A multi-round interaction method, device and storage medium

Abstract

The present application discloses a multi-round interaction method, device and storage medium, which relates to the field of computer processing, and is used to solve the problem that after voice wake-up ends, the last round of interaction cannot be continued after waking up again. The method includes: after receiving the wake-up information, identifying the user's voice information; extracting features from the voice information to obtain sentence information in the voice information; if the interactive information matching the sentence information is found in the multi-round interaction database through semantic analysis, the reply information of the voice information is determined according to the sentence information and the interactive information; wherein the interactive information in the multi-round interaction database is obtained by extracting features from the user's voice information before receiving the wake-up information; and the reply information is broadcast. In this way, even if the user wakes up the interactive device again, the interactive device can still continue the last round of interaction, so that the multi-round interaction can be better completed, and the user experience during multi-round interaction is improved.

Description

Technical Field

[0001] The present application relates to the field of computer processing, and in particular to a multi-round interaction method, device and storage medium. Background Art

[0002] Many voice products lack the ability to interact multiple times, and the product must be awakened each time it is used, which causes a lot of trouble for users. Although some voice products have the ability to interact multiple times, multiple rounds of interaction are also limited to one round of awakening. After the voice awakening is completed, the next awakening cannot continue the previous round of interaction, which also brings inconvenience to the user's interaction. Summary of the invention

[0003] The embodiments of the present application provide a multi-round interaction method, device and storage medium to solve the problem that after voice wake-up ends, the previous round of interaction cannot be continued when waked up again.

[0004] In a first aspect, an embodiment of the present application provides a multi-round interaction method, the method comprising:

[0005] After receiving the wake-up message, recognize the user's voice message;

[0006] Extracting features from the speech information to obtain sentence information in the speech information;

[0007] If interaction information matching the sentence information is found in the multi-round interaction database through semantic analysis, reply information of the voice information is determined according to the sentence information and the interaction information; wherein the interaction information in the multi-round interaction database is obtained by extracting features of the user's voice information before receiving the wake-up information;

[0008] The reply information is broadcasted.

[0009] After receiving the wake-up information, the above method obtains the sentence information in the user's voice information and searches the multi-round interaction database for interaction information matching the sentence information through semantic analysis to determine the reply information. In this way, even if the user wakes up the interaction device again, the interaction device can still continue the previous round of interaction, so that the multi-round interaction can be completed better, improving the user's experience during the multi-round interaction.

[0010] In a possible implementation, the sentence information includes sentence entity information and sentence relationship information;

[0011] The extracting features of the speech information to obtain semantic information of the speech information includes:

[0012] Converting the voice information into text to obtain text information of the voice information;

[0013] Inputting the text information into at least two recognition models to obtain at least two sentence entity information in the speech information;

[0014] The sentence relationship information is determined according to the relationship between at least two sentence entity information obtained.

[0015] The above method obtains sentence entity information of voice information through a recognition model, and obtains sentence relationship information based on the obtained sentence entity information, so as to accurately understand the user's semantics and accurately perform multiple rounds of interaction.

[0016] In a possible implementation, inputting the text information into at least two recognition models to obtain sentence entity information in the speech information includes:

[0017] Inputting the text information into at least two recognition models to obtain sentence entity information corresponding to each recognition model;

[0018] If the sentence entity information obtained according to each recognition model is not completely the same, the sentence entity information with the largest number of occurrences is used as the sentence entity information in the speech information; or,

[0019] If the number of occurrences of each sentence entity information obtained according to each recognition model is the same, a sentence entity information is randomly selected as the sentence entity information in the speech information.

[0020] The above method recognizes text information through at least two recognition models, which can better obtain sentence entity information and improve the accuracy of sentence entity information.

[0021] In a possible implementation, the interaction information in the multi-round interaction database is stored in a manner in which sentence entity information is a node and sentence relationship information is a connection between nodes;

[0022] After finding the interaction information matching the sentence information in the multi-round interaction database through semantic analysis, the method further includes:

[0023] The sentence entity information and sentence relationship information extracted from the voice information are stored in the multi-round interaction database.

[0024] The above method stores the sentence entity information and sentence relationship information of the current voice information into the multi-round interaction database, which provides convenient conditions for subsequent multi-round interactions.

[0025] In a possible implementation, the method further includes:

[0026] If a shutdown instruction is received, the interaction information stored in the multi-round interaction database is deleted.

[0027] The above method deletes the data stored in the multi-round interaction database before shutting down, which can save the data storage space of the interactive device.

[0028] In a second aspect, an embodiment of the present application provides a multi-round interaction device, including:

[0029] A receiving module, used to recognize the user's voice information after receiving the wake-up information;

[0030] A feature extraction module, used to extract features from the voice information to obtain sentence information in the voice information;

[0031] A determination module, configured to determine reply information of the voice information according to the sentence information and the interaction information if interaction information matching the sentence information is found in the multi-round interaction database through semantic analysis; wherein the interaction information in the multi-round interaction database is obtained by extracting features of the user's voice information before receiving the wake-up information;

[0032] The broadcast module is used to broadcast the reply information.

[0033] In a possible implementation, the sentence information includes sentence entity information and sentence relationship information; and the feature extraction module includes:

[0034] A conversion unit, used for performing text conversion on the voice information to obtain text information of the voice information;

[0035] A recognition unit, used for inputting the text information into at least two recognition models to obtain at least two sentence entity information in the speech information;

[0036] The relationship determination unit is used to determine the sentence relationship information according to the relationship between at least two sentence entity information obtained.

[0037] In a possible implementation, the identification unit includes:

[0038] A sentence entity information acquisition subunit is used to input the text information into at least two recognition models to obtain sentence entity information corresponding to each recognition model;

[0039] A first sentence entity information determination subunit is configured to, if the sentence entity information obtained according to each recognition model is not completely the same, use the sentence entity information with the largest number of occurrences as the sentence entity information in the speech information; or

[0040] The second sentence entity information determination subunit is used to randomly select a sentence entity information as the sentence entity information in the speech information if the number of occurrences of the sentence entity information obtained according to the recognition models is the same.

[0041] In a possible implementation, the interaction information in the multi-round interaction database is stored in a manner of sentence entity information as nodes and sentence relationship information as connections between nodes; the device further includes:

[0042] The storage module is used to store the sentence entity information and sentence relationship information extracted from the voice information into the multi-round interaction database after the determination module finds the interaction information matching the sentence information in the multi-round interaction database through semantic analysis.

[0043] In a possible implementation manner, the device further includes:

[0044] The deleting module is configured to delete the interaction information stored in the multi-round interaction database if a shutdown instruction is received.

[0045] According to a third aspect, a computing device is provided, comprising at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program, and when the program is executed by the processing unit, the processing unit executes the steps of any one of the above-mentioned multi-round interaction methods.

[0046] In one embodiment, the computing device may be a server or a terminal device.

[0047] In a fourth aspect, a computer-readable medium is provided, which stores a computer program executable by a terminal device. When the program runs on the terminal device, the terminal device executes the steps of any one of the above-mentioned multi-round interaction methods.

[0048] Other features and advantages of the present application will be described in the following description, and partly become apparent from the description, or be understood by practicing the present application. The purpose and other advantages of the present application can be realized and obtained by the structures specifically pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0050] Figure 1 A schematic diagram of a multi-round interaction method in an embodiment of the present application;

[0051] Figure 2A schematic diagram of a process for obtaining sentence entity information and sentence relationship information in an embodiment of the present application;

[0052] Figure 3 This is a schematic diagram of the structure of a multi-round interaction device in an embodiment of the present application;

[0053] Figure 4 It is a schematic diagram of the structure of a computing device according to an embodiment of the present application. DETAILED DESCRIPTION

[0054] In order to solve the problem in the prior art that after voice wake-up, the last round of interaction cannot be continued when waked up again, the embodiments of the present application provide a multi-round interaction method, device and storage medium. In order to better understand the technical solution provided by the embodiments of the present application, the basic principle of the solution is briefly explained here:

[0055] Although existing voice products have multi-round interaction capabilities, the multi-round interaction is also limited to a limited number of rounds of interaction. After the voice wake-up is completed, the next wake-up cannot continue the previous round of interaction, which will also bring inconvenience to the user's interaction.

[0056] In view of this, the embodiments of the present application provide a multi-round interaction method, device and storage medium. After receiving the wake-up information, the sentence information in the user's voice information is obtained, and the interaction information matching the sentence information is found in the multi-round interaction database through semantic analysis to determine the reply information. In this way, even if the user wakes up the interaction device again, the interaction device can still continue the previous round of interaction, so that the multi-round interaction can be better completed, and the user experience during the multi-round interaction is improved.

[0057] The preferred embodiments of the present application are described below in conjunction with the drawings in the specification. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present application, and are not used to limit the present application. In addition, the embodiments and features in the embodiments of the present application may be combined with each other if there is no conflict.

[0058] The following further explains the multi-round interaction method provided in the embodiment of the present application. Figure 1 As shown, the following steps are included:

[0059] S101: After receiving the wake-up message, recognize the user's voice information.

[0060] In an embodiment of the present application, if a user interacts with an interactive device and does not interact again within a preset time, the interactive device will enter a dormant state and can only continue to interact after being awakened again.

[0061] The wake-up information may be a wake-up word or a wake-up gesture, etc. For example, if the wake-up word of the interactive device is "Hello", when the interactive device is in a dormant state, the user says to the interactive device "Hello, play a song for me", and at this time, the wake-up word exists in the user's voice information, so the interactive device enters the awake state and interacts with the user.

[0062] S102: Extract features from the voice information to obtain sentence information in the voice information.

[0063] Among them, sentence information includes sentence entity information and sentence relationship information. Sentence entity information refers to the entities in this sentence, and sentence relationship information refers to the relationship between the entities in this sentence. Usually, sentence entity information is nouns, and sentence relationship information is verbs. For example, if the voice information is "Xiaohong is going to Beijing tomorrow", then the sentence entity information of this sentence is "Xiaohong" and "Beijing", and the sentence relationship information is "go", that is, the connection between "Xiaohong" and "Beijing".

[0064] In the embodiment of the present application, to obtain the sentence information of the voice information, it is necessary to convert the voice information into text information and recognize the text information, which can be specifically implemented as steps A1-A3:

[0065] Step A1: convert the voice information into text to obtain text information of the voice information.

[0066] Step A2: input the text information into at least two recognition models to obtain at least two sentence entity information in the speech information.

[0067] There are usually two sentence entity information in a sentence, but sometimes there may be more than two sentence entity information.

[0068] In the embodiment of the present application, in order to better obtain sentence entity information and improve the accuracy of sentence entity information, text information is input into at least two recognition models, wherein each recognition model is different, such as the toolkit jiagu, the Bert pre-trained model, and the BiLSTM model.

[0069] The sentence entity information is obtained according to different recognition models, so as to select the most accurate sentence entity information.

[0070] Specifically, it can be implemented as follows: inputting the text information into at least two recognition models to obtain sentence entity information corresponding to each recognition model.

[0071] If the sentence entity information obtained according to each recognition model is not completely the same, the sentence entity information that appears most times is used as the sentence entity information in the speech information.

[0072] For example: if there are three recognition models, and after recognition, sentence entity information A appears twice, and sentence entity information B appears once, then sentence entity information A is taken as the sentence entity information in the voice information.

[0073] If the number of occurrences of each sentence entity information obtained according to each recognition model is the same, a sentence entity information is randomly selected as the sentence entity information in the speech information.

[0074] For example: if there are three recognition models, and after recognizing that sentence entity information A, sentence entity information B and sentence entity information C each appear once, then one is randomly selected from ABC as the sentence entity information in the voice information.

[0075] Step A3: Determine the sentence relationship information according to the relationship between at least two sentence entity information obtained.

[0076] like Figure 2 As shown in FIG. 1 , it is a flow chart of obtaining sentence entity information and sentence relationship information. After the user inputs voice information, the voice information is converted into text information, and the text information is input into three recognition models. The sentence entity information is obtained through the output results of the three recognition models, and the sentence concern information is determined based on the sentence entity information.

[0077] In this way, the sentence entity information of the voice information is obtained through the recognition model, and the sentence relationship information is obtained based on the obtained sentence entity information, so that the user's semantics can be accurately understood, and multiple rounds of interaction can be accurately performed.

[0078] S103: If interaction information matching the sentence information is found in the multi-round interaction database through semantic analysis, the reply information of the voice information is determined according to the sentence information and the interaction information; wherein the interaction information in the multi-round interaction database is obtained by extracting features of the user's voice information before receiving the wake-up information.

[0079] Among them, the interaction information in the multi-round interaction database is stored in the form of sentence entity information as nodes and sentence relationship information as connections between nodes;

[0080] In an embodiment of the present application, after the interaction information matching the sentence information is found in the multi-round interaction database, the sentence entity information and sentence relationship information extracted from the voice information are stored in the multi-round interaction database.

[0081] In this way, the sentence entity information and sentence relationship information of the current voice information are stored in the multi-round interaction database, which provides convenient conditions for subsequent multi-round interactions.

[0082] S104: broadcasting the reply information.

[0083] In this way, even if the user wakes up the interactive device again, the interactive device can still continue the previous round of interaction, so that multiple rounds of interaction can be better completed, improving the user experience when performing multiple rounds of interaction.

[0084] In an embodiment of the present application, if the interactive device is about to shut down, the data stored in the multi-round interaction database needs to be deleted, which can be specifically implemented as follows: if a shutdown command is received, the interaction information stored in the multi-round interaction database is deleted.

[0085] In this way, the data stored in the multi-round interaction database is deleted before shutting down, which can save the data storage space of the interactive device.

[0086] Based on the same inventive concept, the present application embodiment also provides a multi-round interaction device. Figure 3 As shown, the device comprises:

[0087] The receiving module 301 is used to recognize the user's voice information after receiving the wake-up information;

[0088] A feature extraction module 302 is used to extract features from the voice information to obtain sentence information in the voice information;

[0089] The determination module 303 is used to determine the reply information of the voice information according to the sentence information and the interaction information if the interaction information matching the sentence information is found in the multi-round interaction database through semantic analysis; wherein the interaction information in the multi-round interaction database is obtained by extracting features from the user's voice information before receiving the wake-up information;

[0090] The reporting module 304 is used to report the reply information.

[0091] In a possible implementation, the sentence information includes sentence entity information and sentence relationship information; the feature extraction module 302 includes:

[0092] A conversion unit, used for performing text conversion on the voice information to obtain text information of the voice information;

[0093] A recognition unit, used for inputting the text information into at least two recognition models to obtain at least two sentence entity information in the speech information;

[0094] The relationship determination unit is used to determine the sentence relationship information according to the relationship between at least two sentence entity information obtained.

[0095] In a possible implementation, the identification unit includes:

[0096] A sentence entity information acquisition subunit is used to input the text information into at least two recognition models to obtain sentence entity information corresponding to each recognition model;

[0097] A first sentence entity information determination subunit is configured to, if the sentence entity information obtained according to each recognition model is not completely the same, use the sentence entity information with the largest number of occurrences as the sentence entity information in the speech information; or

[0098] The second sentence entity information determination subunit is used to randomly select a sentence entity information as the sentence entity information in the speech information if the number of occurrences of the sentence entity information obtained according to the recognition models is the same.

[0099] In a possible implementation, the interaction information in the multi-round interaction database is stored in a manner of sentence entity information as nodes and sentence relationship information as connections between nodes; the device further includes:

[0100] The storage module is used to store the sentence entity information and sentence relationship information of the extracted voice information into the multi-round interaction database after the determination module 303 finds the interaction information matching the sentence information in the multi-round interaction database through semantic analysis.

[0101] In a possible implementation manner, the device further includes:

[0102] The deleting module is configured to delete the interaction information stored in the multi-round interaction database if a shutdown instruction is received.

[0103] Based on the same technical concept, the present application embodiment also provides a terminal device 400, referring to Figure 4 As shown, the terminal device 400 is used to implement the methods described in the above-mentioned various method embodiments, for example, Figure 2 In the illustrated embodiment, the terminal device 400 may include a memory 401 , a processor 402 , an input unit 403 , and a display panel 404 .

[0104] The memory 401 is used to store the computer program executed by the processor 402. The memory 401 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function, etc.; the data storage area may store data created according to the use of the terminal device 400, etc. The processor 402 may be a central processing unit (CPU) or a digital processing unit, etc. The input unit 403 may be used to obtain user instructions input by the user. The display panel 404 is used to display information input by the user or information provided to the user. In the embodiment of the present application, the display panel 404 is mainly used to display the display interface of each application in the terminal device and the control entity displayed in each display interface. Optionally, the display panel 404 may be configured in the form of a liquid crystal display (LCD) or an OLED (organic light-emitting diode).

[0105] The specific connection medium between the memory 401, the processor 402, the input unit 403 and the display panel 404 is not limited in the embodiment of the present application. Figure 4 In the embodiment, the memory 401, the processor 402, the input unit 403, and the display panel 404 are connected via a bus 405. Figure 4 The connections between other components are shown in bold lines, which are only for illustration and are not intended to be limiting. Bus 405 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 4 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0106] The memory 401 may be a volatile memory, such as a random-access memory (RAM); the memory 401 may also be a non-volatile memory, such as a read-only memory, a flash memory, a hard disk drive (HDD) or a solid-state drive (SSD), or the memory 401 may be any other medium that can be used to carry or store the desired program code in the form of instructions or data structures and can be accessed by a computer, but is not limited thereto. The memory 401 may be a combination of the above memories.

[0107] Processor 402, used to implement Figure 1 The embodiment shown includes:

[0108] Processor 402, used to call the computer program stored in memory 401 to execute the implementation Figure 1 The embodiment shown.

[0109] An embodiment of the present application also provides a computer-readable storage medium that stores computer-executable instructions required to execute the above-mentioned processor, which includes a program required to execute the above-mentioned processor.

[0110] In some possible implementations, various aspects of a multi-round interaction method provided by the present application may also be implemented in the form of a program product, which includes a program code. When the program product is run on a terminal device, the program code is used to enable the terminal device to execute the steps of a multi-round interaction method according to various exemplary implementations of the present application described above in this specification. For example, the terminal device may execute the implementation Figure 1 The embodiment shown.

[0111] The program product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. The readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or device, or any combination thereof. More specific examples of readable storage media (a non-exhaustive list) include: an electrical connection with one or more wires, a portable disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof.

[0112] The multi-round interactive program product of the embodiment of the present application can adopt a portable compact disk read-only memory (CD-ROM) and include program code, and can be run on a computing device. However, the program product of the present application is not limited to this. In this document, a readable storage medium can be any tangible medium containing or storing a program, which can be used by or in combination with an instruction execution system, an apparatus or a device.

[0113] A readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, wherein readable program code is carried. Such propagated data signals may take a variety of forms, including, but not limited to, electromagnetic signals, optical signals, or any suitable combination of the foregoing. A readable signal medium may also be any readable medium other than a readable storage medium, which may transmit, propagate, or transfer a program for use by or in conjunction with an instruction execution system, apparatus, or device.

[0114] The program code embodied on the readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

[0115] Program code for performing the operations of the present application may be written in any combination of one or more programming languages, including entity-oriented programming languages ​​such as Java, C++, etc., and conventional procedural programming languages ​​such as "C" or similar programming languages. The program code may be executed entirely on the user computing device, partially on the user device, as a separate software package, partially on the user computing device and partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computing device (e.g., via the Internet using an Internet service provider).

[0116] It should be noted that, although several units or subunits of the device are mentioned in the above detailed description, this division is merely exemplary and not mandatory. In fact, according to the embodiments of the present application, the features and functions of two or more units described above can be embodied in one unit. Conversely, the features and functions of one unit described above can be further divided into multiple units to be embodied.

[0117] In addition, although the operations of the method of the present application are described in a specific order in the drawings, this does not require or imply that the operations must be performed in this specific order, or that all the operations shown must be performed to achieve the desired results. Additionally or alternatively, some steps may be omitted, multiple steps may be combined into one step, and / or one step may be decomposed into multiple steps.

[0118] Those skilled in the art will appreciate that the embodiments of the present application may be provided as methods, systems, or computer program products. Therefore, the present application may adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment in combination with software and hardware. Moreover, the present application may adopt the form of a computer program product implemented in one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.

[0119] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable file processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable file processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0120] These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable file processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0121] These computer program instructions may also be loaded onto a computer or other programmable file processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0122] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications falling within the scope of the present application.

[0123] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

Claims

1. A multi-round interaction method, characterized in that: The method comprises: After receiving the wake-up message, recognize the user's voice message; Converting the speech information into text to obtain text information of the speech information; inputting the text information into at least two recognition models to obtain at least two sentence entity information in the speech information; determining sentence relationship information according to the relationship between the at least two sentence entity information obtained; If interaction information matching the sentence information including the sentence entity information and the sentence relationship information is found in the multi-round interaction database through semantic analysis, the reply information of the voice information is determined according to the sentence information and the interaction information; wherein the interaction information in the multi-round interaction database is obtained by extracting features from the user's voice information before receiving the wake-up information; After finding the interactive information matching the sentence information including the sentence entity information and the sentence relationship information in the multi-round interactive database through semantic analysis, the sentence entity information and the sentence relationship information extracted from the voice information are stored in the multi-round interactive database; the interactive information in the multi-round interactive database is stored in the form of sentence entity information as nodes and sentence relationship information as connections between nodes; The reply information is broadcasted.

2. The method according to claim 1, characterized in that The step of inputting the text information into at least two recognition models to obtain at least two sentence entity information in the speech information includes: Inputting the text information into at least two recognition models to obtain sentence entity information corresponding to each recognition model; If the sentence entity information obtained according to each recognition model is not completely the same, the sentence entity information with the largest number of occurrences is used as the sentence entity information in the speech information; or, If the number of occurrences of each sentence entity information obtained according to each recognition model is the same, a sentence entity information is randomly selected as the sentence entity information in the speech information.

3. The method according to any one of claims 1 to 2, characterized in that: The method further comprises: If a shutdown instruction is received, the interaction information stored in the multi-round interaction database is deleted.

4. A multi-round interaction device, characterized in that: The device comprises: A receiving module, used to recognize the user's voice information after receiving the wake-up information; The feature extraction module includes a conversion unit, a recognition unit and a relationship determination unit. The conversion unit is used to perform text conversion on the voice information to obtain text information of the voice information; the recognition unit is used to input the text information into at least two recognition models to obtain at least two sentence entity information in the voice information; the relationship determination unit is used to determine sentence relationship information according to the relationship between the at least two sentence entity information obtained; A determination module, configured to determine reply information of the voice information according to the sentence information and the interaction information if interaction information matching the sentence information including the sentence entity information and the sentence relationship information is found in the multi-round interaction database through semantic analysis; wherein the interaction information in the multi-round interaction database is obtained by extracting features from the user's voice information before receiving the wake-up information; A storage module, configured to store the extracted sentence entity information and sentence relationship information of the voice information in the multi-round interaction database after finding the interaction information matching the sentence information including the sentence entity information and the sentence relationship information in the multi-round interaction database through semantic analysis; the interaction information in the multi-round interaction database is stored in a manner in which the sentence entity information is a node and the sentence relationship information is a connection between the nodes; The broadcast module is used to broadcast the reply information.

5. The device according to claim 4, characterized in that The identification unit includes: A sentence entity information acquisition subunit is used to input the text information into at least two recognition models to obtain sentence entity information corresponding to each recognition model; A first sentence entity information determination subunit is configured to, if the sentence entity information obtained according to each recognition model is not completely the same, use the sentence entity information with the largest number of occurrences as the sentence entity information in the speech information; or The second sentence entity information determination subunit is used to randomly select a sentence entity information as the sentence entity information in the speech information if the number of occurrences of the sentence entity information obtained according to the recognition models is the same.

6. The device according to any one of claims 4 to 5, characterized in that: The device also includes: The deleting module is configured to delete the interaction information stored in the multi-round interaction database if a shutdown instruction is received.

7. An electronic device, characterized in that: The method comprises a processor and a memory, wherein the memory stores program codes, and when the program codes are executed by the processor, the processor executes the steps of any one of the methods of claims 1 to 3.

8. A computer-readable storage medium, characterized in that: It includes program codes. When the program product is run on an electronic device, the program codes are used to enable the electronic device to execute the steps of any one of the methods described in claims 1 to 3.

Images