entered until the conversation process is finished.
In one embodiment, the computer program, when executed by a processor, implements the interactive device to initiate a dialog, comprising: acquiring a biological identification feature of a user, acquiring identity information of the user according to the biological identification feature, and starting a conversation according to the identity information; and/or acquiring service requirement information input by a user, matching a corresponding service session flow according to the service requirement information, and starting a session according to the service session flow.
In an embodiment, before the comparing the user input information with the pre-stored intention parameters of all session nodes of the interaction device, the computer program, when executed by the processor, includes: acquiring intention parameters contained in all conversation nodes of the interactive equipment, and constructing an intention parameter table of corresponding services according to the intention parameters, wherein the intention parameters comprise intentions of the conversation nodes and identification information of the conversation nodes; inputting the user input information acquired by the current conversation node into a preset intention recognition model to obtain the user intention; and acquiring matched intention parameters from the intention parameter table according to the user intention.
In one embodiment, the computer program, when executed by a processor, implements a manner of obtaining the intent recognition model comprising: obtaining synonyms or synonyms of each dialogue intention to construct a training sample set; and constructing a neural network framework, inputting the training sample set into the neural network framework, adjusting the neural network parameters, and obtaining the intention recognition model.
In an embodiment, the determining, when the computer program is executed by the processor, whether a to-be-filled session node other than the current session node matching the user input information exists according to the comparison result, and if so, loading an intention parameter of the corresponding to-be-filled session node to the current session node and associating the intention parameter with a corresponding statement in the input information includes: judging whether a conversation node to be filled is matched with user input information in the current conversation node or not, if so, skipping to the previous conversation node, and acquiring an intention parameter loaded by the previous conversation node; judging whether the intention parameter loaded by the previous dialogue node corresponds to the dialogue node between the previous dialogue node and the current dialogue node, recording the intention parameter as an intermediate node, skipping the inquiry statement of the intermediate node, and executing the node skipping from the previous dialogue node to the current dialogue node.
In an embodiment, the computer program, when executed by the processor, after obtaining the user input information corresponding to the query statement, further includes: acquiring and storing user input sentences associated with the intention parameters, acquiring corresponding inquiry sentences of the dialogue nodes to be filled according to the intention parameters of the dialogue nodes to be filled, and outputting the inquiry sentences and the associated corresponding user input sentences to a preset interactive interface for information display; and after the conversation process of the corresponding service is completed, acquiring the service requirement of the user according to the user input information contained in each jumped conversation node and the user input statement related to the intention parameter loaded by each conversation node.
In one embodiment, the instructions, when executed by the processor, implement obtaining a business requirement output of a user, including: matching a preset business rule according to the conversation node and the jump logic of the loaded intention parameter; and acquiring corresponding service response data from a service database according to the service rule.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by a computer program, which can be stored in a non-volatile computer-readable storage medium, and can include the processes of the embodiments of the methods described above when the program is executed. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), or the like.
In summary, the present invention provides a conversation method, apparatus, device and medium for conversation node parameter pre-population, and designs an intention and parameter pre-population mechanism of a robot conversation node; by pre-filling intentions and parameters of the rear dialogue nodes of the current turn, the problem that the robot lacks memory for user input is solved, so that the robot is more intelligent in context understanding; if the method is matched with the routing intention skip and the routing parameter skip for use, the original path can be returned after upward modification, an intermediate node without modification is skipped, the whole semantic understanding closed loop of the conversation robot is formed, the semantic understanding closed loop can be understood downwards in advance, and the semantic understanding closed loop can also be directly returned to the current state after upward modification, so that the conversation process is simplified, and the user experience is enhanced. Therefore, the invention effectively overcomes various defects in the prior art and has high industrial utilization value.
The foregoing embodiments are merely illustrative of the principles and utilities of the present invention and are not intended to limit the invention. Any person skilled in the art can modify or change the above-mentioned embodiments without departing from the spirit and scope of the present invention. Accordingly, it is intended that all equivalent modifications or changes which can be made by those skilled in the art without departing from the spirit and technical spirit of the present invention be covered by the claims of the present invention.
Claims (10)
1. A conversation method of conversation node parameter pre-population, comprising:
the interactive equipment starts a conversation, outputs an inquiry statement corresponding to a current conversation node and acquires user input information corresponding to the inquiry statement;
comparing the user input information with intention parameters of all conversation nodes prestored in the interaction equipment, judging whether conversation nodes to be filled except the current conversation node matched with the user input information exist according to a comparison result, if so, loading the intention parameters of the corresponding conversation nodes to be filled into the current conversation node, and associating the intention parameters with corresponding sentences in the input information;
and after the question and answer of the current conversation node are finished, conversation node skipping is carried out, whether the skipped conversation node is the conversation node to be filled is judged, if yes, the inquiry of the conversation node to be filled is skipped, and the next conversation node is entered until the conversation process is finished.
2. The conversation method of conversation node parameter pre-population of claim 1, wherein the interactive device initiates a conversation comprising:
acquiring a biological identification feature of a user, acquiring identity information of the user according to the biological identification feature, and starting a conversation according to the identity information; and/or the presence of a gas in the gas,
acquiring service requirement information input by a user, matching a corresponding service session flow according to the service requirement information, and starting a session according to the service session flow.
3. The session node parameter pre-populated session method of claim 1, wherein prior to comparing the user input information with the pre-stored intent parameters of all session nodes of the interactive device, comprises:
acquiring intention parameters contained in all conversation nodes of the interactive equipment, and constructing an intention parameter table of corresponding services according to the intention parameters, wherein the intention parameters comprise intentions of the conversation nodes and identification information of the conversation nodes;
inputting the user input information acquired by the current conversation node into a preset intention recognition model to obtain the user intention;
and acquiring matched intention parameters from the intention parameter table according to the user intention.
4. The conversation method of conversation node parameter pre-population of claim 3, wherein the intent recognition model is obtained in a manner comprising:
obtaining synonyms or synonyms of each dialogue intention to construct a training sample set;
and constructing a neural network framework, inputting the training sample set into the neural network framework, adjusting the neural network parameters, and obtaining the intention recognition model.
5. The conversation method according to claim 1, wherein the conversation node parameter pre-filling conversation method comprises the steps of determining whether there is a conversation node to be filled other than the current conversation node matching the user input information according to the comparison result, and if there is a conversation node to be filled, loading an intention parameter of the corresponding conversation node to be filled to the current conversation node and associating the intention parameter with a corresponding statement in the input information, and comprises:
judging whether a conversation node to be filled is matched with user input information in the current conversation node or not, if so, skipping to the previous conversation node, and acquiring an intention parameter loaded by the previous conversation node;
judging whether the intention parameter loaded by the previous dialogue node corresponds to the dialogue node between the previous dialogue node and the current dialogue node, recording the intention parameter as an intermediate node, skipping the inquiry statement of the intermediate node, and executing the node skipping from the previous dialogue node to the current dialogue node.
6. The session method for session node parameter pre-population according to claim 4, after obtaining the user input information corresponding to the query statement, further comprising:
acquiring and storing user input sentences associated with the intention parameters, acquiring corresponding inquiry sentences of the dialogue nodes to be filled according to the intention parameters of the dialogue nodes to be filled, and outputting the inquiry sentences and the associated corresponding user input sentences to a preset interactive interface for information display;
and after the conversation process of the corresponding service is completed, acquiring the service requirement of the user according to the user input information contained in each jumped conversation node and the user input statement related to the intention parameter loaded by each conversation node.
7. The session method for session node parameter pre-population of claim 6, wherein obtaining a traffic demand output of a user comprises:
matching a preset business rule according to the conversation node and the jump logic of the loaded intention parameter;
and acquiring corresponding service response data from a service database according to the service rule.
8. A conversation device with conversation node parameter pre-population, comprising:
the system comprises a wake-up module, a session starting module and a session starting module, wherein the wake-up module is used for starting a session by interactive equipment, outputting an inquiry statement corresponding to a current session node and acquiring user input information corresponding to the inquiry statement;
the parameter pre-filling module is used for comparing the user input information with intention parameters of all conversation nodes pre-stored by the interactive device, judging whether conversation nodes to be filled except the current conversation node matched with the user input information exist according to a comparison result, and if so, loading the intention parameters of the corresponding conversation nodes to be filled into the current conversation node and associating the intention parameters with corresponding sentences in the input information;
and the node skipping module is used for skipping the conversation nodes after the question and answer of the current conversation node are finished, judging whether the skipped conversation nodes are the conversation nodes to be filled, if so, skipping the inquiry of the conversation nodes to be filled, and entering the next conversation node until the conversation process is finished.
9. A computer device, comprising: memory, processor and computer program stored on the memory and executable on the processor, characterized in that the processor implements the steps of the method of any of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111269845.7A CN113901192A (en) | 2021-10-29 | 2021-10-29 | Conversation method, device, equipment and medium for conversation node parameter pre-filling |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111269845.7A CN113901192A (en) | 2021-10-29 | 2021-10-29 | Conversation method, device, equipment and medium for conversation node parameter pre-filling |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113901192A true CN113901192A (en) | 2022-01-07 |
Family
ID=79027550
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111269845.7A Pending CN113901192A (en) | 2021-10-29 | 2021-10-29 | Conversation method, device, equipment and medium for conversation node parameter pre-filling |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113901192A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114662960A (en) * | 2022-04-02 | 2022-06-24 | 深圳市神州云海智能科技有限公司 | Business process generation method, terminal device and computer readable storage medium |
| CN115408510A (en) * | 2022-11-02 | 2022-11-29 | 深圳市人马互动科技有限公司 | Plot interaction node-based skipping method and assembly and dialogue development system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111488447A (en) * | 2020-04-16 | 2020-08-04 | 上海茂声智能科技有限公司 | Intention node skipping method and device and skipping equipment |
| CN112000784A (en) * | 2020-03-17 | 2020-11-27 | 北京来也网络科技有限公司 | User data processing method, device and equipment combining RPA and AI and storage medium |
| US20210216594A1 (en) * | 2020-05-22 | 2021-07-15 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for backtracking common scenario dialog in multi-round dialog |
-
2021
- 2021-10-29 CN CN202111269845.7A patent/CN113901192A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112000784A (en) * | 2020-03-17 | 2020-11-27 | 北京来也网络科技有限公司 | User data processing method, device and equipment combining RPA and AI and storage medium |
| CN111488447A (en) * | 2020-04-16 | 2020-08-04 | 上海茂声智能科技有限公司 | Intention node skipping method and device and skipping equipment |
| US20210216594A1 (en) * | 2020-05-22 | 2021-07-15 | Beijing Baidu Netcom Science And Technology Co., Ltd. | Method and apparatus for backtracking common scenario dialog in multi-round dialog |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114662960A (en) * | 2022-04-02 | 2022-06-24 | 深圳市神州云海智能科技有限公司 | Business process generation method, terminal device and computer readable storage medium |
| CN115408510A (en) * | 2022-11-02 | 2022-11-29 | 深圳市人马互动科技有限公司 | Plot interaction node-based skipping method and assembly and dialogue development system |
| CN115408510B (en) * | 2022-11-02 | 2023-01-17 | 深圳市人马互动科技有限公司 | Plot interaction node-based skipping method and assembly and dialogue development system |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| EP3891596B1 (en) | Expediting interaction with a digital assistant by predicting user responses | |
| KR102222421B1 (en) | Save metadata related to captured images | |
| EP4004719B1 (en) | Ambiguity resolution with dialogue search history | |
| US9607102B2 (en) | Task switching in dialogue processing | |
| DE112016004863T5 (en) | Parameter collection and automatic dialog generation in dialog systems | |
| CN104583927B (en) | User's interface device in user terminal and the method for supporting the user's interface device | |
| US20060155546A1 (en) | Method and system for controlling input modalities in a multimodal dialog system | |
| CN113901192A (en) | Conversation method, device, equipment and medium for conversation node parameter pre-filling | |
| CN117573834B (en) | Multi-robot dialogue method and system for software-oriented instant service platform | |
| CN110019716B (en) | Multi-turn question and answer method, terminal equipment and storage medium | |
| US11106536B2 (en) | Error recovery for conversational systems | |
| CN109271503A (en) | Intelligent answer method, apparatus, equipment and storage medium | |
| CN109902163B (en) | Intelligent response method, device, equipment and storage medium | |
| CN117520498A (en) | Virtual digital human interaction processing method, system, terminal, equipment and medium | |
| CN115658875B (en) | Data processing method based on chat service and related products | |
| US11715470B2 (en) | Method and system for tracking in extended reality | |
| US8051026B2 (en) | Rules collector system and method with user interaction | |
| CN115062629A (en) | Session information identification method and device, storage medium and computer equipment | |
| CN114970559A (en) | Intelligent response method and device | |
| CN115461709A (en) | Hierarchical context-specific actions from ambient speech | |
| KR20200058612A (en) | Artificial intelligence speaker and talk progress method using the artificial intelligence speaker | |
| KR20210052972A (en) | Apparatus and method for supporting voice agent involving multiple users | |
| CN112204656A (en) | Efficient dialog configuration | |
| CN110019718B (en) | Method for modifying multi-turn question-answering system, terminal equipment and storage medium | |
| CN118332100A (en) | Dialogue content summarizing method and device, computer equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |