CN110609893A - Question-answer interaction method, device, equipment and storage medium in multi-turn conversation scene - Google Patents
Question-answer interaction method, device, equipment and storage medium in multi-turn conversation scene Download PDFInfo
- Publication number
- CN110609893A CN110609893A CN201910906760.1A CN201910906760A CN110609893A CN 110609893 A CN110609893 A CN 110609893A CN 201910906760 A CN201910906760 A CN 201910906760A CN 110609893 A CN110609893 A CN 110609893A
- Authority
- CN
- China
- Prior art keywords
- scene
- conversation
- question
- answer interaction
- matched
- 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
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
Abstract
The embodiment of the invention discloses a question-answer interaction method, a question-answer interaction device, question-answer interaction equipment and a storage medium in a multi-turn conversation scene. The method comprises the following steps: when detecting that the current conversation belongs to a new conversation scene, establishing a scene identifier matched with the new conversation scene; in the multi-turn question-answer interaction process matched with the current conversation, obtaining conversation scene description information in real time, and correspondingly storing the conversation scene description information and the scene identification; executing an interrupt processing strategy matched with the interrupt type when the multi-round question-answer interaction is detected to be interrupted; and when detecting that the user initiates the dialogue matched with the scene identification again, continuing to perform question-answer interaction with the user according to the dialogue scene description information. By using the technical scheme of the embodiment of the invention, under the multi-scene and multi-turn conversation scene, continuous and coherent multi-turn voice conversation can be realized according to the memory of the scene, and the conversation efficiency is improved.
Description
Technical Field
The embodiment of the invention relates to an artificial intelligence technology, in particular to a question-answer interaction method, a question-answer interaction device, question-answer interaction equipment and a storage medium in a multi-turn conversation scene.
Background
With the increasing application of artificial intelligence technology in life, people's life is more and more convenient. The intelligent dialogue system can understand human language and effectively communicate with human, and further execute specific tasks or answer according to understanding of human intention.
Current intelligent dialogs can be divided into single-turn dialogs, which are usually presented in the form of a question and a answer, and a plurality of turns of dialogs, in which a user makes a question or a request, and a system recognizes the user's intention, makes an answer, or performs a specific operation. For multiple rounds of conversation, it is usually presented in the form of questions and answers. In the conversation process, the robot also initiates an inquiry, and compared with a single-turn conversation, the robot is more intelligent and has richer and more diverse application scenes.
In the process of implementing the invention, the inventor finds that the prior art has the following defects: in the current stage, multiple rounds of conversations are inquired with a user according to defined contents, if the user is silent for a period of time or scene conversations are switched, the previous voice conversation cannot be recovered, and the user needs to initiate the conversation again.
Disclosure of Invention
The embodiment of the invention provides a question-answer interaction method, a question-answer interaction device, question-answer interaction equipment and a storage medium in a multi-round conversation scene, so that sustainable and coherent multi-round voice conversations are realized under multi-round conversations of multiple scenes according to the memory of the scenes.
In a first aspect, an embodiment of the present invention provides a question-answer interaction method in a multi-turn dialog scenario, where the method includes:
when detecting that the current conversation belongs to a new conversation scene, establishing a scene identifier matched with the new conversation scene;
in the multi-turn question-answer interaction process matched with the current conversation, obtaining conversation scene description information in real time, and correspondingly storing the conversation scene description information and the scene identification;
executing an interrupt processing strategy matched with the interrupt type when the multi-round question-answer interaction is detected to be interrupted;
and when detecting that the user initiates the dialogue matched with the scene identification again, continuing to perform question-answer interaction with the user according to the dialogue scene description information.
In a second aspect, an embodiment of the present invention further provides a question-answer interaction device in a multi-turn dialog scenario, where the device includes:
the scene identification establishing module is used for establishing a scene identification matched with a new conversation scene when the current conversation is detected to belong to the new conversation scene;
the conversation scene description information storage module is used for acquiring the conversation scene description information in real time in the multi-turn question-answer interaction process matched with the current conversation and correspondingly storing the conversation scene description information and the scene identification;
the interrupt processing module is used for executing an interrupt processing strategy matched with the interrupt type when the multi-round question-answer interaction is detected to be interrupted;
and the question-answer interaction continuing module is used for continuing to carry out question-answer interaction with the user according to the dialog scene description information when detecting that the user initiates the dialog matched with the scene identifier again.
In a third aspect, an embodiment of the present invention further provides a computer device, which includes a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor, when executing the program, implements the question-answer interaction method in a multi-turn dialog scenario provided in any embodiment of the present invention.
In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, on which a computer program is stored, where the program, when executed by a computer processor, implements a question-answer interaction method in a multi-turn dialog scenario as provided in any embodiment of the present invention.
According to the embodiment of the invention, the conversation scene information of the conversation in the same multi-turn conversation scene is stored, and the new conversation is matched with the saved conversation scene after the conversation is interrupted, so that the problem that the previous conversation cannot be recovered after the conversation is interrupted is solved, the scene is memorized, and the effect of continuously and continuously carrying out multi-turn voice conversation is achieved.
Drawings
Fig. 1 is a flowchart of a question-answer interaction method in a multi-turn dialog scenario according to a first embodiment of the present invention;
FIG. 2 is a flowchart of a question-answer interaction method in a multi-turn dialog scenario according to a second embodiment of the present invention;
fig. 3 is a flowchart of a question-answer interaction method in a multi-turn dialog scenario according to a third embodiment of the present invention;
fig. 4 is a schematic structural diagram of a question-answer interaction device in a multi-turn dialog scenario according to a fourth embodiment of the present invention;
fig. 5 is a schematic structural diagram of a question-answer interaction device in a multi-turn dialog scenario in the fifth embodiment of the present invention.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings and examples. It is to be understood that the specific embodiments described herein are merely illustrative of the invention and are not limiting of the invention. It should be further noted that, for the convenience of description, only some of the structures related to the present invention are shown in the drawings, not all of the structures.
Example one
Fig. 1 is a flowchart of a question-answer interaction method in a multi-turn dialog scenario according to an embodiment of the present invention, where this embodiment is applicable to a situation where a user initiates a dialog for a multi-turn voice dialog again after the dialog is interrupted when the user performs a multi-turn voice dialog, and the method may be executed by a question-answer interaction device in the multi-turn dialog scenario, where the device may be implemented by software and/or hardware, and may be generally integrated in various terminal devices with a voice recognition function, for example, a car-mounted device or a vehicle-mounted rearview mirror, and specifically includes the following steps:
step 110, when detecting that the current conversation belongs to a new conversation scene, establishing a scene identification matched with the new conversation scene;
the dialog is a multi-turn dialog, and the multi-turn dialog is a mode of acquiring necessary information after preliminarily defining the user intention in the man-machine dialog so as to finally obtain a clear user instruction. The conversation is divided into an offline conversation and an online conversation according to whether the conversation depends on the network or not, the offline conversation does not depend on the network, and the result is directly returned according to the local database; the on-line conversation depends on the network, and if the conditions of network delay or network interruption occur in the conversation process, the conversation is continued after the network is connected. The topic of the conversation scene conversation around which the conversation is developed, e.g., a user initiates a voice conversation, asks a restaurant in the vicinity of a place, and the conversation scene is a meal. The scene identifier is essentially identification information, and corresponds to the dialog scene one to one, for example, defining a meal as scene one, a meal as the dialog scene, and a scene one as the scene identifier. The current conversation is detected to belong to a new conversation scene, for example, after the user inquires about restaurants nearby and obtains a search result, the user initiates a conversation to inquire about movie theaters nearby. When a user inquires about a restaurant nearby a place, the conversation scene is dining, and when the user inquires about a movie theater nearby the place, the conversation scene is watching, and is different from the conversation scene of the previous dining, the conversation scene is switched at the moment, namely, the current conversation belongs to a new conversation scene.
Step 120, in the multi-turn question-answer interaction process matched with the current conversation, obtaining conversation scene description information in real time, and correspondingly storing the conversation scene description information and the scene identification;
specifically, the dialog scene description information includes at least one of the following: at least one user interaction question, at least one interaction response, and at least one contextual content. The user interaction questions are questions put forward by the user to the system, the interaction response is a solution made by the system aiming at the questions put forward by the user, and the context content is other relevant information in the conversation. Since there are cases where a user's question may not get a search result, and there are cases where there is no context but a question-answer, the user interaction question, the interaction response, and the context are sometimes not in one-to-one correspondence. But the dialog scene description information corresponding to the same scene is stored corresponding to its scene identification. For example, a user searches restaurants nearby a place to obtain a search result, the restaurants nearby the place are used as user interaction questions, and the search result is saved under the scene identification of dining as an interaction response. After a period of time, the user initiates a dialogue again to search the western-style dining rooms nearby the place, a sentence of everyone is supplemented by about 100 yuan, at the moment, the system obtains the previous search result according to the saved dining scene identifier, further searches are carried out according to the search result, the search result is not obtained, at the moment, the western-style dining rooms nearby the place are used as the user interaction problem, and the everyone 100 yuan is used as context content and is saved in the dining scene identifier. Since this dialog does not result in a search result, there is no corresponding interactive response.
Step 130, when detecting that the multi-round question-answer interaction is interrupted, executing an interrupt processing strategy matched with the interrupt type;
the interruption type may be a case where the interactive response provided for the multiple rounds of question-answering interactions is not responded within a first preset time (e.g., 3 minutes, or 5 minutes, etc.), that is, the user does not respond to the system for a long time; or the interruption type may also be the detection that a new dialog does not belong to the dialog scenario, i.e. the user proposes a dialog that is different from the current dialog scenario. For different interrupt types, different processing strategies are executed, or the interrupt type may also be that the terminal device is abnormally shut down, and the like, which is not limited in this embodiment.
And 140, when the situation that the user initiates the dialogue matched with the scene identification again is detected, continuing to perform question-answer interaction with the user according to the dialogue scene description information.
After the multiple rounds of question-answer interaction are interrupted, if the user initiates a conversation again and the conversation initiated again is matched with the stored scene identification, the corresponding conversation scene description information is obtained according to the stored scene identification, and the question-answer interaction process is continued according to the conversation scene description information and the question initiated again by the user. For example, after the user inquires about restaurants nearby a certain place and obtains a search result, the scene identification of dining is saved, and if the user does not respond to the reply of the system for a long time, the question-answering interaction process is interrupted. After a period of time, the user inquires about western-style restaurants nearby somewhere, the system detects that the current conversation scene is meal and is matched with the stored scene identification, and finds that the stored conversation scene comprises meal. And then further obtaining the content to be searched by the user according to the search result returned by the system in the information corresponding to the stored scene identification. In addition, there may be a case where the reinitiated dialog matches the saved scene identifier, but the dialog scene description information corresponding to the scene identifier may not be combined with the current dialog to obtain an interactive response. For example, the user asks western-style restaurants near a place without obtaining search results, and after a while, the user asks Chinese restaurants near the place again, although the current conversation scene can be matched with the previously saved scene identifier, but no interactive response exists in the saved dining scene identifier, and at this time, the system directly searches for Chinese restaurants near the place and returns the search results.
According to the technical scheme of the embodiment, the dialog related information of different scenes is stored under multi-turn dialog of multiple scenes, when the user initiates the dialog again after the dialog is detected to be interrupted and the dialog initiated by the user is matched with the stored scene identification, the dialog is continued according to the dialog scene description information corresponding to the stored scene identification. The problem that the previous conversation cannot be recovered once the conversation is interrupted in the multi-turn conversation process is solved, continuous and coherent multi-turn voice conversation is realized, the conversation efficiency is improved, and the time consumed by a user for obtaining an actual result is saved.
Example two
Fig. 2 is a flowchart of a question-answer interaction method in a multi-turn dialog scenario in the embodiment of the present invention, and the embodiment of the present invention is further refined on the basis of a technical solution of the embodiment of the present invention. Further embodying the interrupt type and the interrupt processing strategy, and correspondingly, the method of the embodiment of the invention specifically comprises the following steps:
step 210, when detecting that the current conversation belongs to a new conversation scene, establishing a scene identifier matched with the new conversation scene;
step 220, obtaining dialog scene description information in real time in the multi-turn question-answer interaction process matched with the current dialog, and correspondingly storing the dialog scene description information and the scene identification;
step 230, in the multi-round question-answer interaction process, if it is detected that the interaction response provided for the multi-round question-answer interaction is not responded within a first preset time, determining that the multi-round question-answer interaction is interrupted;
wherein it is detected that the interactive response provided for the multiple rounds of question-and-answer interaction is not responded within a first preset time, i.e. the user is silent for a long time, and no further indication is given to the system, which is an interrupt type
Step 240, detecting whether the user initiates a conversation again every second preset time;
if the user is silent for a long time, the system monitors the conversation state of the user continuously according to the set refreshing time, resets a new waiting time every preset time, and monitors whether the user initiates the conversation again continuously. Polling sessions are identified based on saved scenes, and polling is used to address server overstress. If multiple long connections are maintained, server stress may be excessive and, therefore. An interface of a polling request is specially established, only one task ID (identification) is reserved in the interface, and the condition of the current task can be acquired only by sending the task ID. If the result is returned, the polling is finished, and if the result is not returned, the polling is waited for a while and the request is continuously sent. That is, in the technical solution of the present application, each predetermined time is used to detect whether there is a session that is reinitiated and can correspond to a saved scene identifier, if there is a session, the session is continued according to the session scene description information corresponding to the saved scene identifier, and if not, the session is continued to wait.
In this embodiment, the first preset time and the second preset time may be preset according to an actual situation, which is not limited in this embodiment.
Step 250, if the user is determined to initiate a conversation again, acquiring a conversation scene matched with the conversation, and matching the conversation scene with at least one prestored scene identifier;
if the situation that the user initiates the conversation again is detected, the scene of the conversation initiated again is judged so as to be matched with the pre-stored scene identification, if the pre-stored scene identification cannot be matched with the scene of the conversation initiated again, the scene of the conversation initiated again is a new scene, the scene of the conversation initiated again is obtained as a brand new conversation scene, and the operation of establishing the scene identification matched with the brand new conversation scene is returned.
Step 260, when it is detected that the user initiates the dialogue matched with the scene identification again, continuing to perform question-answer interaction with the user according to the dialogue scene description information;
if the fact that the user initiates the conversation again is detected, and the scene identification matched with the scene of the conversation which is initiated again exists in the pre-stored scene identification, the stored conversation scene description information corresponding to the scene identification is called, and question-answer interaction with the user is continued according to the interaction problem which is proposed again by the user.
Step 270, after receiving the instruction of ending the conversation from the user, ending the conversation process;
the instruction for ending the dialog by the user can be directly sent to the system, for example, the user directly initiates the dialog to indicate that the voice dialog is closed; or the user can directly force the system to be closed at the terminal. The present embodiment does not limit this.
Step 280, deleting the saved scene identification and the corresponding dialog scene description information.
If the user has finished the dialog, it indicates that the dialog process has finished, and the system does not need to keep the saved scene identifier and the corresponding dialog scene description information. If the user does not actively end the session, the session may be reinitiated to continue the previous session, and at this time, the saved scene identifier and the corresponding session scene description information should be saved continuously, and it is continuously detected whether the user reinitiates the session.
According to the technical scheme of the embodiment of the invention, the related information of the conversation which is carried out is stored according to different scenes, so that the conversation can be quickly reacted when the user does not respond for a long time and initiates the conversation again, the related conversation scene description information is called after the conversation which is initiated again is matched with the stored conversation scene, the user can continue the previous conversation, an ideal result is quickly obtained, and the conversation efficiency is improved.
EXAMPLE III
Fig. 3 is a flowchart of a question-answer interaction method in a multi-turn dialog scenario in the embodiment of the present invention, and the embodiment of the present invention is further refined on the basis of a technical solution of the embodiment of the present invention. Further embodying the interrupt type and the interrupt processing strategy, and correspondingly, the method of the embodiment of the invention specifically comprises the following steps:
step 310, when detecting that the current conversation belongs to a new conversation scene, establishing a scene identification matched with the new conversation scene;
step 320, obtaining dialog scene description information in real time in the multi-turn question-answer interaction process matched with the current dialog, and correspondingly storing the dialog scene description information and the scene identification;
step 330, in the multi-turn question-answer interaction process, if it is detected that the new dialog does not belong to the dialog scene, determining that the multi-turn question-answer interaction is interrupted;
wherein the detection that a new dialog does not belong to the dialog scenario, i.e. the user has proposed a dialog different from the current dialog scenario, is the second category of the interruption types. For example, if a user initiates a voice conversation, queries a restaurant in a certain place, obtains a search result, and queries a movie theater in the certain place, at this time, the conversation scene is changed from dining to viewing, at this time, a switch of the conversation scene occurs, and the conversation process is interrupted.
Step 340, judging whether a scene identifier matched with the detected new conversation exists or not;
when the conversation scene is judged to be switched, after the conversation process is interrupted, judging whether the scene corresponding to the new conversation corresponds to the stored scene identification, if the stored scene identification is determined to have the scene corresponding to the scene identification, indicating that the conversation related to the scene has occurred, and continuing the conversation under the current scene according to the conversation scene description information corresponding to the scene identification.
Step 350, if the dialog scene description information exists, continuing to perform question-answer interaction with the user according to the dialog scene description information;
on the basis of the above embodiment, if it is determined that the scene identifier does not exist, returning to perform the operation of establishing the scene identifier matched with the new dialog scene; if there is no scene identifier matching with the new dialog, it indicates that no dialog in this scene has occurred, and therefore, the scene identifier matching with the new dialog is taken as a new dialog scene, and the operation returns to step 310.
Step 360, after receiving the command of the user for finishing the conversation, finishing the conversation process;
step 370, delete the saved scene id and the corresponding dialog scene description information.
On the basis of the above embodiment, the dialog scene description information includes at least one of the following: at least one user interaction question, at least one interaction response, and at least one contextual content. The advantage of this arrangement is that the user interaction question, the interaction response and the context content do not correspond one to one. Not all user interaction issues will get an interactive response of the system and not all dialogs will have contextual content. Moreover, one user interaction question may result in multiple interaction responses, and one interaction response may also correspond to multiple user interaction questions. Therefore, it is only necessary to limit the dialog scene description information to include the three types, and the dialog contents in the same scene are all stored as the dialog scene description information and correspond to the scene identifier.
The technical scheme of the embodiment of the invention can judge the conversation scene of the new conversation when the user switches the conversation scene, and if the stored scene identification can be matched with the conversation scene of the new conversation, the interactive response can be directly made according to the stored conversation scene description information, so that the conversation of the user is more coherent, and the conversation efficiency is improved.
Example four
Fig. 4 is a schematic structural diagram of a question-answer interaction device in a multi-turn dialog scenario according to an embodiment of the present invention, where the question-answer interaction device in the multi-turn dialog scenario includes: a scene identifier establishing module 410, a dialog scene description information saving module 420, an interrupt handling module 430, and a question-answer interaction proceeding module 440, wherein:
a scene identifier establishing module 410, configured to, when it is detected that the current session belongs to a new session scene, establish a scene identifier matching the new session scene;
a dialog scene description information storage module 420, configured to obtain dialog scene description information in real time during a multi-turn question-answer interaction process matching the current dialog, and store the dialog scene description information and the scene identifier correspondingly;
the interrupt processing module 430 is configured to execute an interrupt processing policy matching an interrupt type when detecting that the multi-round question-answer interaction is interrupted;
and a question-answer interaction continuing module 440, configured to continue to perform question-answer interaction with the user according to the dialog scene description information when it is detected that the user initiates a dialog matching the scene identifier again.
The question-answer interaction device in the multi-round conversation scene can realize continuous and coherent multi-round voice conversation according to the memory of the scene in the multi-round conversation scene of the multi-scene, and improve the conversation efficiency.
On the basis of the foregoing embodiments, the interrupt processing module 430 includes:
the first interruption judging unit is used for determining that the multi-round question-answer interaction is interrupted if detecting that the interaction response provided for the multi-round question-answer interaction is not responded within a first preset time in the multi-round question-answer interaction process;
and the conversation reinitiation detection unit is used for detecting whether the user reinitiates the conversation at intervals of second preset time.
On the basis of the foregoing embodiments, the interrupt processing module 430 further includes:
and the scene identification matching unit is used for detecting whether the user initiates the conversation again every second preset time, acquiring a conversation scene matched with the conversation if the user is determined to initiate the conversation again, and matching the conversation scene with at least one prestored scene identification.
On the basis of the foregoing embodiments, the interrupt processing module 430 includes:
a second interruption judging unit, configured to determine that the multi-turn question-answer interaction is interrupted if it is detected that a new dialog does not belong to the dialog scene in the multi-turn question-answer interaction process;
and the question-answer interaction continuing unit is used for continuing to perform question-answer interaction with the user according to the dialog scene description information matched with the scene identifier if the scene identifier matched with the detected new dialog is determined to exist.
On the basis of the foregoing embodiments, the interrupt processing module 430 further includes:
and the scene identification establishing unit is used for acquiring a scene corresponding to the new dialogue as a new dialogue scene and returning to execute the operation of establishing the scene identification matched with the new dialogue scene if the scene identification matched with the detected new dialogue does not exist after the multi-turn question-answer interaction is determined to be interrupted.
On the basis of the foregoing embodiments, the dialog scene description information includes at least one of the following items:
at least one user interaction question, at least one interaction response, and at least one contextual content.
On the basis of the above embodiments, the question-answer interaction device in the multi-turn dialog scene further includes:
the conversation process ending module is used for ending the conversation process after receiving a command of ending the conversation of the user;
and the information deleting module is used for deleting the stored scene identification and the corresponding conversation scene description information.
The question-answer interaction device in the multi-turn dialog scene provided by the embodiment of the invention can execute the question-answer interaction method in the multi-turn dialog scene provided by any embodiment of the invention, and has corresponding functional modules and beneficial effects of the execution method.
EXAMPLE five
Fig. 5 is a schematic structural diagram of an apparatus according to an embodiment of the present invention, as shown in fig. 5, the apparatus includes a processor 50, a memory 51, an input device 52, and an output device 53; the number of processors 50 in the device may be one or more, and one processor 50 is taken as an example in fig. 5; the processor 50, the memory 51, the input device 52 and the output device 53 in the apparatus may be connected by a bus or other means, which is exemplified in fig. 5.
The memory 51 is used as a computer-readable storage medium, and can be used for storing software programs, computer-executable programs, and modules, such as program instructions/modules corresponding to the question-answer interaction method in the multi-turn dialog scenario in the embodiment of the present invention (for example, the question-answer interaction device scenario identifier creating module 410, the dialog scenario description information saving module 420, the interrupt handling module 430, and the question-answer interaction proceeding module 450 in the multi-turn dialog scenario). The processor 50 executes various functional applications of the device and data processing by running software programs, instructions and modules stored in the memory 51, namely, implements the question-answer interaction method in the multi-turn dialog scenario described above. The method comprises the following steps:
when detecting that the current conversation belongs to a new conversation scene, establishing a scene identifier matched with the new conversation scene;
in the multi-turn question-answer interaction process matched with the current conversation, obtaining conversation scene description information in real time, and correspondingly storing the conversation scene description information and the scene identification;
executing an interrupt processing strategy matched with the interrupt type when the multi-round question-answer interaction is detected to be interrupted;
and when detecting that the user initiates the dialogue matched with the scene identification again, continuing to perform question-answer interaction with the user according to the dialogue scene description information.
The memory 51 may mainly include a storage program area and a storage data area, wherein the storage program area may store an operating system, an application program required for at least one function; the storage data area may store data created according to the use of the terminal, and the like. Further, the memory 51 may include high speed random access memory, and may also include non-volatile memory, such as at least one magnetic disk storage device, flash memory device, or other non-volatile solid state storage device. In some examples, the memory 51 may further include memory located remotely from the processor 50, which may be connected to the device over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The input device 52 is operable to receive input numeric or character information and to generate key signal inputs relating to user settings and function controls of the apparatus. The output device 53 may include a display device such as a display screen.
EXAMPLE six
Embodiments of the present invention also provide a storage medium containing computer-executable instructions, which when executed by a computer processor, are configured to perform a question-answer interaction method in a multi-turn dialog scenario, where the method includes:
when detecting that the current conversation belongs to a new conversation scene, establishing a scene identifier matched with the new conversation scene;
in the multi-turn question-answer interaction process matched with the current conversation, obtaining conversation scene description information in real time, and correspondingly storing the conversation scene description information and the scene identification;
executing an interrupt processing strategy matched with the interrupt type when the multi-round question-answer interaction is detected to be interrupted;
and when detecting that the user initiates the dialogue matched with the scene identification again, continuing to perform question-answer interaction with the user according to the dialogue scene description information.
Of course, the storage medium provided by the embodiment of the present invention contains computer-executable instructions, and the computer-executable instructions are not limited to the method operations described above, and may also perform related operations in the question-answer interaction method in a multi-turn dialog scenario provided by any embodiment of the present invention.
From the above description of the embodiments, it is obvious for those skilled in the art that the present invention can be implemented by software and necessary general hardware, and certainly, can also be implemented by hardware, but the former is a better embodiment in many cases. Based on such understanding, the technical solutions of the present invention may be embodied in the form of a software product, which may be stored in a computer-readable storage medium, such as a floppy disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a FLASH Memory (FLASH), a hard disk or an optical disk of a computer, and includes several instructions for enabling a computer device (which may be a personal computer, a server, or a network device) to execute the methods according to the embodiments of the present invention.
It should be noted that, in the embodiment of the question-answering interaction device in the multi-turn dialog scenario, the included units and modules are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be implemented; in addition, specific names of the functional units are only for convenience of distinguishing from each other, and are not used for limiting the protection scope of the present invention.
It is to be noted that the foregoing is only illustrative of the preferred embodiments of the present invention and the technical principles employed. It will be understood by those skilled in the art that the present invention is not limited to the particular embodiments described herein, but is capable of various obvious changes, rearrangements and substitutions as will now become apparent to those skilled in the art without departing from the scope of the invention. Therefore, although the present invention has been described in greater detail by the above embodiments, the present invention is not limited to the above embodiments, and may include other equivalent embodiments without departing from the spirit of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims (15)
1. A question-answer interaction method in a multi-turn dialogue scene is characterized by comprising the following steps:
when detecting that the current conversation belongs to a new conversation scene, establishing a scene identifier matched with the new conversation scene;
in the multi-turn question-answer interaction process matched with the current conversation, obtaining conversation scene description information in real time, and correspondingly storing the conversation scene description information and the scene identification;
executing an interrupt processing strategy matched with the interrupt type when the multi-round question-answer interaction is detected to be interrupted;
and when detecting that the user initiates the dialogue matched with the scene identification again, continuing to perform question-answer interaction with the user according to the dialogue scene description information.
2. The method of claim 1, wherein executing an interrupt handling policy matching an interrupt type upon detecting that the multiple rounds of question-answer interactions are interrupted comprises:
during the multiple rounds of question-answer interaction, if it is detected that the interaction response provided for the multiple rounds of question-answer interaction is not responded within a first preset time, determining that the multiple rounds of question-answer interaction are interrupted;
and detecting whether the user initiates the conversation again or not every second preset time.
3. The method of claim 2, after detecting whether the user initiates the dialog again at every second preset time, further comprising:
and if the user is determined to initiate the conversation again, acquiring a conversation scene matched with the conversation, and matching the conversation scene with at least one prestored scene identifier.
4. The method of claim 1, wherein executing an interrupt handling policy matching an interrupt type upon detecting that the multiple rounds of question-answer interactions are interrupted comprises:
in the multi-turn question-answer interaction process, if it is detected that a new conversation does not belong to the conversation scene, determining that the multi-turn question-answer interaction is interrupted;
and if the scene identification matched with the detected new dialogue is determined to exist, continuing to perform question-answering interaction with the user according to the dialogue scene description information matched with the scene identification.
5. The method of claim 4, after determining that the multiple rounds of question-answering interactions are interrupted, further comprising:
and if the scene identification matched with the detected new conversation does not exist, acquiring the scene corresponding to the new conversation as a brand new conversation scene, and returning to execute the operation of establishing the scene identification matched with the brand new conversation scene.
6. The method according to any of claims 1-5, wherein the dialog scenario description information comprises at least one of:
at least one user interaction question, at least one interaction response, and at least one contextual content.
7. The method according to claim 1, wherein after the step of continuing the question-answer interaction with the user according to the dialog scene description information when it is detected that the user initiates the dialog matching with the scene identifier again, the method further comprises:
when receiving an instruction of a user for finishing the conversation, finishing the conversation process;
and deleting the stored scene identification and the corresponding dialog scene description information.
8. A question-answer interaction device in a multi-turn dialog scenario, comprising:
the scene identification establishing module is used for establishing a scene identification matched with a new conversation scene when the current conversation is detected to belong to the new conversation scene;
the conversation scene description information storage module is used for acquiring the conversation scene description information in real time in the multi-turn question-answer interaction process matched with the current conversation and correspondingly storing the conversation scene description information and the scene identification;
the interrupt processing module is used for executing an interrupt processing strategy matched with the interrupt type when the multi-round question-answer interaction is detected to be interrupted;
and the question-answer interaction continuing module is used for continuing to carry out question-answer interaction with the user according to the dialog scene description information when detecting that the user initiates the dialog matched with the scene identifier again.
9. The apparatus of claim 8, wherein the interrupt handling module comprises:
the first interruption judging unit is used for determining that the multi-round question-answer interaction is interrupted if detecting that the interaction response provided for the multi-round question-answer interaction is not responded within a first preset time in the multi-round question-answer interaction process;
and the conversation reinitiation detection unit is used for detecting whether the user reinitiates the conversation at intervals of second preset time.
10. The apparatus of claim 9, wherein the interrupt handling module further comprises:
and the scene identification matching unit is used for detecting whether the user initiates the conversation again every second preset time, acquiring a conversation scene matched with the conversation if the user is determined to initiate the conversation again, and matching the conversation scene with at least one prestored scene identification.
11. The apparatus of claim 8, wherein the interrupt handling module comprises:
a second interruption judging unit, configured to determine that the multi-turn question-answer interaction is interrupted if it is detected that a new dialog does not belong to the dialog scene in the multi-turn question-answer interaction process;
and the question-answer interaction continuing unit is used for continuing to perform question-answer interaction with the user according to the dialog scene description information matched with the scene identifier if the scene identifier matched with the detected new dialog is determined to exist.
12. The apparatus of claim 11, wherein the interrupt handling module further comprises:
and the scene identification establishing unit is used for acquiring a scene corresponding to the new dialogue as a brand new dialogue scene and returning to execute the operation of establishing the scene identification matched with the brand new dialogue scene if the scene identification matched with the detected new dialogue does not exist after the multi-turn question-answer interaction is determined to be interrupted.
13. The apparatus according to any one of claims 8-12, wherein the dialog scene description information comprises at least one of:
at least one user interaction question, at least one interaction response, and at least one contextual content.
14. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the method of question-answer interaction in a multi-turn dialog scenario as claimed in any one of claims 1 to 7 when executing the program.
15. A storage medium containing computer-executable instructions for performing the method of question-answer interaction in a multi-turn dialog scenario of any of claims 1-7 when executed by a computer processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910906760.1A CN110609893A (en) | 2019-09-24 | 2019-09-24 | Question-answer interaction method, device, equipment and storage medium in multi-turn conversation scene |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910906760.1A CN110609893A (en) | 2019-09-24 | 2019-09-24 | Question-answer interaction method, device, equipment and storage medium in multi-turn conversation scene |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110609893A true CN110609893A (en) | 2019-12-24 |
Family
ID=68892043
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910906760.1A Pending CN110609893A (en) | 2019-09-24 | 2019-09-24 | Question-answer interaction method, device, equipment and storage medium in multi-turn conversation scene |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110609893A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111611368A (en) * | 2020-05-22 | 2020-09-01 | 北京百度网讯科技有限公司 | Method and device for backtracking common scene conversation in multi-turn conversation |
| CN111813912A (en) * | 2020-06-29 | 2020-10-23 | 北京百度网讯科技有限公司 | Man-machine conversation method, device, equipment and storage medium |
| CN112699221A (en) * | 2021-01-07 | 2021-04-23 | 中通天鸿（北京）通信科技股份有限公司 | Multi-scene task flow system in intelligent customer service |
| CN113488036A (en) * | 2020-06-10 | 2021-10-08 | 海信集团有限公司 | Multi-round voice interaction method, terminal and server |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107193978A (en) * | 2017-05-26 | 2017-09-22 | 武汉泰迪智慧科技有限公司 | A kind of many wheel automatic chatting dialogue methods and system based on deep learning |
| CN108197191A (en) * | 2017-12-27 | 2018-06-22 | 神思电子技术股份有限公司 | A kind of scene of more wheel dialogues is intended to interrupt method |
| US20180239837A1 (en) * | 2017-02-17 | 2018-08-23 | Salesforce.Com, Inc. | Intelligent embedded self-help service |
| CN109754806A (en) * | 2019-03-21 | 2019-05-14 | 问众智能信息科技（北京）有限公司 | A kind of processing method, device and the terminal of more wheel dialogues |
-
2019
- 2019-09-24 CN CN201910906760.1A patent/CN110609893A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20180239837A1 (en) * | 2017-02-17 | 2018-08-23 | Salesforce.Com, Inc. | Intelligent embedded self-help service |
| CN107193978A (en) * | 2017-05-26 | 2017-09-22 | 武汉泰迪智慧科技有限公司 | A kind of many wheel automatic chatting dialogue methods and system based on deep learning |
| CN108197191A (en) * | 2017-12-27 | 2018-06-22 | 神思电子技术股份有限公司 | A kind of scene of more wheel dialogues is intended to interrupt method |
| CN109754806A (en) * | 2019-03-21 | 2019-05-14 | 问众智能信息科技（北京）有限公司 | A kind of processing method, device and the terminal of more wheel dialogues |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111611368A (en) * | 2020-05-22 | 2020-09-01 | 北京百度网讯科技有限公司 | Method and device for backtracking common scene conversation in multi-turn conversation |
| CN113488036A (en) * | 2020-06-10 | 2021-10-08 | 海信集团有限公司 | Multi-round voice interaction method, terminal and server |
| CN111813912A (en) * | 2020-06-29 | 2020-10-23 | 北京百度网讯科技有限公司 | Man-machine conversation method, device, equipment and storage medium |
| CN112699221A (en) * | 2021-01-07 | 2021-04-23 | 中通天鸿（北京）通信科技股份有限公司 | Multi-scene task flow system in intelligent customer service |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110609893A (en) | Question-answer interaction method, device, equipment and storage medium in multi-turn conversation scene | |
| US11844121B2 (en) | Wireless connection establishment method, apparatus, device, and storage medium | |
| JP6305033B2 (en) | Method and system for providing a multi-user messenger service | |
| KR102429695B1 (en) | A method for initiating a group chat based on a shared document, an apparatus therefor, and a storage medium | |
| CN109712624A (en) | A kind of more voice assistant coordination approach, device and system | |
| KR102043365B1 (en) | Local maintenance of data for voice actions that can be selectively performed offline on a speech recognition electronic device | |
| WO2022016669A1 (en) | Bluetooth network configuration method, device, and storage medium | |
| CN112913219B (en) | Connection management method, system and computer readable medium for dynamic endpoint communication channel | |
| CN107623621B (en) | Chat corpus collection method and device | |
| US11271877B2 (en) | Primary chat bot service and secondary chat bot service integration | |
| CN107479957A (en) | A kind of switching method of Multiple system, device, equipment and storage medium | |
| JP2010134517A (en) | Device, method and program for message exchange | |
| CN110391970A (en) | For the message management system of communication applications, method, equipment and storage medium | |
| WO2019192203A1 (en) | Control system and method for air-conditioning device, storage medium and processor | |
| CN115424624B (en) | Man-machine interaction service processing method and device and related equipment | |
| CN111131005A (en) | Dialogue method, device, equipment and storage medium of customer service system | |
| CN110995706B (en) | Authentication system, method, device and storage medium for communication application | |
| CN107967363A (en) | A kind of data processing method, device and electronic equipment | |
| CN114036277A (en) | Dialogue robot route skipping method and device, electronic equipment and medium | |
| CN115529391A (en) | Method, device, equipment and storage medium for local area network voice call | |
| JP2019159880A (en) | Answering device, answering method, answering program and answering system | |
| CN108724203A (en) | A kind of exchange method and device | |
| US11082826B2 (en) | Emergency communication method and apparatus | |
| US20180212789A1 (en) | Communication Method and Storage Medium Storing Communication Program | |
| CN105991691B (en) | Method for transmitting information, equipment and system |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20191224 |
|
| RJ01 | Rejection of invention patent application after publication |