CN112836098A - Multi-role-based conversation assistance method and device - Google Patents
Multi-role-based conversation assistance method and device Download PDFInfo
- Publication number
- CN112836098A CN112836098A CN202110133476.2A CN202110133476A CN112836098A CN 112836098 A CN112836098 A CN 112836098A CN 202110133476 A CN202110133476 A CN 202110133476A CN 112836098 A CN112836098 A CN 112836098A
- Authority
- CN
- China
- Prior art keywords
- user
- role
- conversation
- assistant
- information
- 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.)
- Granted
Links
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
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/9032—Query formulation
- G06F16/90332—Natural language query formulation or dialogue systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/903—Querying
- G06F16/9035—Filtering based on additional data, e.g. user or group profiles
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Abstract
The application discloses a multi-role-based conversation assistance method and device, wherein the method comprises the following steps: under a normal interactive working mode, acquiring conversation content, user emotion information and environment information based on real-time conversation voice of a user and an assistant role; judging whether the conversation is in a state of impasse of conversation or a preset emergency environment event occurs at present according to the conversation content, the emotion information of the user and the environment information; if so, selecting the switched target assistant role according to the conversation content, and switching the assistant role of the user to the target assistant role. By the method and the device, the intellectuality of role switching can be improved.
Description
Technical Field
The invention relates to a computer application technology, in particular to a multi-role-based conversation assistance method and device.
Background
With the development and popularization of intelligent devices, people have more and more demands on intelligent voice interaction, and an intelligent voice interaction system gradually becomes a research hotspot.
A scheme for training and configuring an intelligent assistant team and switching the intelligent assistant roles is proposed at present. In the scheme, a plurality of assistant roles are arranged in an intelligent assistant team, wherein each assistant has different characteristics (such as different functions, basic language models, training degrees, avatars, personalities and the like). The features of each assistant may be configured and trained by the trainer, end user, or assistant server. The assistant may perform different functions in a dialog with the end user on the user interface. Different intelligent assistants may be appropriate for different conversation contexts. Meanwhile, different assistants can interact with each other to jointly complete tasks related to the user, and the interactions are displayed on the user interaction interface.
The inventor discovers that in the process of implementing the invention: the above scheme has the problem of poor intelligence of the switching mechanism of the assistant role, and the specific analysis is as follows:
the helper role switching mechanism adopted in the scheme is as follows: when the assistant role interacts with the user, whether the assistant role needs to be switched is judged according to the keywords input by the current user and the change of the objective background. In practical applications, when a user converses with a real person, there is a need to switch helper roles under the influence of emotion and conversation atmosphere. For example, when a conversation falls into an impasse, the user may not want to continue the conversation and may wish to change to a helper role. At this time, since the switching mechanism does not distinguish the emotion and the conversation state of the user in real time, the helper role is not automatically switched for the user, so that the user can obtain a better conversation experience.
Disclosure of Invention
In view of the above, the present invention is directed to a multi-role based dialog assistance method and apparatus, which can improve the intelligence of role switching.
In order to achieve the purpose, the technical scheme provided by the invention is as follows:
a multi-role based dialog assistance method, comprising:
under a normal interactive working mode, acquiring conversation content, user emotion information and environment information based on real-time conversation voice of a user and an assistant role;
judging whether the conversation is in a state of impasse of conversation or a preset emergency environment event occurs at present according to the conversation content, the emotion information of the user and the environment information; if so, selecting the switched target assistant role according to the conversation content, and switching the assistant role of the user to the target assistant role.
In one embodiment, the determining whether the current session impasse state is met comprises:
judging whether the conversation content exists: the times of the same keywords appearing in the dialogue information of the user reach a preset repetition time threshold value, and if so, the current state of the dialogue impasse is judged;
judging whether the conversation content exists: the conversation information of the assistant role comprises all preset candidate feedback information of the user consultation problem, and if yes, the current conversation impasse state is judged;
and judging whether the user has negative emotion currently according to the emotion information of the user, wherein the corresponding emotion fluctuation degree reaches a preset emotion threshold value, and if so, judging that the user is in a state of session impasse currently.
In one embodiment, the selecting the target helper role for handover comprises:
determining the relevance of the background story of the assistant role to the conversation content or the emergency event based on the profile of the assistant role in the assistant role database; taking the assistant role to which the background story with the relevance degree larger than a preset relevance threshold belongs as a candidate assistant role;
determining affinity of the candidate helper role to the user based on social relationship configuration information in a profile of the candidate helper role;
determining a degree of matching of the personality traits of the candidate helper character with the dialog content or the emergency event based on the personality traits in the profile of the candidate helper character;
calculating a comprehensive matching score of the candidate helper role according to a weight calculation method based on the correlation degree, the intimacy degree, the matching degree and corresponding weight parameters;
and selecting the candidate helper role corresponding to the maximum comprehensive matching score as the target helper role.
In one embodiment, the method further comprises:
when the real-time dialogue voice of the user and the assistant role is detected for the first time, whether the user is a new user is judged by using a voiceprint identity recognition method, and if so, a file is created for the user.
In one embodiment, the method further comprises:
and after the assistant role is switched, the conversation content triggering the switching is used as a background story and is stored in a user file of the user, and the background story is stored in a file of the corresponding assistant role having a social relationship with the user according to a preset event sharing rule.
In one embodiment, the method further comprises:
when judging that the user is not in a conversation impasse state and the emergency environment event does not occur currently, or judging that the user is satisfied with the target assistant role after the switching is completed, searching feedback information related to the current real-time conversation voice of the user in the internet and background stories in respective files of the user and the current assistant role according to the current conversation content of the user; generating a feedback information candidate set according to the search result;
for each piece of feedback information in the feedback information candidate set, calculating the character feature matching degree and the conversation logic feature matching degree of the piece of feedback information and the current helper role, and calculating the information comprehensive matching degree of the piece of feedback information based on the character feature matching degree and the conversation logic feature matching degree;
based on the conversation logic characteristics of the current assistant role, performing wording modification on the feedback information corresponding to the maximum value of the information comprehensive matching degree to generate text feedback information;
and processing the text feedback information according to the language characteristics in the file of the current assistant role, generating corresponding voice feedback information and playing the voice feedback information to the user.
In one embodiment, the method further comprises:
under a role generation working mode, when non-real-time voice is detected, voiceprint identity recognition is carried out according to the non-real-time voice; if the identification fails, a file is created for the corresponding user or the assistant role;
analyzing the external voice to obtain the language characteristics of the corresponding user or the assistant role and adding the language characteristics into the corresponding file;
extracting conversation content and emotion information from the external voice; analyzing the extracted conversation content and emotion information to obtain conversation logic characteristics and character characteristics of corresponding user or assistant role and adding the conversation logic characteristics and character characteristics to corresponding files;
judging whether the emotion fluctuation degree of the corresponding user or the assistant role reaches a preset emotion threshold value or not according to the emotion information, and if so, adding the corresponding conversation content serving as the background story into the file of the corresponding user or the assistant role;
and triggering the user to configure a social relationship for the user or the assistant role corresponding to the external voice, and adding the social relationship into the file of the corresponding user or the assistant role.
In one embodiment, the method further comprises:
when an interactive scene mode setting instruction of a user is received, setting a current interactive scene mode as a target interactive scene mode indicated by the instruction according to the interactive scene mode setting instruction; and switching the assistant role to the assistant role matched with the target interaction scene mode.
In one embodiment, the method further comprises:
and judging whether an interactive scene mode needs to be switched or not based on the environment information, and if so, switching the current interactive scene mode to an interactive scene mode matched with the environment information.
The embodiment of the present application further discloses a multi-role based dialog assistance device, which includes:
the information extraction unit is used for acquiring conversation content, user emotion information and environment information based on real-time conversation voice of the user and the assistant role in a normal interactive working mode;
the role control unit is used for judging whether the mobile terminal is in a conversation impasse state or a preset emergency environment event occurs at present according to the conversation content, the user emotion information and the environment information; if so, selecting the switched target assistant role according to the conversation content, and switching the assistant role of the user to the target assistant role.
Also disclosed in embodiments herein is a non-transitory computer readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the steps of the multi-role based dialog assistance method as described above.
An electronic device is also disclosed in an embodiment of the present application, and includes the aforementioned nonvolatile computer-readable storage medium, and the processor having access to the nonvolatile computer-readable storage medium.
According to the technical scheme, under a normal interaction working mode, conversation content, user emotion information and environment information are obtained from conversation voice of a user and an assistant role in real time, whether the user is in a conversation impasse state or a preset emergency environment event occurs is monitored based on the obtained content and information, when the user is monitored to be in the conversation impasse state or the preset emergency environment event occurs, a switched target assistant role is selected according to the obtained conversation content, and the assistant role of the user is switched to the selected target assistant role. Therefore, when the user interacts with the assistant role, the emotion and the conversation state of the user are distinguished in real time, and the assistant role is automatically triggered to be switched for the user when the user is in an impasse state or an emergency environment event occurs, so that the assistant role matched with the current conversation situation is switched for the user, the intelligence of role switching can be improved, and the user can obtain better conversation experience.
Drawings
FIG. 1 is a schematic flow chart of a method according to an embodiment of the present invention;
fig. 2 is a schematic structural diagram of an apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail with reference to the accompanying drawings and specific embodiments.
Fig. 1 is a schematic method flow diagram of an embodiment of the present invention, and as shown in fig. 1, the conversation assistance method based on multiple roles implemented in the embodiment mainly includes:
The step is used for acquiring conversation content, user emotion information and environment information from the conversation between the user and the assistant role in real time when the user and the assistant role normally interact, so that whether the user is in a conversation impasse state or a preset emergency environment event occurs is monitored in step 102 based on the content and the information, and the assistant role is automatically triggered to be switched for the user when the conversation impasse state or the preset emergency environment event occurs, so that the intelligence of role switching is improved.
Specifically, in this step, the current emotional information and the ambient information of the user can be obtained by monitoring the language change of the user in the conversation (including but not limited to wording change, tone change, accent change, use of specific vocabularies and the like).
In practical applications, the specific method for acquiring the conversation content, the emotion information of the user and the environmental information from the real-time conversation voice is known by those skilled in the art, and will not be described herein again.
In this step, it is determined whether the current state is a session impasse state or a preset emergency environment event occurs according to the session content, the user emotion information, and the environment information acquired from the real-time session content in step 101, so as to trigger the switching of the helper role in time when the session impasse or the emergency environment event occurs.
When the helper roles need to be switched, the helper roles are switched according to the current conversation content, so that the switched helper roles can be matched with the current conversation scene, the helper roles capable of better providing interaction service for the user can be used as new helper roles, the interaction experience of the user and the helper roles can be improved, and the intelligence of helper role switching is enhanced.
In one embodiment, in order to accurately identify the session impasse state, the following method may be specifically adopted to determine whether the session impasse state is currently in the session impasse state:
judging whether the conversation content exists: the times of the same keywords appearing in the dialogue information of the user reach a preset repetition time threshold value, and if so, the current state of the dialogue impasse is judged;
judging whether the conversation content exists: the conversation information of the assistant role comprises all preset candidate feedback information of the user consultation problem, and if yes, the current conversation impasse state is judged;
and judging whether the user has negative emotion currently according to the emotion information of the user, wherein the corresponding emotion fluctuation degree reaches a preset emotion threshold value, and if so, judging that the user is in a state of session impasse currently.
In the method, when the times of the same keyword appearing in the dialogue information of the user reach the preset repetition time threshold, the user may ask the same question for multiple times in the dialogue, and effective information is not provided through multiple rounds of interactive helper roles at the moment with high probability, so that the dialogue enters the impasse.
When the conversation information of the assistant role contains all the preset candidate feedback information of the user consultation problem, the assistant role means that all the preset candidate feedback information of the user consultation problem is provided for the user, so that effective feedback information can not be provided for the user in terms of the current conversation any more, and the conversation enters a impasse.
When the current negative emotions (such as negative emotions of anger, fear, sadness and the like) of the user are judged, and the corresponding emotion fluctuation degree reaches a preset emotion threshold value, the current interaction with the assistant role is unsatisfied by the user, the emotion fluctuation is severe, the conversation enters an impasse, and the switching of the assistant role needs to be triggered, so that the assistant role capable of providing service is switched for the user.
In the above method, the keywords may be set in advance by those skilled in the art according to actual service needs. The repetition time threshold and the emotion threshold can be set by those skilled in the art according to the needs of the actual application scenario.
In practical applications, the emergency event may be preset according to practical application scenarios, for example, the emergency event may be an event that a user is injured or the surrounding environment is changed greatly, but is not limited thereto.
In one embodiment, in order to select a helper role that better matches the current dialog scenario, the following method may be specifically used to select a target helper role to be switched:
step x1, determining the relevance of the background story of the helper character to the dialog content or the emergency event based on the profile of the helper character in the helper character database; and taking the assistant role to which the background story with the relevance degree larger than a preset relevance threshold belongs as a candidate assistant role.
Step x2, determining the affinity of the candidate helper role to the user based on social relationship configuration information in the profile of the candidate helper role.
Step x3, determining the degree of matching of the personality traits of the candidate helper character with the dialog content or the emergency event based on the personality traits in the profile of the candidate helper character.
The personality characteristics may specifically include: impatience, inward, gentle, calm, etc., but not limited thereto.
Step x4, calculating the comprehensive matching score of the candidate helper role according to a weight calculation method based on the correlation degree, the intimacy degree, the matching degree and the corresponding weight parameters.
Step x5, selecting the candidate helper role corresponding to the maximum comprehensive matching score as the target helper role.
By adopting the method, the background stories closely related to the current conversation content or the occurred emergency event, namely the background stories with the correlation degree larger than the preset threshold value, are selected from the assistant character database, the assistant characters to which the background stories belong are taken as candidate assistant characters, the comprehensive matching scores of the candidate assistant characters are calculated by combining the intimacy degree of the assistant characters and the user and the matching degree of the character characteristics of the assistant characters and the user, and the candidate assistant character with the maximum comprehensive matching score is selected as the switched target assistant character, so that the assistant character more matched with the current conversation content or the occurred emergency event can be selected, and the user conversation experience after the character switching can be improved.
In one embodiment, when the real-time dialogue voice of the user and the assistant role is detected for the first time, whether the user is a new user or not can be judged based on the voiceprint recognition method, and if the user is the new user, a file is created for the new user so as to perform user management.
In one embodiment, after the helper role switch is completed, the user profile may be further refined to improve the accuracy of the subsequent helper role switch, and specifically, the method may further include the following steps:
and after the assistant role is switched, the conversation content triggering the switching is used as a background story and is stored in a user file of the user, and the background story is stored in a file of the corresponding assistant role having a social relationship with the user according to a preset event sharing rule.
Specifically, the event sharing rules may be set according to the actual application scenario needs, so as to define which helper roles of the user may (or may not) share the background story, for example, events generated between the user and certain specific roles (such as a psychological consultant, a lawyer, etc.) will not be shared as appropriate; for experiences where the user explicitly indicates a reluctance to share, the sharing will not be performed.
In one embodiment, in order to improve the sense of reality of the user interacting with the assistant character, feedback information given to the user in the interaction process can be generated according to the conversation logic characteristics and the character characteristics of the current assistant character in real time, and specifically, the method further comprises the following steps:
step y1, when it is judged that the user is not in the state of session impasse and the emergency environment event does not occur, or when it is judged that the user is satisfied with the target assistant role after the switching is completed, according to the current session content of the user, searching feedback information related to the current real-time session voice of the user in the internet and background stories in respective files of the user and the current assistant role; and generating a feedback information candidate set according to the search result.
This step is used to search for feedback information related to the current real-time conversational speech of the user to construct a candidate set of feedback information.
And step y2, for each piece of feedback information in the feedback information candidate set, calculating the character feature matching degree and the dialogue logic feature matching degree of the piece of feedback information and the current helper role, and calculating the information comprehensive matching degree of the piece of feedback information based on the character feature matching degree and the dialogue logic feature matching degree.
The step is used for calculating the information comprehensive matching degree of each piece of feedback information in the feedback information candidate set so as to select the information most suitable for the feedback of the assistant role in the subsequent steps.
The dialog logic features may specifically include: whether to adhere to certain qualifications or to be more user friendly, whether to say a good lie on a particular occasion, whether to change the direction of conversation due to the user's mood, etc., but is not limited thereto.
In this step, a weight calculation mode may be specifically adopted to obtain the information comprehensive matching degree of each piece of feedback information.
And step y3, based on the conversation logic characteristics of the current assistant role, performing wording modification on the feedback information corresponding to the maximum value of the information comprehensive matching degree to generate text feedback information.
In this step, the feedback information with the maximum information comprehensive matching degree is selected, and the wording of the feedback information is modified by combining the conversation logic characteristics of the assistant role, so as to obtain the corresponding text feedback information.
The specific method for generating the text feedback information according to the dialog logic features is known to those skilled in the art and will not be described herein.
And step y4, processing the text feedback information according to the language characteristics in the file of the current assistant role, generating corresponding voice feedback information and playing the voice feedback information to the user.
The language features may specifically include: sentence making style by words, vocal buddhist, speed of speech and accent.
In this step, the specific method for processing the text feedback information according to the language features to obtain the voice feedback information is known by those skilled in the art and will not be described herein again.
In one embodiment, the following method may be further employed to create a profile for a user or a helper role based on non-real-time speech in the role generation mode of operation:
step z1, under the role generation working mode, when non-real-time voice is detected, voiceprint identity recognition is carried out according to the non-real-time voice; if the identification fails, a profile is created for the corresponding user or helper role.
Here, the non-real-time voice may be a voice call, a recorded broadcast, or the like provided by the user.
In the step, when the voiceprint identity is identified, voiceprint features are obtained from played voice, if no user matched with the voiceprint features exists, the identification fails, and the fact that a file is not established for a corresponding voice owner is indicated, so that the file is required to be established for the voice owner through triggering, and language features, conversation logic features, character features and social relations are obtained from the voice through subsequent steps and added to the corresponding file.
And step z2, analyzing the external voice to obtain the language characteristics of the corresponding user or assistant role and adding the language characteristics to the corresponding file.
Step z3, extracting dialogue content and emotion information from the external voice; and analyzing the extracted conversation content and emotion information to obtain conversation logic characteristics and character characteristics of the corresponding user or assistant role, and adding the conversation logic characteristics and character characteristics to the corresponding file.
Step z4, according to the emotion information, judging whether the emotion fluctuation degree of the corresponding user or assistant role reaches a preset emotion threshold value, if so, adding the corresponding conversation content as the background story to the file of the corresponding user or assistant role.
In the step, the conversation content when the emotion fluctuation exists in the non-real-time voice is added into the file as a background story, so that the file can contain the background story reflecting the character characteristics and emotion information of the user or the character, and the screening of assistant characters is facilitated.
Step z5, triggering the user to configure social relations for the user or the assistant role corresponding to the external voice, and adding the social relations into the profile of the corresponding user or the assistant role.
Here, the social relationship may specifically include: the relationship such as the relationship of the relationship, friends, teachers and students, colleagues and the like, the degree of relationship affinity and sparseness, whether the experience story can be shared, and the like, but not limited thereto.
In an embodiment, an interaction scene mode may be further set according to a user instruction, and switching of an assistant role may be triggered according to a current interaction scene mode, which may specifically be implemented by the following method:
when an interactive scene mode setting instruction of a user is received, setting a current interactive scene mode as a target interactive scene mode indicated by the instruction according to the interactive scene mode setting instruction; and switching the assistant role to the assistant role matched with the target interaction scene mode.
In an embodiment, the switching of the interactive scene mode may be automatically triggered when the current dialog background changes (for example, a transition from an entertainment scene to a meeting or a work scene) further according to the environmental information extracted from the current real-time dialog content, and may be specifically implemented by the following method:
and judging whether an interactive scene mode needs to be switched or not based on the environment information, and if so, switching the current interactive scene mode to an interactive scene mode matched with the environment information.
Specifically, if the current dialog background is known to change based on the environment information, it is determined that the interactive scene mode needs to be switched currently, otherwise, the interactive scene mode does not need to be switched.
Based on the above method embodiment, it can be seen that the intelligence of role switching can be effectively improved by using the above method embodiment, and specific implementation of the above embodiment is further described below with reference to several specific application scenarios.
Scene one:
when a user is interacting with a voice assistant, different voice assistant characters may have different voice feedback in different interaction modes for the user's input. For example, for a user to ask "what to do on weekends" in an entertainment mode, an assistant character with an outward style character characteristic, liking outdoor sports, may reply to "go to kickbar"; an assistant character with an in-type personality trait that likes silence might revert to "go to a library bar"; family member characters such as mom character may reply to "go to park bar"; while a female friend character might reply to a female user "shopping a street bar". If the user says "play music" in the learning mode, the helper character with the outward style personality may play a relatively cheerful music, the helper character with the inward personality may play a relatively soft pure music, the character helper with the severe personality may reply "concentrate on learning and do not listen to music", the friend character with the common music preference may play music consistent with the user preference, if the helper character currently serving has a background story about the user learning, it may reply "listen to music again when there is a need to rest, otherwise it may not complete a lesson just listening to music as when xx".
When the user is in the entertainment mode and is communicating with the voice assistant, if the current user suddenly enters the teleconference, the system is switched to the conference-opening mode through background analysis and is switched to the voice assistant corresponding to the conference-opening mode.
And in a second scenario, when the user interacts with the voice system in the entertainment mode and intends to send a birthday for a friend, the user can switch the current voice assistant role to the friend role. The relevant preparation (such as location, arrangement, food, background music, game segments, etc.) is sent to the voice assistant, which replies according to the personality trait setting for the character in the system. For example, when the user asks "what background music should be selected", the voice assistant replies to the friend with the type of music (rock music) that the friend may prefer according to the personality setting (extroversion, impatience). As another example, when the user asks about a party's game circle, the voice assistant may set a reply to the friend's likely favorite game type based on the previous background story (e.g., a more happy game that was experienced together).
And a third scenario, when the user interacts with the voice system in the learning mode and prepares for a piece of paper, the user can communicate with a default voice assistant (learning assistant) in the learning mode preset in the system. When the user feels vexed to the thinking of the current learning-class voice assistant, repeatedly asks the same question and shows unsatisfied answer, the user can automatically switch to a certain friend role. The friend character can play the background story (such as scenic spots once gone together, performances once seen together, and the like) related to the topic of the paper according to the background story setting.
And a fourth scenario, when the child user interacts with the voice system, if accidental injury occurs, the child user can be switched to a certain character assistant (such as a family member mother) with placating property through strategy control, instructive voice is provided according to the prior setting (such as the position of an emergency kit in a home is provided according to the background story of the mother character), and the background story playing of the character is triggered when the child user has violent emotion and needs placating (such as the story about courage and placating in the assistant mother character). The user experience of injury will be supplemented by the personality traits and language characteristics of the child user (response to accidental injury, verbal feature expression when excited, etc.) and the children user's background stories (time and place of occurrence of the experience and subsequent processing). At the same time, the mom role will update the background story (the unexpected experience and subsequent processing of the child user at the time and place, the reaction to the soothing story, etc.), and at the same time, decide whether to share the experience among the roles (whether to share the story experience among family members, whether to inform the child user's friend roles, etc.) according to the relationship between the service roles corresponding to the child user. In addition, whether to trigger the switching of the helper role can be decided according to the reaction of the child user to the mom role (whether the child user is satisfied with the voice interaction with the mom role, whether to switch to dad or other friend roles, and the like).
And in a fifth scenario, when the old user interacts with the voice system in the monitoring mode, if the user feels annoyance to the voice assistant feedback of the current worker role (such as unwilling to do regular exercise or take medicines according to the suggestions of the worker role), when the counterfeeling emotion of the user reaches a certain threshold value and the counterwillingness causes unsmooth or stagnation of communication, the assistant role can be switched to other more convincing roles (a certain doctor role) or more appeasing roles (such as a certain family member) so as to achieve the purpose of continuing the communication. At the same time, the context causing the switching is used as a background story and is supplemented to the file of the user and the assistant role which can be shared.
Corresponding to the above method embodiment, an embodiment of the present application further discloses a dialog assisting apparatus based on multiple roles, as shown in fig. 2, including:
an information extraction unit 201, configured to, in a normal interactive working mode, obtain conversation content, user emotion information, and environment information based on real-time conversation voice of a user and an assistant character;
the role control unit 202 is used for judging whether the current state is in a session impasse state or a preset emergency environment event occurs according to the session content, the user emotion information and the environment information; if so, selecting the switched target assistant role according to the conversation content, and switching the assistant role of the user to the target assistant role.
Embodiments of the present application also provide a non-transitory computer-readable storage medium storing instructions that, when executed by a processor, cause the processor to perform the steps of the multi-role based dialog assistance method as described above.
Embodiments of the present application further provide an electronic device, including the aforementioned non-volatile computer-readable storage medium, and the processor having access to the non-volatile computer-readable storage medium.
In summary, the above description is only a preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (12)
1. A multi-role based dialog assistance method, comprising:
under a normal interactive working mode, acquiring conversation content, user emotion information and environment information based on real-time conversation voice of a user and an assistant role;
judging whether the conversation is in a state of impasse of conversation or a preset emergency environment event occurs at present according to the conversation content, the emotion information of the user and the environment information; if so, selecting the switched target assistant role according to the conversation content, and switching the assistant role of the user to the target assistant role.
2. The method of claim 1, wherein determining whether a session impasse state is currently present comprises:
judging whether the conversation content exists: the times of the same keywords appearing in the dialogue information of the user reach a preset repetition time threshold value, and if so, the current state of the dialogue impasse is judged;
judging whether the conversation content exists: the conversation information of the assistant role comprises all preset candidate feedback information of the user consultation problem, and if yes, the current conversation impasse state is judged;
and judging whether the user has negative emotion currently according to the emotion information of the user, wherein the corresponding emotion fluctuation degree reaches a preset emotion threshold value, and if so, judging that the user is in a state of session impasse currently.
3. The method of claim 1, wherein selecting the target helper role for handover comprises:
determining the relevance of the background story of the assistant role to the conversation content or the emergency event based on the profile of the assistant role in the assistant role database; taking the assistant role to which the background story with the relevance degree larger than a preset relevance threshold belongs as a candidate assistant role;
determining affinity of the candidate helper role to the user based on social relationship configuration information in a profile of the candidate helper role;
determining a degree of matching of the personality traits of the candidate helper character with the dialog content or the emergency event based on the personality traits in the profile of the candidate helper character;
calculating a comprehensive matching score of the candidate helper role according to a weight calculation method based on the correlation degree, the intimacy degree, the matching degree and corresponding weight parameters;
and selecting the candidate helper role corresponding to the maximum comprehensive matching score as the target helper role.
4. The method of claim 1, further comprising:
when the real-time dialogue voice of the user and the assistant role is detected for the first time, whether the user is a new user is judged by using a voiceprint identity recognition method, and if so, a file is created for the user.
5. The method of claim 1, further comprising:
and after the assistant role is switched, the conversation content triggering the switching is used as a background story and is stored in a user file of the user, and the background story is stored in a file of the corresponding assistant role having a social relationship with the user according to a preset event sharing rule.
6. The method of claim 1, further comprising:
when judging that the user is not in a conversation impasse state and the emergency environment event does not occur currently, or judging that the user is satisfied with the target assistant role after the switching is completed, searching feedback information related to the current real-time conversation voice of the user in the internet and background stories in respective files of the user and the current assistant role according to the current conversation content of the user; generating a feedback information candidate set according to the search result;
for each piece of feedback information in the feedback information candidate set, calculating the character feature matching degree and the conversation logic feature matching degree of the piece of feedback information and the current helper role, and calculating the information comprehensive matching degree of the piece of feedback information based on the character feature matching degree and the conversation logic feature matching degree;
based on the conversation logic characteristics of the current assistant role, performing wording modification on the feedback information corresponding to the maximum value of the information comprehensive matching degree to generate text feedback information;
and processing the text feedback information according to the language characteristics in the file of the current assistant role, generating corresponding voice feedback information and playing the voice feedback information to the user.
7. The method of claim 1, further comprising:
under a role generation working mode, when non-real-time voice is detected, voiceprint identity recognition is carried out according to the non-real-time voice; if the identification fails, a file is created for the corresponding user or the assistant role;
analyzing the external voice to obtain the language characteristics of the corresponding user or the assistant role and adding the language characteristics into the corresponding file;
extracting conversation content and emotion information from the external voice; analyzing the extracted conversation content and emotion information to obtain conversation logic characteristics and character characteristics of corresponding user or assistant role and adding the conversation logic characteristics and character characteristics to corresponding files;
judging whether the emotion fluctuation degree of the corresponding user or the assistant role reaches a preset emotion threshold value or not according to the emotion information, and if so, adding the corresponding conversation content serving as the background story into the file of the corresponding user or the assistant role;
and triggering the user to configure a social relationship for the user or the assistant role corresponding to the external voice, and adding the social relationship into the file of the corresponding user or the assistant role.
8. The method of claim 1, further comprising:
when an interactive scene mode setting instruction of a user is received, setting a current interactive scene mode as a target interactive scene mode indicated by the instruction according to the interactive scene mode setting instruction; and switching the assistant role to the assistant role matched with the target interaction scene mode.
9. The method of claim 1, further comprising:
and judging whether an interactive scene mode needs to be switched or not based on the environment information, and if so, switching the current interactive scene mode to an interactive scene mode matched with the environment information.
10. A multi-character based conversation assistance apparatus comprising:
the information extraction unit is used for acquiring conversation content, user emotion information and environment information based on real-time conversation voice of the user and the assistant role in a normal interactive working mode;
the role control unit is used for judging whether the mobile terminal is in a conversation impasse state or a preset emergency environment event occurs at present according to the conversation content, the user emotion information and the environment information; if so, selecting the switched target assistant role according to the conversation content, and switching the assistant role of the user to the target assistant role.
11. A non-transitory computer readable storage medium storing instructions which, when executed by a processor, cause the processor to perform the steps of the multi-role based dialog assistance method of any of claims 1 to 9.
12. An electronic device comprising the non-volatile computer-readable storage medium of claim 11, and the processor having access to the non-volatile computer-readable storage medium.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110133476.2A CN112836098B (en) | 2021-02-01 | 2021-02-01 | Multi-role-based conversation assistance method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110133476.2A CN112836098B (en) | 2021-02-01 | 2021-02-01 | Multi-role-based conversation assistance method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112836098A true CN112836098A (en) | 2021-05-25 |
| CN112836098B CN112836098B (en) | 2023-07-07 |
Family
ID=75932715
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110133476.2A Active CN112836098B (en) | 2021-02-01 | 2021-02-01 | Multi-role-based conversation assistance method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112836098B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023159536A1 (en) * | 2022-02-28 | 2023-08-31 | 华为技术有限公司 | Human-computer interaction method and apparatus, and terminal device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106776926A (en) * | 2016-12-01 | 2017-05-31 | 竹间智能科技（上海）有限公司 | Improve the method and system of responsibility when robot talks with |
| CN107340991A (en) * | 2017-07-18 | 2017-11-10 | 百度在线网络技术（北京）有限公司 | Switching method, device, equipment and the storage medium of speech roles |
| CN110265021A (en) * | 2019-07-22 | 2019-09-20 | 深圳前海微众银行股份有限公司 | Personalized speech exchange method, robot terminal, device and readable storage medium storing program for executing |
| CN111026932A (en) * | 2019-12-20 | 2020-04-17 | 北京百度网讯科技有限公司 | Man-machine conversation interaction method and device, electronic equipment and storage medium |
-
2021
- 2021-02-01 CN CN202110133476.2A patent/CN112836098B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106776926A (en) * | 2016-12-01 | 2017-05-31 | 竹间智能科技（上海）有限公司 | Improve the method and system of responsibility when robot talks with |
| CN107340991A (en) * | 2017-07-18 | 2017-11-10 | 百度在线网络技术（北京）有限公司 | Switching method, device, equipment and the storage medium of speech roles |
| CN110265021A (en) * | 2019-07-22 | 2019-09-20 | 深圳前海微众银行股份有限公司 | Personalized speech exchange method, robot terminal, device and readable storage medium storing program for executing |
| CN111026932A (en) * | 2019-12-20 | 2020-04-17 | 北京百度网讯科技有限公司 | Man-machine conversation interaction method and device, electronic equipment and storage medium |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2023159536A1 (en) * | 2022-02-28 | 2023-08-31 | 华为技术有限公司 | Human-computer interaction method and apparatus, and terminal device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112836098B (en) | 2023-07-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP6601069B2 (en) | Dialog control apparatus, dialog control method, and program | |
| CN111564202B (en) | Psychological counseling method based on man-machine conversation, psychological counseling terminal and storage medium | |
| US10388281B2 (en) | Dialogue control apparatus, dialogue control method, and non-transitory recording medium | |
| CN105991847B (en) | Call method and electronic equipment | |
| CN111696559B (en) | Providing emotion management assistance | |
| US11222633B2 (en) | Dialogue method, dialogue system, dialogue apparatus and program | |
| Traum et al. | Incremental dialogue understanding and feedback for multiparty, multimodal conversation | |
| Fang et al. | Sounding board–university of washington’s alexa prize submission | |
| JP6699010B2 (en) | Dialogue method, dialogue system, dialogue device, and program | |
| CN111542814A (en) | Method, computer device and computer readable storage medium for changing responses to provide rich-representation natural language dialog | |
| US20190304451A1 (en) | Dialogue method, dialogue system, dialogue apparatus and program | |
| CN112632242A (en) | Intelligent conversation method and device and electronic equipment | |
| JPWO2018230345A1 (en) | Dialogue robot, dialogue system, and dialogue program | |
| US20190103106A1 (en) | Command processing program, image command processing apparatus, and image command processing method | |
| JP3676981B2 (en) | KANSEI GENERATION METHOD, KANSEI GENERATION DEVICE, AND SOFTWARE | |
| CN111556999B (en) | Method, computer device and computer readable storage medium for providing natural language dialogue by providing substantive answer in real time | |
| CN112836098A (en) | Multi-role-based conversation assistance method and device | |
| CN111767386A (en) | Conversation processing method and device, electronic equipment and computer readable storage medium | |
| US20230351661A1 (en) | Artificial intelligence character models with goal-oriented behavior | |
| JP2022531994A (en) | Generation and operation of artificial intelligence-based conversation systems | |
| Kopp et al. | Incremental multimodal feedback for conversational agents | |
| KR101891495B1 (en) | Method and computer device for controlling a display to display conversational response candidates to a user utterance input, and computer readable recording medium | |
| CN109359177B (en) | Multi-mode interaction method and system for story telling robot | |
| US11960983B1 (en) | Pre-fetching results from large language models | |
| US11954570B2 (en) | User interface for construction of artificial intelligence based characters |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |