Data processing method, device and equipment

Abstract

The present invention provides a data processing method, device and equipment to solve the problem that the existing method of obtaining user satisfaction scores through telephone surveys is easily affected by human factors, making the scores less objective. The method of the present invention includes: obtaining a first voice segment of an interviewee recorded in the satisfaction survey of a target issue and an initial satisfaction score given by the interviewee; performing emotion recognition on the first voice segment to obtain an emotional index of the interviewee, and the emotional index is used to reflect the emotional state of the interviewee; based on the emotional index, correcting the initial satisfaction score to obtain a final satisfaction score. The present invention can eliminate the influence of the interviewee's emotions on the satisfaction score, make the score more objective, and improve the accuracy and reliability of satisfaction survey data.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a data processing method, device and equipment. Background Art

[0002] In today's increasingly competitive market, customer satisfaction has gradually become a core corporate strategy, goal, and competitive tool. Customer orientation, striving to meet customer needs, pursuing customer satisfaction, and hoping to gain customer loyalty have become key priorities for businesses. In line with this trend, customer satisfaction measurement has gradually become a specialized practice. In the carrier sector, third-party telephone surveys are commonly used to gauge customer satisfaction. This data serves not only as a guide for operator improvement but also as an external indicator of operator performance. Therefore, obtaining accurate and reliable satisfaction data is extremely important.

[0003] In existing technical solutions, a user's rating of an operator's services depends on a single call with a third-party staff member. The rating given by the user in a single call may be subject to human influence, making the rating lose objectivity. Summary of the Invention

[0004] The purpose of the present invention is to provide a data processing method, device and equipment for solving the problem that the existing method of obtaining user satisfaction scores through telephone surveys is easily affected by human factors, making the scores lose objectivity.

[0005] In order to achieve the above object, the present invention provides a data processing method, comprising:

[0006] Obtaining a first voice clip of a respondent collected during a satisfaction survey on a target issue and an initial satisfaction score given by the respondent;

[0007] Performing emotion recognition on the first voice segment to obtain an emotion index of the interviewee, where the emotion index is used to reflect the emotional state of the interviewee;

[0008] Based on the sentiment index, the initial satisfaction score is revised to obtain a final satisfaction score.

[0009] The first voice clip of the interviewee recorded in the satisfaction survey on the target question is obtained, including:

[0010] Obtaining a voice call recording of the respondent, wherein the voice call recording includes N questions that the respondent is required to rate during a satisfaction survey, where N is a positive integer and N is greater than or equal to 1.

[0011] Based on the N questions, the voice call recording is segmented to obtain a first voice segment of the respondent recorded during the satisfaction survey on the target question, where the target question is at least one of the N questions.

[0012] The performing emotion recognition on the first voice segment to obtain the respondent's emotion index includes:

[0013] The first speech segment is passed through a speech emotion recognition model to calculate the emotion index of the respondent.

[0014] The step of revising the initial satisfaction score based on the sentiment index to obtain a final satisfaction score includes:

[0015] Obtaining attribute parameters of the interviewee;

[0016] The emotion index and the attribute parameters of the respondent are used as feature vectors, and a score to be corrected is calculated using a pre-established correction model;

[0017] The initial satisfaction score and the score to be revised are summed to obtain a final satisfaction score.

[0018] The method further comprises:

[0019] By collecting voice data, a voice data sample library is constructed, wherein the voice data sample library stores voice data of different interviewees, each voice data including a second voice segment and a score of a question corresponding to the second voice segment;

[0020] Extract features from each voice data and calculate the corresponding respondent's emotion index;

[0021] Obtaining a scoring model by machine learning based on the attribute parameters of the respondent corresponding to the first voice data, wherein the first voice data is voice data corresponding to a sentiment index falling within a first preset range;

[0022] The attribute parameters and emotional index of the interviewee corresponding to the second voice data are used to obtain a correction model through machine learning. The second voice data is the voice data corresponding to the emotional index falling outside the first preset range, wherein the score corresponding to the second voice data is matched with the scoring model through the correction model.

[0023] The attribute parameters of the interviewee include the interviewee's own attributes and the attributes of the service used by the interviewee.

[0024] The present invention also provides a data processing device, comprising:

[0025] An acquisition module, configured to acquire a first voice clip of a respondent recorded in a satisfaction survey on a target question and an initial satisfaction score given by the respondent;

[0026] an emotion recognition module, configured to perform emotion recognition on the first voice segment to obtain an emotion index of the respondent, wherein the emotion index is used to reflect the emotional state of the respondent;

[0027] The correction module is used to correct the initial satisfaction score based on the emotion index to obtain a final satisfaction score.

[0028] The present invention further provides a data processing device, comprising a processor and a transceiver, wherein the transceiver receives and sends data under the control of the processor, and the processor is configured to perform the following operations:

[0029] Obtaining a first voice clip of a respondent collected during a satisfaction survey on a target issue and an initial satisfaction score given by the respondent;

[0030] Performing emotion recognition on the first voice segment to obtain an emotion index of the interviewee, where the emotion index is used to reflect the emotional state of the interviewee;

[0031] Based on the sentiment index, the initial satisfaction score is revised to obtain a final satisfaction score.

[0032] The processor is further configured to execute the following process:

[0033] Obtaining a voice call recording of the respondent, wherein the voice call recording includes N questions that the respondent is required to rate during a satisfaction survey, where N is a positive integer and N is greater than or equal to 1.

[0034] Based on the N questions, the voice call recording is segmented to obtain a first voice segment of the respondent recorded during the satisfaction survey on the target question, where the target question is at least one of the N questions.

[0035] The processor is further configured to execute the following process:

[0036] The first speech segment is passed through a speech emotion recognition model to calculate the emotion index of the respondent.

[0037] The processor is further configured to execute the following process:

[0038] Obtaining attribute parameters of the interviewee;

[0039] The emotion index and the attribute parameters of the respondent are used as feature vectors, and a score to be corrected is calculated using a pre-established correction model;

[0040] The initial satisfaction score and the score to be revised are summed to obtain a final satisfaction score.

[0041] The processor is further configured to execute the following process:

[0042] By collecting voice data, a voice data sample library is constructed, wherein the voice data sample library stores voice data of different interviewees, each voice data including a second voice segment and a score of a question corresponding to the second voice segment;

[0043] Extract features from each voice data and calculate the corresponding respondent's emotion index;

[0044] Obtaining a scoring model by machine learning based on the attribute parameters of the respondent corresponding to the first voice data, wherein the first voice data is voice data corresponding to a sentiment index falling within a first preset range;

[0045] The attribute parameters and emotional index of the interviewee corresponding to the second voice data are used to obtain a correction model through machine learning. The second voice data is the voice data corresponding to the emotional index falling outside the first preset range, wherein the score corresponding to the second voice data is matched with the scoring model through the correction model.

[0046] The attribute parameters of the interviewee include the interviewee's own attributes and the attributes of the service used by the interviewee.

[0047] The present invention also provides a data processing device, comprising a memory, a processor, and a program stored in the memory and executable on the processor; when the processor executes the program, the data processing method described above is implemented.

[0048] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which implements the steps in the data processing method described above when the program is executed by a processor.

[0049] The above technical solution of the present invention has at least the following beneficial effects:

[0050] In an embodiment of the present invention, a first voice clip of an interviewee recorded in a satisfaction survey on a target issue and an initial satisfaction score given by the interviewee are obtained; emotion recognition is performed on the first voice clip to obtain an emotion index of the interviewee, and the emotion index is used to reflect the emotional state of the interviewee; based on the emotion index, the initial satisfaction score is corrected to obtain a final satisfaction score. In this way, the influence of the interviewee's emotion on the satisfaction score can be eliminated, the score can be made more objective, and the accuracy and reliability of the satisfaction survey data can be improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 A schematic diagram showing a flow chart of a data processing method according to an embodiment of the present invention;

[0052] Figure 2 A schematic diagram showing a module of a data processing device according to an embodiment of the present invention;

[0053] Figure 3 A schematic structural diagram of a data processing device according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0054] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, a detailed description will be given below with reference to the accompanying drawings and specific embodiments.

[0055] The present invention aims to solve the problem that user satisfaction scores obtained by telephone surveys in the prior art are easily affected by human factors, resulting in a loss of objectivity. The present invention provides a data processing method, device and equipment.

[0056] Here, the main steps of implementing the existing third-party telephone survey method are as follows:

[0057] 1. Set the number of surveys for each operator in each province, and randomly select users in each province for telephone surveys.

[0058] 2. Inquire about the satisfaction of connected users with voice and Internet access (mobile phone and broadband) services and obtain user ratings.

[0059] 3. Analyze user ratings by operator, province, business, and other dimensions to derive user satisfaction indicators for each operator.

[0060] In existing technical solutions, users' ratings of operator services depend on a single call with a third-party staff member. The inventors have found that the ratings given by users in a single call are easily influenced by human factors, including the user's emotions at the time of the call, making the ratings less objective.

[0061] In order to solve the above problems, embodiments of the present invention provide a data processing method, apparatus, and device.

[0062] like Figure 1 FIG. 1 is a flow chart of a data processing method according to an embodiment of the present invention. The method specifically includes:

[0063] Step 101: obtaining a first voice clip of a respondent collected during a satisfaction survey on a target issue and an initial satisfaction score given by the respondent;

[0064] It should be noted that, optionally, the satisfaction score is normalized into a unified score range, for example, 0 to 10 points.

[0065] Step 102: performing emotion recognition on the first voice segment to obtain an emotion index of the respondent, where the emotion index is used to reflect the respondent's emotional state;

[0066] In this step, the emotional state includes a positive emotional state, a negative emotional state, and a neutral emotional state.

[0067] It should be noted that the emotion recognition method can be achieved by extracting the acoustic feature parameters of the interviewee's voice and establishing a corresponding model for discrimination; or, it can be achieved by pre-establishing an emotion database that contains the correspondence between the emotional state and the feature parameters of the interviewee's voice, thereby obtaining the emotional state of the interviewee through the correspondence.

[0068] Here, without loss of generality, the sentiment index is represented as a value within a preset value range, for example, the sentiment index is represented as a value between [-5, 5].

[0069] Step 103: Based on the emotion index, the initial satisfaction score is revised to obtain a final satisfaction score.

[0070] In this step, the initial satisfaction score is revised based on the emotional index in order to eliminate the influence of emotional factors on the satisfaction score.

[0071] The data processing method of the embodiment of the present invention obtains the first voice clip of the interviewee recorded in the satisfaction survey of the target problem and the initial satisfaction score given by the interviewee; performs emotion recognition on the first voice clip to obtain the interviewee's emotion index, and the emotion index is used to reflect the emotional state of the interviewee; based on the emotion index, the initial satisfaction score is corrected to obtain the final satisfaction score. In this way, the influence of the interviewee's emotion on the satisfaction score can be eliminated, the score is made more objective, and the accuracy and reliability of the satisfaction survey data are improved.

[0072] As an optional implementation, in step 101 of the method of the embodiment of the present invention, obtaining the first voice segment of the interviewee recorded during the satisfaction survey on the target question may include:

[0073] Obtaining a voice call recording of the respondent, wherein the voice call recording includes N questions that the respondent is required to rate during a satisfaction survey, where N is a positive integer and N is greater than or equal to 1.

[0074] In this step, the satisfaction survey specifically refers to a satisfaction survey conducted by telephone. Here, the respondent's voice call recording specifically refers to a complete voice call recording from the time the respondent answers the call to the time the call is hung up in a satisfaction survey.

[0075] Based on the N questions, the voice call recording is segmented to obtain a first voice segment of the respondent recorded during the satisfaction survey on the target question, where the target question is at least one of the N questions.

[0076] It should be noted that the target question can be one or more. In other words, each of the N questions can correspond to one satisfaction score, or multiple questions can correspond to one satisfaction score.

[0077] Here, when multiple (greater than or equal to 2) questions correspond to one satisfaction score, the multiple questions may be a group of questions for the same survey level or the same business type.

[0078] It should be noted that the first voice segment is one or more voice segments for the target question, which is not specifically limited here.

[0079] Here, based on the questions involved in the satisfaction survey, the interviewees' voice call recordings are segmented and refined into different voice segments, and then subsequent emotion recognition is performed. The satisfaction score is corrected based on the emotion index obtained based on emotion recognition, which can further improve the accuracy and reliability of the satisfaction survey data.

[0080] Specifically, in step 101 of the method of the embodiment of the present invention, obtaining the initial satisfaction score given by the respondent during the satisfaction survey on the target issue includes:

[0081] The initial satisfaction scores given by the respondents during the satisfaction survey on the target issue are obtained through the satisfaction survey score list.

[0082] That is, the initial satisfaction scores given by the respondents in the satisfaction survey process for different questions can be pre-stored in a list format.

[0083] As an optional implementation, step 102 of the method in this embodiment of the present invention may include:

[0084] The first speech segment is passed through a speech emotion recognition model to calculate the emotion index of the respondent.

[0085] Here, the speech emotion recognition model is a pre-established and trained model. Specifically, the model can be established by extracting acoustic feature parameters of the user's speech, and then trained using a large amount of training data and a preset algorithm to obtain the speech emotion recognition model.

[0086] Of course, the method is not limited to the above. A neural network can also be built to obtain the correspondence between the emotion index and the characteristic parameters of the interviewee's voice, and then the emotion index of the interviewee can be obtained through the correspondence.

[0087] As an optional implementation, step 103 of the method in this embodiment of the present invention may include:

[0088] Obtaining attribute parameters of the interviewee;

[0089] In this step, the attribute parameters of the interviewee include the interviewee's own attributes and the attributes of the service used by the interviewee.

[0090] It should be noted that the attributes of the interviewee include but are not limited to: the interviewee's gender, age, education level, and location information.

[0091] The attributes of the services used by the interviewee include but are not limited to: basic services and value-added services activated by the interviewee.

[0092] The emotion index and the attribute parameters of the respondent are used as feature vectors, and a score to be corrected is calculated using a pre-established correction model;

[0093] Here, as an optional implementation, the correction model can be established by the following method, as follows:

[0094] By collecting voice data, a voice data sample library is constructed, wherein the voice data sample library stores voice data of different interviewees, each voice data including a second voice segment and a score of a question corresponding to the second voice segment;

[0095] It's important to note that to obtain a sufficiently comprehensive sample, the constructed voice data sample library needs to cover as many interviewee groups as possible. The interviewee attributes can be more diverse, for example, including interviewees of different genders, age groups, regions, and educational backgrounds. Furthermore, the voice data should reflect as many emotions as possible. Specifically, for each interviewee group, a variety of emotional expressions, ranging from negative to neutral to positive, should be included.

[0096] It should be noted that the voice data sample library can be constructed through active access or through past historical evaluation recordings.

[0097] Extract features from each voice data and calculate the corresponding respondent's emotion index;

[0098] In this step, specifically, features can be extracted based on frequency, duration, resonance peaks, MFCC (Mel-Fregunency Ceptral Coefficients), etc., but are not limited to these. Afterwards, the emotion index in the speech is identified through an SVM (Support Vector Machine) algorithm (not limited to this type of algorithm).

[0099] Without loss of generality, the sentiment index is represented as a value within a preset range, for example, the sentiment index is represented as a value between [-5, 5].

[0100] Obtaining a scoring model by machine learning based on the attribute parameters of the respondent corresponding to the first voice data, wherein the first voice data is voice data corresponding to a sentiment index falling within a first preset range;

[0101] Optionally, the attribute parameters of the interviewee include the interviewee's own attributes and attributes of the service used by the interviewee.

[0102] It should be noted that the emotional state corresponding to the emotional index within the first preset range is a neutral emotional state. For example, the first preset range is [-0.1, 0.1].

[0103] Here, the neutral emotional state sample score is used as the standard, and the scoring model is obtained through machine learning. Specifically, the first voice data and the score corresponding to the first voice data are used as the standard, and the scoring model is constructed by machine learning methods to convert the sample features into scores.

[0104] The attribute parameters and emotional index of the interviewee corresponding to the second voice data are used to obtain a correction model through machine learning. The second voice data is the voice data corresponding to the emotional index falling outside the first preset range, wherein the score corresponding to the second voice data is matched with the scoring model through the correction model.

[0105] It should be noted that for voice data with an emotion index outside the first preset range, the corresponding emotional state of the respondent is either negative or positive. In either a negative or positive emotional state, the respondent's score for the question is influenced by their current mood, so the model needs to be modified to eliminate the influence of emotional factors on the score.

[0106] Here, the correction model can make the scores of all samples (second voice data) whose emotion index falls outside the first preset range plus the correction model match the model of samples whose emotion index falls within the first preset range (i.e., the scoring model) as much as possible. That is to say, after the correction model is used, the final satisfaction score obtained is an objective score that excludes the influence of emotions.

[0107] The initial satisfaction score and the score to be revised are summed to obtain a final satisfaction score.

[0108] Here, the initial satisfaction score is the score given by the respondent voluntarily plus the score to be corrected to obtain the score after correcting the emotional factors, that is, the final satisfaction score.

[0109] The data processing method of the embodiment of the present invention obtains the first voice clip of the interviewee recorded in the satisfaction survey of the target problem and the initial satisfaction score given by the interviewee; performs emotion recognition on the first voice clip to obtain the interviewee's emotion index, and the emotion index is used to reflect the emotional state of the interviewee; based on the emotion index, the initial satisfaction score is corrected to obtain the final satisfaction score. In this way, the influence of the interviewee's emotion on the satisfaction score can be eliminated, the score is made more objective, and the accuracy and reliability of the satisfaction survey data are improved.

[0110] like Figure 2 As shown, an embodiment of the present invention further provides a data processing device, the device comprising:

[0111] An acquisition module 201 is configured to acquire a first voice segment of a respondent collected during a satisfaction survey on a target issue and an initial satisfaction score given by the respondent;

[0112] An emotion recognition module 202 is configured to perform emotion recognition on the first voice segment to obtain an emotion index of the respondent, where the emotion index is used to reflect the respondent's emotional state;

[0113] The correction module 203 is configured to correct the initial satisfaction score based on the emotion index to obtain a final satisfaction score.

[0114] Optionally, the acquisition module 201 may include:

[0115] A first acquiring unit is configured to acquire a voice call recording of the respondent, wherein the voice call recording includes N questions for which the respondent is required to give a score during a satisfaction survey, where N is greater than or equal to 1 and is a positive integer;

[0116] The first processing unit is used to perform voice segmentation on the voice call recording based on the N questions to obtain a first voice segment of the interviewee recorded in the satisfaction survey of the target question, where the target question is at least one of the N questions.

[0117] Optionally, the emotion recognition module 202 may include:

[0118] The emotion recognition unit is used to calculate the emotion index of the interviewee by applying the first speech segment to a speech emotion recognition model.

[0119] Optionally, the correction module 203 may include:

[0120] A second acquiring unit, configured to acquire attribute parameters of the respondent;

[0121] A second processing unit is configured to use the emotion index and the attribute parameters of the respondent as feature vectors and calculate a score to be corrected using a pre-established correction model;

[0122] The correction unit is used to perform a sum operation on the initial satisfaction score and the score to be corrected to obtain a final satisfaction score.

[0123] Optionally, the data processing device may further include:

[0124] A sample library construction module is used to construct a voice data sample library by collecting voice data, wherein the voice data sample library stores voice data of different interviewees, each voice data including a second voice segment and a score corresponding to a question of the second voice segment;

[0125] The calculation module is used to extract features from each voice data and calculate the corresponding emotional index of the interviewee;

[0126] a first processing module configured to obtain a scoring model by machine learning based on the attribute parameters of the respondent corresponding to the first voice data, wherein the first voice data is voice data corresponding to an emotion index falling within a first preset range;

[0127] The second processing module is used to obtain a correction model through machine learning for the attribute parameters and emotional index of the respondent corresponding to the second voice data, wherein the second voice data is voice data corresponding to the emotional index falling outside the first preset range, wherein the score corresponding to the second voice data is matched with the scoring model through the correction model.

[0128] Optionally, the attribute parameters of the interviewee include the interviewee's own attributes and attributes of the service used by the interviewee.

[0129] The data processing device of an embodiment of the present invention obtains a first voice clip of an interviewee recorded in a satisfaction survey on a target issue and an initial satisfaction score given by the interviewee; performs emotion recognition on the first voice clip to obtain an emotion index of the interviewee, and the emotion index is used to reflect the emotional state of the interviewee; based on the emotion index, the initial satisfaction score is corrected to obtain a final satisfaction score. In this way, the influence of the interviewee's emotion on the satisfaction score can be eliminated, the score can be made more objective, and the accuracy and reliability of the satisfaction survey data can be improved.

[0130] It should be noted here that the above-mentioned device provided by the embodiment of the present invention can implement all the method steps implemented by the above-mentioned method embodiment and can achieve the same technical effect. The parts and beneficial effects that are the same as the method embodiment in this embodiment will not be described in detail here.

[0131] In order to better achieve the above goals, Figure 3 As shown, an embodiment of the present invention further provides a data processing device, including a processor 300 and a transceiver 310. The transceiver 310 receives and sends data under the control of the processor. The processor 300 is configured to perform the following process:

[0132] Obtaining a first voice clip of a respondent collected during a satisfaction survey on a target issue and an initial satisfaction score given by the respondent;

[0133] Performing emotion recognition on the first voice segment to obtain an emotion index of the interviewee, where the emotion index is used to reflect the emotional state of the interviewee;

[0134] Based on the sentiment index, the initial satisfaction score is revised to obtain a final satisfaction score.

[0135] Optionally, the processor 300 is further configured to execute the following process:

[0136] Obtaining a voice call recording of the respondent, wherein the voice call recording includes N questions that the respondent is required to rate during a satisfaction survey, where N is a positive integer and N is greater than or equal to 1.

[0137] Based on the N questions, the voice call recording is segmented to obtain a first voice segment of the respondent recorded during the satisfaction survey on the target question, where the target question is at least one of the N questions.

[0138] Optionally, the processor 300 is further configured to execute the following process:

[0139] The first speech segment is passed through a speech emotion recognition model to calculate the emotion index of the respondent.

[0140] Optionally, the processor 300 is further configured to execute the following process:

[0141] Obtaining attribute parameters of the interviewee;

[0142] The emotion index and the attribute parameters of the respondent are used as feature vectors, and a score to be corrected is calculated using a pre-established correction model;

[0143] The initial satisfaction score and the score to be revised are summed to obtain a final satisfaction score.

[0144] Optionally, the processor 300 is further configured to execute the following process:

[0145] By collecting voice data, a voice data sample library is constructed, wherein the voice data sample library stores voice data of different interviewees, each voice data including a second voice segment and a score of a question corresponding to the second voice segment;

[0146] Extract features from each voice data and calculate the corresponding respondent's emotion index;

[0147] Obtaining a scoring model by machine learning based on the attribute parameters of the respondent corresponding to the first voice data, wherein the first voice data is voice data corresponding to a sentiment index falling within a first preset range;

[0148] The attribute parameters and emotional index of the interviewee corresponding to the second voice data are used to obtain a correction model through machine learning. The second voice data is the voice data corresponding to the emotional index falling outside the first preset range, wherein the score corresponding to the second voice data is matched with the scoring model through the correction model.

[0149] Optionally, the attribute parameters of the interviewee include the interviewee's own attributes and attributes of the service used by the interviewee.

[0150] The data processing device of an embodiment of the present invention obtains a first voice clip of an interviewee recorded in a satisfaction survey on a target issue and an initial satisfaction score given by the interviewee; performs emotion recognition on the first voice clip to obtain an emotion index of the interviewee, and the emotion index is used to reflect the emotional state of the interviewee; based on the emotion index, the initial satisfaction score is corrected to obtain a final satisfaction score. In this way, the influence of the interviewee's emotion on the satisfaction score can be eliminated, the score can be made more objective, and the accuracy and reliability of the satisfaction survey data can be improved.

[0151] An embodiment of the present invention also provides a data processing device, including a memory, a processor, and a computer program stored in the memory and runnable on the processor. When the processor executes the program, the various processes in the data processing method embodiment described above are implemented, and the same technical effect can be achieved. To avoid repetition, they will not be described here.

[0152] The present invention also provides a computer-readable storage medium having a computer program stored thereon. When executed by a processor, the program implements the various processes of the data processing method embodiment described above and achieves the same technical effects. To avoid repetition, the details are not described here. The computer-readable storage medium may be, for example, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0153] 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 take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to magnetic disk storage and optical storage, etc.) containing computer-usable program code.

[0154] 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 data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 means for performing functions specified in one or more processes and / or one or more blocks.

[0155] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable storage medium produce a paper product including an instruction device that implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0156] These computer program instructions can also be loaded onto a computer or other programmable data processing device to cause the computer or other programmable device to execute a series of operating steps 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 A step that specifies a function in one or more boxes.

[0157] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as within the scope of protection of the present invention.

Claims

1. A data processing method, characterized in that: include: Obtaining a first voice clip of a respondent collected during a satisfaction survey on a target issue and an initial satisfaction score given by the respondent; Performing emotion recognition on the first voice segment to obtain an emotion index of the interviewee, where the emotion index is used to reflect the emotional state of the interviewee; Based on the sentiment index, the initial satisfaction score is revised to obtain a final satisfaction score; The initial satisfaction score is modified based on the sentiment index to obtain a final satisfaction score, including: Obtaining attribute parameters of the interviewee; The emotion index and the attribute parameters of the respondent are used as feature vectors, and a score to be corrected is calculated using a pre-established correction model; Performing a sum operation on the initial satisfaction score and the score to be revised to obtain a final satisfaction score; The correction model is established by the following method, specifically including: By collecting voice data, a voice data sample library is constructed, wherein the voice data sample library stores voice data of different interviewees, each voice data including a second voice segment and a score of a question corresponding to the second voice segment; Extract features from each voice data and calculate the corresponding respondent's emotion index; A scoring model is obtained by machine learning based on the attribute parameters of the respondent corresponding to the first voice data, wherein the first voice data is voice data corresponding to an emotion index falling within a first preset range, and the emotion index within the first preset range corresponds to a neutral emotional state; The attribute parameters and emotional index of the respondent corresponding to the second voice data are used to obtain a correction model through machine learning. The second voice data is the voice data corresponding to the emotional index falling outside the first preset range. Among them, the voice data with the emotional index falling outside the first preset range corresponds to the respondent's emotional state being a negative emotional state or a positive emotional state. The score corresponding to the second voice data is matched with the scoring model through the correction model.

2. The method according to claim 1, characterized in that Obtain the first voice clip of the interviewee during the satisfaction survey on the target question, including: Obtaining a voice call recording of the respondent, wherein the voice call recording of the respondent specifically refers to a complete voice call recording from the time the respondent answers the call to the time the call is hung up in a satisfaction survey, and the voice call recording includes N questions that the interviewer requires the respondent to give a rating during the satisfaction survey, where N ≥ 1, and N is a positive integer; Based on the N questions, the voice call recording is segmented to obtain a first voice segment of the respondent recorded during the satisfaction survey on the target question, where the target question is at least one of the N questions.

3. The method according to claim 1, characterized in that The performing emotion recognition on the first voice segment to obtain the respondent's emotion index includes: The first speech segment is passed through a speech emotion recognition model to calculate the emotion index of the respondent.

4. The method according to claim 1, wherein The attribute parameters of the interviewee include the interviewee's own attributes and the attributes of the service used by the interviewee.

5. A data processing device, characterized in that: include: An acquisition module, configured to acquire a first voice clip of a respondent recorded in a satisfaction survey on a target question and an initial satisfaction score given by the respondent; an emotion recognition module, configured to perform emotion recognition on the first voice segment to obtain an emotion index of the respondent, wherein the emotion index is used to reflect the emotional state of the respondent; A correction module, configured to correct the initial satisfaction score based on the sentiment index to obtain a final satisfaction score; The correction modules include: A second acquiring unit, configured to acquire attribute parameters of the respondent; A second processing unit is configured to use the emotion index and the attribute parameters of the respondent as feature vectors and calculate a score to be corrected using a pre-established correction model; a correction unit, configured to perform a sum operation on the initial satisfaction score and the score to be corrected to obtain a final satisfaction score; The data processing device may further include: A sample library construction module is used to construct a voice data sample library by collecting voice data, wherein the voice data sample library stores voice data of different interviewees, each voice data including a second voice segment and a score corresponding to a question of the second voice segment; The calculation module is used to extract features from each voice data and calculate the corresponding emotional index of the interviewee; a first processing module configured to obtain a scoring model by machine learning based on the attribute parameters of the respondent corresponding to the first voice data, wherein the first voice data is voice data corresponding to an emotion index falling within a first preset range, and the emotion index within the first preset range corresponds to a neutral emotional state; The second processing module is used to obtain a correction model through machine learning for the attribute parameters and emotional index of the respondent corresponding to the second voice data, wherein the second voice data is voice data corresponding to the emotional index falling outside the first preset range, wherein the voice data with the emotional index falling outside the first preset range corresponds to the emotional state of the respondent being a negative emotional state or a positive emotional state, and the score corresponding to the second voice data is matched with the scoring model through the correction model.

6. A data processing device comprising a processor and a transceiver, wherein the transceiver receives and sends data under the control of the processor, characterized in that: The processor is configured to perform the following operations: Obtaining a first voice clip of a respondent collected during a satisfaction survey on a target issue and an initial satisfaction score given by the respondent; Performing emotion recognition on the first voice segment to obtain an emotion index of the interviewee, where the emotion index is used to reflect the emotional state of the interviewee; Based on the sentiment index, the initial satisfaction score is revised to obtain a final satisfaction score; The processor is further configured to execute the following process: Obtaining attribute parameters of the interviewee; The emotion index and the attribute parameters of the respondent are used as feature vectors, and a score to be corrected is calculated using a pre-established correction model; Performing a sum operation on the initial satisfaction score and the score to be revised to obtain a final satisfaction score; The processor is further configured to execute the following process: By collecting voice data, a voice data sample library is constructed, wherein the voice data sample library stores voice data of different interviewees, each voice data including a second voice segment and a score of a question corresponding to the second voice segment; Extract features from each voice data and calculate the corresponding respondent's emotion index; A scoring model is obtained by machine learning based on the attribute parameters of the respondent corresponding to the first voice data, wherein the first voice data is voice data corresponding to an emotion index falling within a first preset range, and the emotion index within the first preset range corresponds to a neutral emotional state; The attribute parameters and emotional index of the respondent corresponding to the second voice data are used to obtain a correction model through machine learning. The second voice data is the voice data corresponding to the emotional index falling outside the first preset range. Among them, the voice data with the emotional index falling outside the first preset range corresponds to the respondent's emotional state being a negative emotional state or a positive emotional state. The score corresponding to the second voice data is matched with the scoring model through the correction model.

7. The data processing device according to claim 6, characterized in that The processor is further configured to execute the following process: Obtaining a voice call recording of the respondent, wherein the voice call recording of the respondent specifically refers to a complete voice call recording from the time the respondent answers the call to the time the call is hung up in a satisfaction survey, and the voice call recording includes N questions that the interviewer requires the respondent to give a rating during the satisfaction survey, where N ≥ 1, and N is a positive integer; Based on the N questions, the voice call recording is segmented to obtain a first voice segment of the respondent recorded during the satisfaction survey on the target question, where the target question is at least one of the N questions.

8. The data processing device according to claim 6, characterized in that The processor is further configured to execute the following process: The first speech segment is passed through a speech emotion recognition model to calculate the emotion index of the respondent.

9. The data processing device according to claim 6, characterized in that The attribute parameters of the interviewee include the interviewee's own attributes and the attributes of the service used by the interviewee.

10. A data processing device comprising a memory, a processor, and a program stored in the memory and executable on the processor; When the processor executes the program, the data processing method according to any one of claims 1 to 4 is implemented.

11. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the data processing method according to any one of claims 1 to 4 are implemented.

Images