Data generation method and device, electronic equipment and storage medium

Abstract

The invention provides a data generation method and device, electronic equipment and a storage medium, and the method is applied to the field of intelligent interaction.The method comprises the steps that interaction data input by a user is received, user identity information of the user is determined based on the interaction data, and a feature description text corresponding to the user identity information is generated; determining text information corresponding to the interaction data, and inputting the text information and the feature description text into the interaction model to generate reply data corresponding to the interaction data. According to the method, the interaction model can output the reply data of the feature description text corresponding to the user identity information when receiving the interaction data input by the user, and the reply data output by the interaction model is matched with the user identity information, so that personalized service is provided for the user.

Description

Technical Field

[0001] The present application relates to the field of intelligent interaction, and more specifically, to a data generation method, device, electronic device and storage medium in the field of intelligent interaction. Background Art

[0002] Natural language processing technology has made rapid progress in recent years, and natural language models can usually be used for intelligent interaction. Natural language models are mainly trained based on large-scale corpora. Although they can process a wide range of text data, they cannot provide personalized services during intelligent interaction due to different training data. Summary of the invention

[0003] The present application provides a data generation method, device, electronic device and storage medium. The method enables an interaction model to output reply data corresponding to the user's identity information when receiving interaction data input by the user. The reply data output by the interaction model is highly matched with the user's identity information, thereby providing personalized services for the user.

[0004] In a first aspect, a data generation method is provided, the method comprising: receiving interaction data input by a user, and determining user identity information of the user based on the interaction data;

[0005] Generate a feature description text corresponding to the user identity information;

[0006] Text information corresponding to the interaction data is determined, and the text information and the feature description text are input into the interaction model to generate response data corresponding to the interaction data.

[0007] In the embodiment of the present application, after obtaining the interaction data input by the user, the user identity information of the user is determined based on the interaction data, and then the feature description text corresponding to the user identity information is obtained, and the feature description text and the interaction data are input into the interaction model, and the interaction model generates corresponding reply data for the input feature description text and interaction data. Since the interaction model also uses the feature description text corresponding to the user identity information when generating the reply data corresponding to the interaction data, the reply data is not only related to the interaction data, but also to the user's feature description text, which improves the correlation between the reply data and the user's feature description text and meets the user's personalized needs.

[0008] In a second aspect, a data generating device is provided, the device comprising:

[0009] A receiving unit, configured to receive interaction data input by a user, and determine user identity information of the user based on the interaction data;

[0010] A generating unit, used to generate a feature description text corresponding to the user identity information;

[0011] The determination unit is used to determine text information corresponding to the interaction data, and input the text information and the feature description text into the interaction model to generate reply data corresponding to the interaction data.

[0012] In a third aspect, an electronic device is provided, comprising: a memory for storing executable program code; and a processor for calling and running the executable program code from the memory, so that the electronic device executes the method in the above-mentioned first aspect or any possible implementation of the first aspect.

[0013] In a fourth aspect, a computer program product is provided, comprising: a computer program code, which, when executed on a computer, enables the computer to execute the method in the first aspect or any possible implementation of the first aspect.

[0014] In a fifth aspect, a computer-readable storage medium is provided, which stores a computer program code. When the computer program code runs on a computer, the computer executes the method in the above-mentioned first aspect or any possible implementation manner of the first aspect. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a scenario schematic diagram of a data generation method provided in an embodiment of the present application;

[0016] Figure 2 It is a flowchart of a data generation method provided in an embodiment of the present application;

[0017] Figure 3 It is a flowchart of a data generation method provided in an embodiment of the present application;

[0018] Figure 4 It is a flowchart of a data generation method provided in an embodiment of the present application;

[0019] Figure 5 This is a schematic diagram of a feature information input interface of a data generation method provided in an embodiment of the present application;

[0020] Figure 6 It is a schematic diagram of an interactive interface of a data generation method provided in an embodiment of the present application;

[0021] Figure 7 It is a structural schematic diagram of a data generating device provided in an embodiment of the present application;

[0022] Figure 8 It is a structural schematic diagram of a data generating device provided in an embodiment of the present application;

[0023] Fig. 9 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0024] The technical solution in the present application will be described clearly and in detail below in conjunction with the accompanying drawings. In the description of the embodiments of the present application, unless otherwise specified, " / " means or, for example, A / B can mean A or B: "and / or" in the text is only a description of the association relationship of associated objects, indicating that there can be three relationships, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present application, "multiple" means two or more than two.

[0025] In the following, the terms "first" and "second" are used for descriptive purposes only and are not to be understood as suggesting or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features.

[0026] Figure 1 It is a scenario diagram of a data generation method provided in an embodiment of the present application.

[0027] like Figure 1 As shown, the data generation method provided in the embodiment of this specification can be applied to the terminal device 10 to realize the data generation process. The system architecture provided in the embodiment of this specification mainly includes the terminal device 10 and the user 20. Among them, the terminal device 10 can be a terminal device with data generation, including but not limited to wearable devices, such as watches, etc., as well as conference tablets, mobile phones, personal computers, laptops, smart furniture, vehicle-mounted devices, wearable devices and other terminal devices.

[0028] In related technologies, natural language processing (NLP) has been widely used in intelligent customer service, voice assistants, search engines and other fields. Through the training of massive data, it has achieved remarkable results in text understanding, semantic matching, answer extraction and other aspects, and can provide users with fast and accurate answers. However, since a large amount of data is required to train the corresponding interaction model using natural language processing technology, if it is required to provide targeted replies to different users, large-scale retraining is required, that is, the current interaction model cannot meet the personalized needs of users.

[0029] In the embodiment of the present application, after obtaining the interaction data input by the user 20, the terminal device 10 determines the user identity information of the user 20, and then obtains the feature description text corresponding to the user identity information, inputs the feature description text and the interaction data into the interaction model, and generates corresponding reply data for the input feature description text and interaction data through the interaction model. Since the interaction model also uses the feature description text corresponding to the user identity information when generating the reply data corresponding to the interaction data, the reply data is not only related to the interaction data, but also to the user's feature description text, which improves the correlation between the reply data and the user's feature description text and meets the user's personalized needs.

[0030] based on Figure 1 The scene shown is shown below. Figure 2-Figure 9 , the data generation method provided in the embodiment of the present application is introduced in detail.

[0031] See also Figure 2 , which is a flow chart of a data generation method provided in an embodiment of the present application. Figure 2 As shown, the method of the embodiment of the present application may include the following steps S101-S103.

[0032] S101, receiving interaction data input by a user, and determining user identity information of the user based on the interaction data;

[0033] In one embodiment, the interaction data may be data used to control a conversation with an electronic device, which may include a request instruction, such as "please play a piece of light music", statement information, such as "I am in a good mood today", question information, such as "are there any movies you can recommend", etc. User identity information may be information that uniquely identifies a user, and illustratively, it may be the user's name.

[0034] In one embodiment, a microphone device is installed on the electronic device, so that the user can input the interaction data by voice. In this case, the interaction data can be subjected to voiceprint recognition to obtain the user's user identity information. Specifically, the user's user identity information is identified by analyzing the frequency, pitch, speech speed, pronunciation method, etc. in the interaction data. In the process of voiceprint recognition of interaction data, a user's voiceprint model is first established, and the voiceprint model contains the user's voice features. Then, the user's interaction data is compared with the features in the model to verify the user's user identity information.

[0035] In another embodiment, the user may also input the interaction data in the form of text. For example, the user may input the interaction data of "I am Xiao A, please start playing light music" through the keyboard of the electronic device. The electronic device can thereby determine the user's user identity information. In another case, when the user inputs the interaction data in the form of text, the electronic device may also capture the user's facial data through an installed camera device, and then perform facial recognition on the facial data to obtain the user's user identity information.

[0036] S102, generating a feature description text corresponding to the user identity information;

[0037] In one embodiment, the feature description text can constrain the response data generated by the interaction model to be related to the user identity information when the interaction model generates response data corresponding to the interaction data. After determining the user identity information of the user, the electronic device generates the feature description text corresponding to the user identity information based on the acquired user identity information.

[0038] Optionally, in the embodiment of the present application, after obtaining the user identity information, the user's gender, age and other information can be determined through the user identity information, and then the corresponding feature description text can be generated. For example, the user's gender is determined to be "male" and the age is 3 years old through the user identity information (small A), and the generated feature description text can be "boy".

[0039] S103, determining text information corresponding to the interaction data, and inputting the text information and the feature description text into the interaction model to generate reply data corresponding to the interaction data.

[0040] In one embodiment, after the interaction data is obtained, the text information corresponding to the interaction data is determined, and then the text information and the feature description text are input into the interaction model together. The interaction model is a pre-trained model, which can generate corresponding reply data based on the input text information and feature description text, and the reply data is data generated based on the feature description text and related to the user identity information of the user, and corresponding to the input interaction data.

[0041] It can be understood that in the embodiment of the present application, if the interaction data received from the user is interaction data input by voice, the text information corresponding to the interaction data can be obtained by voice recognition.

[0042] In another embodiment, if the user inputs the interaction data in a textual manner, it is determined that the interaction data is textual information.

[0043] In the embodiment of the present application, after obtaining the interaction data input by the user, the user identity information of the user is determined based on the interaction data, and then the feature description text corresponding to the user identity information is obtained, and the feature description text and the interaction data are input into the interaction model, and the interaction model generates corresponding reply data for the input feature description text and interaction data. Since the interaction model also uses the feature description text corresponding to the user identity information when generating the reply data corresponding to the interaction data, the reply data is not only related to the interaction data, but also to the user's feature description text, which improves the correlation between the reply data and the user's feature description text and meets the user's personalized needs.

[0044] See also Figure 3 , which is a flow chart of a data generation method provided in an embodiment of the present application. Figure 3 As shown, the method of the embodiment of the present application may include the following steps S201-S203.

[0045] S201, receiving interaction data input by a user, and determining user identity information of the user based on the interaction data;

[0046] Specifically, the manner of receiving the interaction data input by the user and determining the user identity information of the user based on the interaction data can be found in the description of step S101 in the above embodiment of the specification, which will not be elaborated here.

[0047] S202, acquiring characteristic information corresponding to the user identity information, the characteristic information including the user's name information, age information, personality information, hobby information, and education information;

[0048] In one embodiment, after obtaining the user identity information of the user, further obtain characteristic information of the user identity information, which is information describing the user's characteristics, behaviors, preferences, needs, etc. In the embodiment of the present application, the characteristic information includes the user's name information, age information, personality information, hobby information, and education information.

[0049] Specifically, in an embodiment of the present application, the user identity information and its corresponding feature information may be mapped and stored, so that after the user identity information of the user is determined, the feature information corresponding to the user identity information may be obtained based on the mapping relationship.

[0050] S203, filtering out target feature information from the feature information, and determining a feature description text corresponding to the target feature information.

[0051] In one embodiment, after determining the characteristic information of the user, the electronic device determines the target characteristic information from the characteristic information, and further determines the target characteristic information.

[0052] Optionally, in the embodiment of the present application, the target feature information is screened out from the feature information, and the feature description text corresponding to the target feature information is determined, including:

[0053] S2031, determining whether the interaction data is interaction data received from the user input for the first time;

[0054] S2032, if yes, determine the name information and hobby information as target feature information;

[0055] In one case, after receiving the user's interactive data and identifying the user's user identity information, the electronic device checks whether there is a record of outputting reply data to the user. If not, it is determined that the currently received interactive data is the first interactive data received from the user. Further, the characteristic information of the user is obtained according to the user's user identity information, and then the name information and the hobby information are determined as the target characteristic information of the user in the characteristic information (name information, age information, personality information, hobby information, education information).

[0056] It is understandable that when the interactive data sent by the user is received for the first time, the electronic device has not stored any record of the interaction with the user, and it is unable to select the target feature information that can be accurately selected from the feature information. Therefore, in an embodiment of the present application, hobby information and name information can be selected from the feature information as the target feature information. When the interactive data sent by the user is received for the first time, it is avoided to use too much or too detailed feature information as the target feature information, that is, to avoid introducing features that cause data sparsity or complicate the model, resulting in the generated reply data not meeting user expectations, thereby improving the accuracy of the generated reply data.

[0057] S2033, if not, determining the user's current emotional tendency based on the interaction data, and determining target feature information in the feature information based on the current emotional tendency;

[0058] In another case of the present application, if it is determined that this is not the first time that the interaction data sent by the user is received, the user's current emotional tendency is determined based on the current interaction data, and then the target feature information is determined in the feature information.

[0059] Optionally, determining the user's current emotional tendency based on the current interaction data includes: after obtaining the interaction data input by the user in the form of voice, removing the background noise of the interaction data to improve the accuracy of subsequent analysis, and then extracting the acoustic feature vector in the interaction data through signal processing technology (such as Fourier transform, wavelet transform, etc.) and speech analysis software (such as Praat, HTK, etc.), and finally inputting the acoustic feature vector into a preset emotion evaluation model, and determining the user's current emotional tendency based on the output result of the emotion evaluation model, wherein the acoustic feature vector is a set of key information that can reflect the speech emotion in the interaction data input in the form of audio. Common acoustic features include: volume, fundamental frequency, resonance peak, short-time energy and short-time zero-crossing rate, Mel frequency cepstral coefficients (MFCCs), intonation and speaking speed, etc. Among them:

[0060] Volume: reflects the volume of interactive data input by voice;

[0061] Fundamental frequency: represents the pitch of speech and its changes, which is closely related to emotions;

[0062] Formant: The resonant frequency of the vocal tract, which reflects the way of pronunciation;

[0063] Short-time energy and short-time zero-crossing rate: used to distinguish between unvoiced and voiced sounds;

[0064] Mel frequency cepstral coefficients (MFCCs): features that simulate the hearing characteristics of the human ear and are often used in speech recognition and sentiment analysis;

[0065] Intonation and speaking speed: reflect dynamic changes in emotional state.

[0066] In an embodiment of the present application, after obtaining the acoustic feature vector, the acoustic feature vector is input into a preset emotion evaluation model. Exemplarily, the preset emotion evaluation model can be obtained by training in advance using a convolutional neural network model.

[0067] In another embodiment, if the interaction data is received by text input, the interaction data can be directly input into a sentiment analysis tool to obtain the user's current sentiment tendency. Exemplarily, the sentiment analysis tool can be SentiWordNet.

[0068] Further, in the embodiment of the present application, after obtaining the current emotional tendency of the user, the historical interaction data corresponding to the target historical emotional tendency consistent with the current emotional tendency is determined in the historical emotional tendency. For example, when the current emotional tendency of the user is determined to be "pleasure", the historical emotional tendency consistent with the current emotional tendency "pleasure" is determined in the historical emotional tendency as the target historical emotional tendency, and then the historical interaction data corresponding to the target historical emotional tendency is determined.

[0069] Further, in an embodiment of the present application, if there are multiple historical interaction data corresponding to historical emotional tendencies that are consistent with the current emotional tendencies, historical evaluation data of historical reply data corresponding to the historical interaction data are obtained, and the historical interaction data corresponding to the historical evaluation data with the highest evaluation score is determined to be the target historical interaction data, and historical feature information corresponding to the target historical interaction data is obtained, and feature information corresponding to the historical feature information is determined to be the target feature information, wherein the historical reply data is reply data for the interaction data that the user has input; and the historical evaluation data is evaluation data obtained by evaluating the historical reply data.

[0070] Exemplarily, in an embodiment of the present application, it is determined that the historical interaction data corresponding to the target historical emotional tendency include: "Play light music", "Why is the sea blue", "Tell me a story", and the historical evaluation data of the corresponding historical reply data are: 80, 60, 95, respectively, then the historical interaction data "Tell me a story" is determined to be the target historical interaction data, and the historical feature information corresponding to the target historical interaction data is further obtained, which may include: name information, hobby information, and education information, then the name information, hobby information, and education information are determined to be the target feature information.

[0071] S2034, input the target feature information into a preset sample template to generate a corresponding feature description text.

[0072] In one embodiment, the sample template may be a template that combines target feature information. After the target feature information is determined, the target feature information is input into a preset sample template, and a feature description text corresponding to the target feature information may be generated.

[0073] Optionally, in the embodiment of the present application, the sample template may change according to the target feature information. For example, the address information may be "Xiao A" and the hobby information may be "reading martial arts novels". Then the target feature information is input into the preset sample template, and the generated feature description text may be "Xiao A likes to read martial arts novels"; if the target feature information is address information (Xiao A), age information (15), and hobby information (reading martial arts novels), the generated feature description text may be "Xiao A, 15 years old, likes to read martial arts novels". Then the generated feature description text is input into the interaction model to generate reply data corresponding to the interaction data.

[0074] It can be understood that when the feature description text is input into the interaction model and the interaction model inputs the reply data corresponding to the interaction data, the generated reply data is consistent with the feature description text under the constraints of the feature description text. The feature description text is generated by the identity information of the user who inputs the interaction data. Therefore, the reply data output by the interaction model is not only related to the interaction data but also to the user.

[0075] In an embodiment of the present application, the target feature information is determined in the feature information through the feature information corresponding to the user identity information, and the feature description text corresponding to the target feature information is determined, so that the feature description text is associated with the user's feature information; and the target feature information is determined according to the judgment result of whether it is the first time to receive the interaction data input by the user. When it is determined that the user is not inputting the interaction data for the first time, the target feature information is determined in the feature information through the user's current emotional tendency, and the target feature information that meets the current emotional tendency can be screened, so that the reply data is more in line with the user's needs; and by inputting the acoustic vector in the interaction data into the emotional evaluation model, the user's current emotional tendency is obtained, and the user's current emotional tendency is accurately determined; and when it is determined in the historical emotional tendency that there are multiple historical interaction data corresponding to the target historical emotional tendency consistent with the current emotional tendency, the historical interaction data corresponding to the historical evaluation data with the highest evaluation score is determined as the target historical interaction data, and the historical feature information corresponding to the target historical interaction data is determined as the target feature information. Since the determined target feature information is determined based on the historical reply data with the highest evaluation score, the target feature information is determined based on the actual situation, and the target feature information can be accurately determined, providing a basis for the subsequent generation of reply data that meets the user's personalized needs.

[0076] See also Figure 4 , which is a flow chart of a data generation method provided in an embodiment of the present application. Figure 4 As shown, the method of the embodiment of the present application may include the following steps S301-S306.

[0077] S301, receiving interaction data input by a user, and determining user identity information of the user based on the interaction data;

[0078] S302, generating a feature description text corresponding to the user identity information;

[0079] S303, determining text information corresponding to the interaction data, and inputting the text information and the feature description text into the interaction model to generate reply data corresponding to the interaction data;

[0080] Specifically, for obtaining reply data of interactive data in the embodiment of the present application, please refer to the description of the above embodiment, which will not be repeated here.

[0081] S304, receiving feedback data from the user, and calculating evaluation data based on the feedback data, wherein the feedback data is obtained based on the user's rating of the reply data;

[0082] In one embodiment, after the reply data is fed back to the user, the user's feedback data for the reply data is obtained, and then the feedback data is scored to obtain evaluation data.

[0083] Exemplarily, in the embodiment of the present application, if the feedback data received from the user regarding the reply data is very satisfactory, then the evaluation data of the feedback data is determined to be 95.

[0084] S305, obtaining adjacent interaction data of the interaction data, where the adjacent interaction data is the interaction data adjacent to the interaction data sent by the user;

[0085] S306, determining the change in the user's emotions between the interaction data and the adjacent interaction data, and determining evaluation data of the reply data based on the change in emotions.

[0086] In another embodiment, the electronic device may also determine the evaluation data according to the emotion change between the interaction data and the adjacent interaction data, wherein the adjacent interaction data is the next interaction data of the interaction data sent by the user.

[0087] For example, if through analysis of the next interaction data, it is found that the historical emotional tendency corresponding to the next interaction data is pleasure, and based on the interaction data it is determined that the user's current emotional tendency is stable, then it can be determined that the user's emotional change from the current interaction data to the next interaction data is from stable to pleasure, and the evaluation data of the reply data is determined to be 90.

[0088] It can be understood that in order to improve the user experience in this application, when determining the evaluation data of the reply data of the interaction data using the emotional changes determined by the interaction data input by the user, relatively high evaluation data can be set for the emotional change of the user from negative emotion to positive emotion.

[0089] Further, in the embodiment of the present application, if it is determined that there is no user identity information corresponding to the interaction data when the interaction data is received, the user identity information corresponding to the user is created, and the user is prompted to enter the characteristic information, wherein the interface for prompting the user to enter the corresponding characteristic information can refer to Figure 5 .exist Figure 5 In the example, the characteristic information input by the user includes name information, age information, hobby information, and education information. Thus, the characteristic information of the user can be collected.

[0090] Furthermore, in the embodiment of the present application, the feature information can also be updated according to the interaction data input by the user. Figure 6 If the interaction data input by the user is "I like to watch skits", "watch skits" may be added to the hobby information corresponding to the user identity information of the user.

[0091] In an embodiment of the present application, evaluation data of reply data corresponding to interaction data is determined by determining the emotion change of the user between the input interaction data and the adjacent interaction data of the interaction data, without the need for the user to make subjective input judgments, thereby improving the accuracy of the determined evaluation data; and when it is determined that there is no user identity information corresponding to the interaction data, the user identity information corresponding to the user can be created, and the user can be prompted to input feature information, and the feature information input by the user can be obtained, thereby providing a basis for the interaction model to output the interaction data corresponding to the feature information input by the user, and the feature information can be updated through the interaction data, thereby the user's feature information can be updated during the interaction process, thereby improving the convenience of updating the user's feature information.

[0092] based on Figure 1 The following is a schematic diagram of the scene. Figure 7-Figure 8 , the data generation device provided in the embodiment of the present application is introduced in detail. It should be noted that, Figure 7-Figure 8 The data generating device in the present application is used to execute Figure 2-Figure 6 For the convenience of explanation, only the part related to the embodiment of the present application is shown. For the specific technical details not disclosed, please refer to the present application. Figure 2-Figure 6 Specifically, the data generating device 1 includes: a receiving unit 11 , a generating unit 12 , and a determining unit 13 .

[0093] The receiving unit 11 is used to receive interaction data input by the user and determine the user identity information of the user based on the interaction data;

[0094] A generating unit 12, used to generate a feature description text corresponding to the user identity information;

[0095] The determination unit 13 is used to determine text information corresponding to the interaction data, and input the text information and the feature description text into the interaction model to generate reply data corresponding to the interaction data.

[0096] Optionally, the generating unit 12 further includes: an acquiring subunit 121 and a screening subunit 122 .

[0097] The acquisition subunit 121 is used to acquire characteristic information corresponding to the user identity information, the characteristic information including the user's name information, age information, personality information, hobby information, and education information;

[0098] The screening subunit 122 is used to screen out target feature information from the feature information and determine feature description text corresponding to the target feature information.

[0099] Optionally, the screening subunit 122 is specifically used for:

[0100] Determining whether the interaction data is interaction data input by the user received for the first time;

[0101] If so, the name information and hobby information are determined as target feature information;

[0102] If not, determining the user's current emotional tendency based on the interaction data, and determining target feature information in the feature information based on the current emotional tendency;

[0103] Input the target feature information into the pre-set sample template to generate the corresponding feature description text.

[0104] Optionally, the screening subunit 122 is specifically used for:

[0105] Extracting acoustic feature vectors from the interaction data, and inputting the acoustic feature vectors into a preset emotion evaluation model;

[0106] The user's current emotional tendency is determined based on the output results of the emotional evaluation model.

[0107] Optionally, the screening subunit 122 is specifically used for:

[0108] Determine a target historical emotional tendency that is consistent with the current emotional tendency in the historical emotional tendency, and determine the historical interaction data corresponding to the target historical emotional tendency;

[0109] If there are multiple historical interaction data, then obtain the historical evaluation data of the historical reply data corresponding to the historical interaction data;

[0110] Determine the historical interaction data corresponding to the historical evaluation data with the highest evaluation score as the target historical interaction data;

[0111] The historical feature information corresponding to the target historical interaction data is obtained, and the feature information corresponding to the historical feature information is determined to be the target feature information.

[0112] Optionally, refer to Figure 8 The data generating device also includes: a calculating unit 14, an acquiring unit 15, and an evaluating and determining unit 16.

[0113] A calculation unit 14 is used to receive feedback data from users and calculate evaluation data based on the feedback data, where the feedback data is obtained based on the user's rating of the reply data; and / or,

[0114] An acquisition unit 15 is used to acquire adjacent interaction data of the interaction data, where the adjacent interaction data is the adjacent interaction data of the interaction data sent by the user;

[0115] The evaluation determination unit 16 is used to determine the change in the user's emotions between the interaction data and the adjacent interaction data, and determine the evaluation data of the reply data based on the change in emotions.

[0116] Optionally, the receiving unit 11 is further configured to:

[0117] If the user identity information corresponding to the interaction data does not exist, the user identity information corresponding to the user is created, and the user is prompted to input feature information.

[0118] Optionally, the determining unit 13 is further configured to:

[0119] Update feature information based on interaction data.

[0120] In the embodiment of the present application, after obtaining the interaction data input by the user, the user identity information of the user is determined based on the interaction data, and then the feature description text corresponding to the user identity information is obtained, and the feature description text and the interaction data are input into the interaction model, and the interaction model generates corresponding reply data for the input feature description text and interaction data. Since the interaction model also uses the feature description text corresponding to the user identity information when generating the reply data corresponding to the interaction data, the reply data is not only related to the interaction data, but also to the user's feature description text, which improves the correlation between the reply data and the user's feature description text and meets the user's personalized needs.

[0121] See also Figure 8 , is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 8 As shown, the electronic device 500 includes a processor 501 and a memory 502. The processor 501 is electrically connected to the memory 502.

[0122] The processor 501 is the control center of the electronic device 500 and may include one or more processing cores. The processor 501 uses various interfaces and lines to connect various parts of the entire electronic device 500, and executes various functions and processes data of the electronic device 500 by running or calling computer programs stored in the memory 502, and calling data stored in the memory 502, so as to control the electronic device 500 as a whole. Optionally, the processor 501 can be implemented in at least one hardware form of digital signal processing (DSP), field programmable gate array (FPGA), and programmable logic array (PLA). The processor 501 can integrate one or a combination of CPU, graphics processing unit (GPU), modem, etc. Among them, the CPU mainly processes the operating system, user pages, and applications; the GPU is responsible for rendering and drawing display content; and the modem is used to process wireless communications. It is understandable that the above-mentioned modem may not be integrated into the processor 501, but may be implemented separately through a communication chip.

[0123] The memory 502 can be used to store software programs and modules, and the processor 501 executes various functional applications and data processing by running the computer programs and modules stored in the memory 502. The memory 502 can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, a computer program required for at least one function, etc.; the data storage area can store data created according to the use of the electronic device 500, etc.

[0124] In addition, the memory 502 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage device. Accordingly, the memory 502 may also include a memory controller to provide the processor 501 with access to the memory 502.

[0125] In this embodiment, the processor 501 in the electronic device 500 loads instructions corresponding to the processes of one or more computer programs into the memory 502 according to the following steps, and the processor 501 runs the computer program stored in the memory 502 to implement various functions, as follows:

[0126] receiving interaction data input by a user, and determining user identity information of the user based on the interaction data;

[0127] Generate a feature description text corresponding to the user identity information;

[0128] Text information corresponding to the interaction data is determined, and the text information and the feature description text are input into the interaction model to generate response data corresponding to the interaction data.

[0129] Optionally, when the processor 501 generates a feature description text corresponding to the user identity information, the processor 501 specifically performs:

[0130] Acquire characteristic information corresponding to the user's identity information, the characteristic information including the user's name information, age information, personality information, hobby information, and education information;

[0131] Target feature information is screened out from the feature information, and feature description text corresponding to the target feature information is determined.

[0132] Optionally, when the processor 501 filters out the target feature information from the feature information and determines the feature description text corresponding to the target feature information, the processor 501 specifically executes:

[0133] Determining whether the interaction data is interaction data input by the user received for the first time;

[0134] If so, the name information and hobby information are determined as target feature information;

[0135] If not, determining the user's current emotional tendency based on the interaction data, and determining target feature information in the feature information based on the current emotional tendency;

[0136] Input the target feature information into the pre-set sample template to generate the corresponding feature description text.

[0137] Optionally, when determining the current emotional tendency of the user based on the interaction data, the processor 501 specifically performs:

[0138] Extracting acoustic feature vectors from the interaction data, and inputting the acoustic feature vectors into a preset emotion evaluation model;

[0139] The user's current emotional tendency is determined based on the output results of the emotional evaluation model.

[0140] Optionally, when the processor 501 determines the target feature information in the feature information based on the current emotional tendency, the processor 501 specifically performs:

[0141] Determine a target historical emotional tendency that is consistent with the current emotional tendency in the historical emotional tendency, and determine the historical interaction data corresponding to the target historical emotional tendency;

[0142] If there are multiple historical interaction data, then obtain the historical evaluation data of the historical reply data corresponding to the historical interaction data;

[0143] Determine the historical interaction data corresponding to the historical evaluation data with the highest evaluation score as the target historical interaction data;

[0144] The historical feature information corresponding to the target historical interaction data is obtained, and the feature information corresponding to the historical feature information is determined to be the target feature information.

[0145] Optionally, after inputting the text information and the feature description text into the interaction model to generate reply data corresponding to the interaction data, the processor 501 may further execute:

[0146] Receiving feedback data from users, and calculating evaluation data based on the feedback data, wherein the feedback data is obtained based on the user's rating of the reply data; and / or,

[0147] Acquire adjacent interaction data of the interaction data, where the adjacent interaction data is the adjacent interaction data of the interaction data sent by the user;

[0148] The change of the user's emotion between the interaction data and the adjacent interaction data is determined, and the evaluation data of the reply data is determined based on the change of emotion.

[0149] Optionally, after receiving the interactive data input by the user, the processor 501 may further execute:

[0150] If the user identity information corresponding to the interaction data does not exist, the user identity information corresponding to the user is created, and the user is prompted to input feature information.

[0151] Optionally, after determining the text information corresponding to the interaction data and inputting the text information and the feature description text into the interaction model, the processor 501 may further execute:

[0152] Update feature information based on interaction data.

[0153] In the embodiment of the present application, after obtaining the interaction data input by the user, the user identity information of the user is determined based on the interaction data, and then the feature description text corresponding to the user identity information is obtained, and the feature description text and the interaction data are input into the interaction model, and the interaction model generates corresponding reply data for the input feature description text and interaction data. Since the interaction model also uses the feature description text corresponding to the user identity information when generating the reply data corresponding to the interaction data, the reply data is not only related to the interaction data, but also to the user's feature description text, which improves the correlation between the reply data and the user's feature description text and meets the user's personalized needs.

[0154] It should be understood that the device provided in the embodiment of the present application is used to execute the above-mentioned data generation method, and thus can achieve the same effect as the above-mentioned implementation method.

[0155] In the case of an integrated unit, the device may include a processing module and a storage module. When the device is applied to an electronic device, the processing module may be used to control and manage the actions of the electronic device. The storage module may be used to support the electronic device in executing related program codes, etc.

[0156] The processing module may be a processor or a controller, which may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of this application. The processor may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of digital signal processing (DSP) and a microprocessor, etc. The storage module may be a memory.

[0157] In addition, the device provided in the embodiment of the present application can specifically be a chip, component or module, and the chip may include a connected processor and memory; wherein the memory is used to store instructions, and when the processor calls and executes the instructions, the chip can execute a data generation method provided in the above embodiment.

[0158] An embodiment of the present application also provides a computer-readable storage medium, in which a computer program code is stored. When the computer program code is executed on a computer, the computer executes the above-mentioned related method steps to implement a data generation method provided in the above-mentioned embodiment.

[0159] This embodiment also provides a computer program product. When the computer program product runs on a computer, it enables the computer to execute the above-mentioned related steps to implement a data generation method provided by the above embodiment.

[0160] Among them, the device, computer-readable storage medium, computer program product or chip provided in this embodiment is used to execute the corresponding method provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding method provided above, and will not be repeated here.

[0161] Through the description of the above implementation methods, technical personnel in the relevant field can understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0162] In the embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of modules or units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0163] The above contents are only specific implementation methods of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A data generation method, characterized in that: The method comprises: receiving interaction data input by a user, and determining user identity information of the user based on the interaction data; Generate a feature description text corresponding to the user identity information; Determine text information corresponding to the interaction data, and input the text information and the feature description text into an interaction model to generate reply data corresponding to the interaction data.

2. The method according to claim 1, characterized in that The generating of a feature description text corresponding to the user identity information includes: Acquire characteristic information corresponding to the user identity information, the characteristic information including the user's name information, age information, personality information, hobby information, and education information; Target feature information is screened out from the feature information, and feature description text corresponding to the target feature information is determined.

3. The method according to claim 2, characterized in that The step of filtering out target feature information from the feature information and determining a feature description text corresponding to the target feature information includes: Determining whether the interaction data is interaction data input by the user received for the first time; If yes, determining the name information and the hobby information as the target characteristic information; If not, determining the current emotional tendency of the user based on the interaction data, and determining the target feature information in the feature information based on the current emotional tendency; The target feature information is input into a preset sample template to generate a corresponding feature description text.

4. The method according to claim 3, characterized in that Determining the current emotional tendency of the user based on the interaction data includes: Extracting acoustic feature vectors from the interaction data, and inputting the acoustic feature vectors into a preset emotion evaluation model; The current emotional tendency of the user is determined based on the output result of the emotional evaluation model.

5. The method according to claim 3, characterized in that: The determining the target feature information in the feature information based on the current emotional tendency includes: Determine a target historical emotional tendency consistent with the current emotional tendency in the historical emotional tendency, and determine the historical interaction data corresponding to the target historical emotional tendency; If there are multiple historical interaction data, obtaining historical evaluation data of historical reply data corresponding to the historical interaction data; Determine the historical interaction data corresponding to the historical evaluation data with the highest evaluation score as the target historical interaction data; Acquire historical feature information corresponding to the target historical interaction data, and determine that feature information corresponding to the historical feature information is the target feature information.

6. The method according to claim 1, characterized in that After inputting the text information and the feature description text into the interaction model to generate reply data corresponding to the interaction data, the method further includes: receiving feedback data from a user, and calculating evaluation data of the reply data based on the feedback data, wherein the feedback data is obtained based on the user's rating of the reply data; and / or, Acquire adjacent interaction data of the interaction data, where the adjacent interaction data is interaction data adjacent to the interaction data sent by the user; The change of the user's emotion between the interaction data and adjacent interaction data is determined, and evaluation data of the reply data is determined based on the change of emotion.

7. The method according to claim 1, characterized in that After receiving the interactive data input by the user, the method further includes: If the user identity information corresponding to the interaction data does not exist, the user identity information corresponding to the user is created, and the user is prompted to input feature information.

8. A data generating device, characterized in that: The device comprises: A receiving unit, configured to receive interaction data input by a user, and determine user identity information of the user based on the interaction data; A generating unit, used to generate a feature description text corresponding to the user identity information; A determination unit is used to determine text information corresponding to the interaction data, and input the text information and the feature description text into an interaction model to generate reply data corresponding to the interaction data.

9. An electronic device, characterized in that: The electronic device comprises: A memory for storing executable program codes; A processor, configured to call and run the executable program code from the memory, so that the electronic device executes the method according to any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program code, and when the computer program code is executed, the method according to any one of claims 1 to 7 is implemented.

Images