Data analysis method and device, storage medium and electronic equipment

Abstract

The invention discloses a data analysis method and device, a storage medium and electronic equipment, and relates to the field of artificial intelligence. The method comprises the steps that initial data information of a target user is acquired through a target system, the initial data information comprises text record information and behavior information of the target user, the text record information is text information provided by the target user for a target demand, and the behavior information is behavior data of the target user when the target user answers a preset question; converting the text record information into a text vector which can be recognized by a target system; extracting a behavior feature vector from the behavior information; and inputting the text vector and the behavior feature vector into a target model, and determining the matching degree between the target user and the target demand according to the text vector and the behavior feature vector through priori knowledge learned by the target model during model training. Through the method and the device, the technical problem of low analysis efficiency during user data analysis in the prior art is solved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence, and more specifically, to a data analysis method, device, storage medium, and electronic device. Background Art

[0002] In today's highly data-driven financial industry, there are more and more data analysis scenarios. Data analysis has become the core force to promote business decisions, optimize operational processes and improve customer experience. Whether it is credit risk assessment, user behavior analysis, employee performance appraisal, or talent screening in recruitment interviews, the financial field is facing unprecedented data challenges. Especially when dealing with large-scale, multi-dimensional data, traditional manual analysis methods have gradually exposed their shortcomings of low efficiency and high cost.

[0003] Taking the recruitment interview scenario as an example, banks need to evaluate applicants' professional knowledge, communication skills, emotional management and other aspects. Relying on the subjective judgment of human resources experts is not only time-consuming but also easily affected by personal bias, which greatly limits the accuracy and coverage of the evaluation. Especially when faced with a large amount of application information, the demand for manpower increases sharply, resulting in increased cost burden.

[0004] It can be seen that in the financial sector, there is a common technical bottleneck of low analysis efficiency in user data analysis. With the increasing amount of data and increasingly sophisticated analysis requirements, how to achieve automation and intelligence in data analysis to improve efficiency and reduce costs has become a key issue that the industry needs to solve urgently.

[0005] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0006] The main purpose of the present application is to provide a data analysis method, device, storage medium and electronic device to at least solve the technical problem of low analysis efficiency in user data analysis in the prior art.

[0007] In order to achieve the above-mentioned purpose, according to one aspect of the present application, a data analysis method is provided, which includes: obtaining initial data information of a target user through a target system, wherein the initial data information includes text record information and behavior information of the target user, the text record information is text information provided by the target user for the target demand, and the behavior information is the behavior data of the target user when answering preset questions; converting the text record information into a text vector recognizable by the target system; extracting a behavior feature vector from the behavior information, wherein the behavior feature vector is used to characterize the facial features, body features and voice features of the target user when answering preset questions; inputting the text vector and the behavior feature vector into a target model, and determining the matching degree between the target user and the target demand based on the text vector and the behavior feature vector through the prior knowledge learned by the target model during model training, wherein the target model is a neural network model trained using the text record information and behavior information of historical users as training samples and the matching degree between historical users and target demands as labels.

[0008] Optionally, before obtaining the initial data information of the target user through the target system, the data analysis method also includes: setting a target test scenario corresponding to the target user, wherein the target test scenario is used to simulate the target user's behavioral performance for the target needs; selecting analysis indicators corresponding to the target user according to the target test scenario, wherein the analysis indicators represent the analysis method for text record information and the analysis method for behavioral information.

[0009] Optionally, obtaining text record information of the target user includes: obtaining voice information of the target user when answering preset questions; converting the voice information into first text information; obtaining text input information of the target user from a user input interface of the target system; and using the text input information and the first text information as text record information of the target user.

[0010] Optionally, obtaining the target user's behavior information includes: obtaining voice information and image information of the target user when answering preset questions; determining the target user's language expression behavior information based on the voice information; determining the target user's body behavior information and facial behavior information based on the image information; and using the language expression behavior information, body behavior information and facial behavior information as the target user's behavior information.

[0011] Optionally, converting the text record information into a text vector recognizable by the target system includes: identifying at least one keyword in the text record information, wherein the keywords include at least words that reflect the tone of the target user and words that are associated with preset questions; and converting at least one keyword into a text vector based on a contextual relationship of the at least one keyword in the text record information.

[0012] Optionally, extracting a behavior feature vector from the behavior information includes: determining facial key point information, body key point information and sound information from the behavior information; determining the facial features of the target user when answering preset questions based on the facial posture composed of the facial key point information; determining the body features of the target user when answering preset questions based on the body posture composed of the body key point information; determining the sound features of the target user when answering the preset questions based on the sound information, wherein the sound features include at least the loudness of the target user's voice, speech speed information, rhythm change information, pause patterns and frequency of occurrence of different types of modal particles; converting the facial features, body features and sound features into feature vectors to obtain a behavior feature vector.

[0013] Optionally, the target model is obtained through the following steps: obtaining a training data set, wherein the training data set includes communication case information of L historical users, wherein L is an integer greater than 1, and the communication case information of each historical user is used to record the text record information, behavior information of the historical user for the target demand and the matching degree of the historical user with the target demand; obtaining the information amount of the communication case information of each historical user; determining the quality score corresponding to the communication case information of each historical user according to the information amount of the communication case information of each historical user; determining the communication case information of M target historical users from the training data set according to the quality score corresponding to the communication case information of each historical user, wherein M is an integer less than or equal to L and greater than 1; using the text record information and behavior information of the target historical user for the target demand recorded in the communication case information of each target historical user as training sample data; using the matching degree of each target historical user with the target demand as a training sample label; inputting the training sample data and the training sample label into the neural network, and obtaining the target model after multiple iterations of training.

[0014] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a data analysis device is also provided, including: an acquisition unit, which acquires initial data information of a target user through a target system, wherein the initial data information includes text record information and behavior information of the target user, the text record information is text information provided by the target user for the target demand, and the behavior information is the behavior data of the target user when answering preset questions; a conversion unit, which converts the text record information into a text vector recognizable by the target system; an extraction unit, which extracts a behavior feature vector from the behavior information, wherein the behavior feature vector is used to characterize the facial features, body features and voice features of the target user when answering preset questions; a determination unit, which inputs the text vector and the behavior feature vector into a target model, and determines the matching degree between the target user and the target demand based on the text vector and the behavior feature vector through the prior knowledge learned by the target model during model training, wherein the target model is a neural network model trained using the text record information and behavior information of historical users as training samples and the matching degree between historical users and target demands as labels.

[0015] Optionally, the data analysis device further includes: a setting unit and a selection unit. The setting unit is used to set a target test scenario corresponding to the target user, wherein the target test scenario is used to simulate the behavior performance of the target user for the target demand; the selection unit is used to select the analysis index corresponding to the target user according to the target test scenario, wherein the analysis index represents the analysis method for the text record information and the analysis method for the behavior information.

[0016] Optionally, the acquisition unit includes: a first acquisition subunit, a first conversion subunit, a second acquisition subunit and a first processing subunit. The first acquisition subunit is used to acquire the voice information of the target user when answering the preset question; the first conversion subunit is used to convert the voice information into first text information; the second acquisition subunit is used to acquire the text input information of the target user from the user input interface of the target system; and the first processing subunit is used to use the text input information and the first text information as the text record information of the target user.

[0017] Optionally, the acquisition unit includes: a third acquisition subunit, a first determination subunit, a second determination subunit, and a second processing subunit. The third acquisition subunit is used to acquire the voice information and image information of the target user when answering the preset question; the first determination subunit is used to determine the language expression behavior information of the target user according to the voice information; the second determination subunit is used to determine the body behavior information and facial behavior information of the target user according to the image information; and the second processing subunit is used to use the language expression behavior information, body behavior information, and facial behavior information as the behavior information of the target user.

[0018] Optionally, the conversion unit includes: a first recognition subunit and a second conversion subunit. The first recognition subunit is used to recognize at least one keyword in the text record information, wherein the keyword at least includes a word reflecting the tone of the target user and a word associated with a preset question; the second conversion subunit is used to convert the at least one keyword into a text vector according to the contextual relationship of the at least one keyword in the text record information.

[0019] Optionally, the extraction unit includes: a third determination subunit, a fourth determination subunit, a fifth determination subunit, a sixth determination subunit and a third conversion subunit. The third determination subunit is used to determine facial key point information, body key point information and sound information from the behavior information; the fourth determination subunit is used to determine the facial features of the target user when answering the preset question based on the facial gesture composed of the facial key point information; the fifth determination subunit is used to determine the body features of the target user when answering the preset question based on the body gesture composed of the body key point information; the sixth determination subunit is used to determine the sound features of the target user when answering the preset question based on the sound information, wherein the sound features at least include the loudness of the target user's voice, speech speed information, rhythm change information, pause rules and the frequency of occurrence of different types of modal particles; the third conversion subunit is used to convert the facial features, body features and sound features into feature vectors to obtain the behavior feature vector.

[0020] Optionally, the determination unit includes: a fourth acquisition subunit, a fifth acquisition subunit, a seventh determination subunit, an eighth determination subunit, a third processing subunit, a fourth processing subunit and a first training subunit. Among them, the fourth acquisition subunit is used to obtain the training data set, wherein the training data set includes the communication case information of L historical users, wherein L is an integer greater than 1, and the communication case information of each historical user is used to record the text record information, behavior information of the historical user for the target demand and the matching degree of the historical user with the target demand; the fifth acquisition subunit is used to obtain the information volume of the communication case information of each historical user; the seventh determination subunit is used to determine the quality score corresponding to the communication case information of each historical user according to the information volume of the communication case information of each historical user; the eighth determination subunit determines the communication case information of M target historical users from the training data set according to the quality score corresponding to the communication case information of each historical user, wherein M is an integer less than or equal to L and greater than 1; the third processing subunit is used to use the text record information and behavior information of the target historical user for the target demand recorded in the communication case information of each target historical user as training sample data; the fourth processing subunit is used to use the matching degree of each target historical user with the target demand as a training sample label; the first training subunit is used to input the training sample data and the training sample label into the neural network, and obtain the target model after multiple iterative training.

[0021] According to another aspect of an embodiment of the present application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein when the executable program is running, the device where the computer-readable storage medium is located is controlled to execute the above-mentioned data analysis method.

[0022] According to another aspect of an embodiment of the present application, there is also provided an electronic device, including: a memory storing an executable program; and a processor for running the program, wherein the above-mentioned data analysis method is executed when the program is running.

[0023] According to another aspect of an embodiment of the present application, a computer program product is also provided, including computer instructions, which implement the steps of the above-mentioned data analysis method when executed by a processor.

[0024] In the present application, firstly, the initial data information of the target user is acquired through the target system, wherein the initial data information includes the text record information and behavior information of the target user, wherein the text record information is the text information provided by the target user for the target demand, and the behavior information is the behavior data of the target user when answering the preset question, thereby ensuring the comprehensiveness of the acquired user data, and then the text record information is converted into a text vector recognizable by the target system, and a behavior feature vector is extracted from the behavior information, wherein the behavior feature vector is used to characterize the facial features, body features and voice features of the target user when answering the preset question, and then the text vector and the behavior feature vector are input into the target model, and the target model is trained according to the prior knowledge learned by the target model during model training, and the target user is trained according to the text vector and the behavior feature vector. The feature vector determines the matching degree between the target user and the target demand, wherein the target model is a neural network model trained with the text record information and behavior information of historical users as training samples and the matching degree between the historical users and the target demand as labels, that is, by obtaining the multi-dimensional data of the user, including facial information, body information, and text information, the multi-dimensional information is converted and features are extracted so that the system can process the data more quickly and accurately, and then the processed user data is input into the trained model to obtain the matching degree between the final user and the target demand, thereby achieving the purpose of automatically, comprehensively and accurately analyzing the user data to obtain the matching result, thereby achieving the technical effect of improving the efficiency of user data analysis, and then solving the technical problem of low analysis efficiency in user data analysis in the prior art. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] The drawings constituting a part of the present application are used to provide a further understanding of the present application. The illustrative embodiments and descriptions of the present application are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0026] Figure 1 A hardware structure block diagram of a computer terminal for implementing a data analysis method is shown;

[0027] Figure 2 is a flow chart of an optional data analysis method according to an embodiment of the present application;

[0028] Figure 3 is a schematic diagram of an optional data analysis device according to an embodiment of the present application;

[0029] Figure 4 It is a structural block diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0031] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] It should be noted that the collected information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for display, data for analysis, etc.) involved in this application are information and data authorized by the user or fully authorized by all parties, and the collection, storage, use, processing, transmission, provision, disclosure and application of relevant data are in compliance with relevant laws, regulations and standards, necessary confidentiality measures are taken, and public order and good customs are not violated, and corresponding operation entrances are provided for users to choose to authorize or refuse. For example, an interface is set up between this system and relevant users or institutions to provide users with corresponding operation entrances for users to choose to agree or refuse the results of automated decision-making; if the user chooses to refuse, the expert decision-making process will be entered.

[0033] Example 1

[0034] According to an embodiment of the present application, an embodiment of a data analysis method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0035] It should be noted that a data analysis system can be used as the execution subject of the data analysis method of the embodiment of the present application. It is understandable that the data analysis method provided in the embodiment of the present application can also be executed by other systems or devices, and the embodiment of the present application does not specifically limit this.

[0036] The method embodiments provided in the present application may be executed in a mobile terminal, a computer terminal or a similar computing device. Figure 1 FIG. 1 shows a hardware structure block diagram of a computer terminal (or mobile device) for implementing a digital analysis method. Figure 1 As shown, the computer terminal 10 (or mobile device) may include one or more (102a, 102b, ..., 102n are used to illustrate) processors 102 (the processor 102 may include but is not limited to a processing device such as a microprocessor MCU or a programmable logic device FPGA), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display, an input / output interface (I / O interface), a universal serial bus (USB) port (which may be included as one of the ports of the BUS bus), a network interface, a power supply and / or a camera. It can be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the above electronic device. Figure 1 More or fewer components as shown, or with Figure 1 Different configurations shown.

[0037] It should be noted that the one or more processors 102 and / or other data processing circuits described above may generally be referred to herein as "data processing circuits". The data processing circuit may be embodied in whole or in part as software, hardware, firmware, or any other combination thereof. In addition, the data processing circuit may be a single independent processing module, or may be incorporated in whole or in part into any of the other components in the computer terminal 10 (or mobile device). As in the data analysis method involved in the embodiments of the present application, the data processing circuit serves as a processor control (e.g., selection of a variable resistor terminal path connected to an interface).

[0038] The memory 104 can be used to store software programs and modules of application software, such as program instructions / data storage devices corresponding to the data analysis method in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, that is, realizing the above-mentioned data analysis method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the computer terminal 10 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0039] The transmission device 106 is used to receive or send data via a network. The specific example of the above network may include a wireless network provided by a communication provider of the computer terminal 10. In one example, the transmission device 106 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices through a base station so as to communicate with the Internet. In one example, the transmission device 106 can be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0040] The display may be, for example, a touch screen liquid crystal display (LCD) that enables a user to interact with a user interface of the computer terminal 10 (or mobile device).

[0041] Under the above operating environment, this application provides Figure 2 The data analysis methods shown. Figure 2 FIG. 1 is a flow chart of an optional data analysis method according to an embodiment of the present application. Figure 2 As shown, the method comprises the following steps:

[0042] Step S201, obtaining initial data information of a target user through a target system.

[0043] In step S201, the initial data information includes text record information and behavior information of the target user. The text record information is text information provided by the target user for target needs, and the behavior information is behavior data of the target user when answering preset questions.

[0044] Optionally, the target system refers to a platform or application for data collection and interaction, which is responsible for interacting with users and obtaining necessary data.

[0045] Optionally, the initial data information includes all text record information and behavior information of the user during the process of interacting with the system.

[0046] Optionally, the data analysis system first obtains text record information and behavior information of the target user through the target system, wherein the text record information can be collected directly from the user input, and the behavior information may be captured in real time through cameras, microphones and other sensors.

[0047] Step S202, converting the text record information into a text vector that can be recognized by the target system.

[0048] Optionally, the data analysis system uses natural language processing technology to convert text record information into text vectors.

[0049] Step S203: extracting a behavior feature vector from the behavior information.

[0050] In step S203, the behavior feature vector is used to characterize the facial features, body features, and voice features of the target user when answering the preset questions.

[0051] Optionally, the data analysis system extracts behavioral feature vectors from the behavioral information, which may include vector representations of facial expressions (such as smiling, frowning), vector representations of body language (such as crossing arms, nodding), and vector representations of voice intonation (such as pitch, rhythm, and intensity), which comprehensively reflect the user's emotional state, confidence level, and adaptability.

[0052] Step S204, input the text vector and the behavior feature vector into the target model, and determine the matching degree between the target user and the target demand based on the text vector and the behavior feature vector through the prior knowledge learned by the target model during model training.

[0053] In step S204, the target model is a neural network model trained using text record information and behavior information of historical users as training samples and the matching degree between historical users and target requirements as labels.

[0054] Optionally, the data analysis system inputs the text vector and behavior feature vector into a pre-trained target model. Through the calculation of the model, the matching score between the target user and the target demand can be obtained. This data-driven approach reduces the impact of human bias and improves the fairness of the analysis. The automated analysis process greatly shortens the analysis cycle, reduces the workload of analysts, and saves a lot of time.

[0055] Optionally, after obtaining the match score between the target user and the target needs, the data analysis system automatically generates a detailed evaluation report, including the score, areas of strength and suggestions for improvement, and provides instant feedback to the user. The system also provides personalized training and development suggestions based on the match score results.

[0056] Optionally, the data analysis system may protect static data (data in storage) and dynamic data (data in transmission) through encryption algorithms. The present application embodiment uses the SSL / TLS protocol (Secure Sockets Layer / Transport Layer Security) to ensure the security of data during network transmission.

[0057] Optionally, the data analysis system strictly restricts access to user data, and only authorized personnel can access sensitive information. Use multi-factor authentication and other security mechanisms to enhance the security of the system, and conduct regular security audits to check for system vulnerabilities and promptly fix any problems found.

[0058] Optionally, the data analysis system allows users to review the information they provide, and provides tools for users to correct or update their personal information, and provides users with a simple and easy-to-use mechanism so that they can request to delete their personal information. Users have the right to choose not to participate in certain data collection activities, or to stop using the service completely without affecting other rights and interests.

[0059] Optionally, in the credit application review process application scenario, the data analysis system can assist in assessing the applicant's credit status by analyzing the applicant's voice and text communication. For example, the system can set a series of questions about financial status, repayment willingness and ability, and judge the applicant's integrity and risk level through comprehensive analysis of the applicant's answer content, voice tone and facial expression. This helps banks obtain more comprehensive information in credit assessment, assist decision-making, and reduce the risk of non-performing loans; in the product recommendation personalization application scenario, when recommending bank products or services to customers, the system can more accurately infer the customer's needs and preferences based on the customer's tone, emotional tendency and conversation content in the interaction, so as to provide personalized suggestions and solutions. This product recommendation method based on emotion and text analysis can improve the accuracy of recommendations and customer acceptance; in the virtual banker service application scenario, in self-service channels such as telephone banking and online customer service, the data analysis system can play the role of a virtual banker, and communicate with customers smoothly through natural language understanding and generation technology, answer inquiries, process transactions, and even provide personal financial advice. This intelligent customer service method can improve service efficiency and reduce operating costs.

[0060] It can be known from the contents of step S201 to step S204 that in the present application, firstly, the initial data information of the target user is acquired through the target system, wherein the initial data information includes the text record information and behavior information of the target user, wherein the text record information is the text information provided by the target user for the target demand, and the behavior information is the behavior data of the target user when answering the preset question, thereby ensuring the comprehensiveness of the acquired user data, and then the text record information is converted into a text vector recognizable by the target system, and a behavior feature vector is extracted from the behavior information, wherein the behavior feature vector is used to characterize the facial features, body features and voice features of the target user when answering the preset question, and then the text vector and the behavior feature vector are input into the target model, and the target model is trained according to the prior knowledge learned by the target model during model training. The text vector and the behavior feature vector determine the matching degree between the target user and the target demand, wherein the target model is a neural network model trained with the text record information and behavior information of historical users as training samples and the matching degree between the historical users and the target demand as labels, that is, by obtaining multi-dimensional data of the user, including facial information, body information, and text information, converting and extracting features of the multi-dimensional information so that the system can process data more quickly and accurately, and then inputting the processed user data into the trained model to obtain the matching degree between the final user and the target demand, thereby achieving the purpose of automatically, comprehensively and accurately analyzing user data to obtain matching results, thereby achieving the technical effect of improving the efficiency of user data analysis, and further solving the technical problem of low analysis efficiency in user data analysis in the prior art.

[0061] In an optional embodiment, the data analysis system first sets up a target test scenario corresponding to the target user, wherein the target test scenario is used to simulate the target user's behavioral performance in response to the target needs, and then selects analysis indicators corresponding to the target user based on the target test scenario, wherein the analysis indicators represent the analysis method for text record information and the analysis method for behavioral information.

[0062] Optionally, before evaluating the match between the target user and the demand, the data analysis system will pre-design one or more target test scenarios, such as customer consulting services, sales product introductions, and handling customer complaints. These scenarios are set based on target needs and are intended to simulate the target user's behavior in actual work or specific tasks. Subsequently, the system selects corresponding analysis indicators based on these test scenarios. The analysis indicators define how to evaluate the collected text record information and behavioral information, which includes a deep understanding of the text content and a quantitative analysis of non-verbal behaviors (such as facial expressions, body movements, and changes in tone). In this way, the system can more accurately evaluate the target user's ability and potential when facing specific needs.

[0063] Optionally, when designing, the analysis indicators will be customized according to the specific requirements of the target test scenario and the capabilities that the target user needs to demonstrate. For example, the accuracy and richness of the target user's answers related to the target needs can be used to assess the target user's understanding of the target needs; the coherence and structure of the target user can be detected to determine the target user's ability to analyze and solve problems; the unique insights or innovative solutions proposed by the target user when solving problems can be identified to assess their innovation potential; the grammar, spelling and text fluency of the target user can be assessed to understand their written communication ability; the emotional stability and non-verbal communication ability of the candidate can be judged through the target user's facial expressions; the pitch, speed, pauses and intonation of the target user's answers can be analyzed to assess the clarity of the target user's voice and emotional expression; and information such as standing posture, gestures, eye contact, etc. can be extracted from the video data to determine the target user's confidence and interactive ability.

[0064] Optionally, the data analysis system quantifies each indicator and expresses it in the form of numbers or scores. The system combines the analysis results of these indicators to help analyze the match between target users and target needs.

[0065] From the above content, it can be seen that the data analysis system can evaluate the performance of target users when facing actual tasks by simulating specific scenarios in the real working environment. Moreover, since the target test scenarios and analysis indicators can be adjusted according to different target requirements and evaluation purposes, the system shows a high degree of adaptability and flexibility. At the same time, the analysis indicators not only cover the in-depth analysis of text record information, but also go deep into the multi-dimensional investigation of behavioral information. This kind of analysis ensures the comprehensiveness of the data and can improve the accuracy of user data analysis.

[0066] In an optional embodiment, the data analysis system first obtains voice information of the target user when answering preset questions, then converts the voice information into first text information, then obtains text input information of the target user from the user input interface of the target system, and finally uses the text input information and the first text information as text record information of the target user.

[0067] Optionally, the data analysis system collects text record information in two ways: one is to interact with the system through a microphone and convert the user's voice information into text using automatic speech recognition technology, where automatic speech recognition converts the user's voice input into understandable text information; the other is to directly obtain the text information entered by the user on the system interface. The combination of these two methods ensures that the system can understand the user's thoughts and expressions from multiple perspectives, thereby conducting a more in-depth and comprehensive analysis.

[0068] Optionally, for the processing of converting user speech to text, the data analysis system receives the user's speech signal, which is a continuous audio waveform. First, preprocessing is performed, wherein the preprocessing step may include noise removal to ensure the clarity of the audio signal and converting the signal into a format suitable for model input, such as converting the audio waveform into Mel-frequency cepstral coefficients or other spectral features. Next, a CNN model (Convolutional Neural Network) is used to extract key acoustic information from the preprocessed audio features, and then RNN (Recurrent Neural Network), especially LSTM (Long Short-Term Memory), is used to capture the sequence information and contextual dependencies of the speech and understand the coherence of words and sentences, which is crucial to ensuring the grammatical correctness and logical consistency of the output text. Through the dynamic update of its internal state, LSTM can remember the long-term contextual information of the speech and ignore irrelevant background noise, thereby improving the accuracy and robustness of recognition. Finally, through the joint action of CNN and RNN, the system can output text information corresponding to the user's speech input.

[0069] Optionally, the data analysis system utilizes advanced acoustic models and vocoder technology to enable the computer to produce natural and fluent human speech. This approach allows the system to dynamically generate responses without pre-recorded speech, enhancing the realism of the interaction.

[0070] Optionally, in order to further improve the recognition accuracy, the data analysis system uses an end-to-end learning framework to integrate the above steps into a unified model. This method can directly predict the text output from the original audio waveform, avoiding the information loss and recognition errors caused by traditional module segmentation, and finally generates text predictions based on the audio input. The recognized text may undergo post-processing steps, such as spelling checking and grammar correction, to improve the accuracy and readability of the output text.

[0071] From the above content, it can be seen that the data analysis system obtains the user's voice information and the input text information, so that it can fully understand the match between the user and the target needs, providing a data basis for subsequent analysis, and through automatic speech recognition technology, it can quickly process a large amount of voice data and convert it into analyzable text form, which can greatly shorten the analysis cycle and effectively improve the efficiency of the analysis.

[0072] In an optional embodiment, the data analysis system obtains the voice information and image information of the target user when answering preset questions, and then determines the language expression behavior information of the target user based on the voice information, and then determines the body behavior information and facial behavior information of the target user based on the image information, and finally uses the language expression behavior information, body behavior information and facial behavior information as the behavior information of the target user.

[0073] Optionally, the data analysis system first obtains the voice signals of the target user when answering the preset questions. These signals are converted into text information through automatic speech recognition technology, and the user's language expression behavior is analyzed from the text, including the tone of voice, speech speed, vocabulary used, etc. At the same time, the system records the image information of the target user when answering the questions through image acquisition devices such as cameras. Then, using computer vision technology, the system extracts body behavior information (such as gestures, postures) and facial behavior information (such as facial expressions, eye contact) from these images. This behavior information can reflect the user's non-verbal communication skills, such as confidence and concentration. Then, the language expression behavior information (such as tone, clarity of expression, and vocabulary use) obtained from the voice information analysis is integrated with the body behavior information (such as the openness of body language) and facial behavior information (such as the frequency of smiles) obtained through image information analysis to form a comprehensive behavior information data set.

[0074] From the above content, it can be seen that the behavioral information dataset obtained by the data analysis system not only contains the details of the user's verbal expression, but also combines observations of non-verbal behavior, such as the frequency of eye contact, changes in smiles or other facial expressions, and the degree of confidence in body language. This behavioral information dataset provides rich data support for evaluating the match between users and target needs.

[0075] In an optional embodiment, the data analysis system identifies at least one keyword in the text record information, wherein the keywords include at least words that reflect the tone of the target user and words that are associated with preset questions, and then converts the at least one keyword into a text vector based on the contextual relationship of the at least one keyword in the text record information.

[0076] Optionally, the data analysis system enhances the evaluation of the target user's communication ability, emotional state, and professional knowledge through deep text analysis. The system first identifies key words from the text record information generated by the user's answer to the preset question. These keywords include words that reflect the user's tone (such as "affirmation", "hesitation"), and professional terms or conceptual words closely related to the preset question, such as words that can reflect the user's emotions and words that can reflect whether the user's answer to the preset question is accurate. Then, according to the contextual relationship of the keywords in the text, the keywords are converted into text vectors. This conversion is based on word embedding technology, which maps each keyword into a high-dimensional vector space. These vectors not only contain the semantic information of the keywords themselves, but also reflect the meaning and usage of the keywords in a specific context, as well as the degree of association with other words.

[0077] Optionally, the process of identifying keywords involves natural language processing techniques, including part-of-speech tagging, entity recognition, and semantic analysis, to ensure that the system can accurately capture the core information and emotional color in the text. The data analysis system is based on LSTM and Transformer architecture (a deep learning model widely used in the field of natural language processing) to understand users' questions and answers and generate appropriate feedback. Transformer is a powerful deep learning model that is particularly good at processing natural language processing tasks. It uses a self-attention mechanism to focus on different parts of the input sequence, so that it can more efficiently capture the complex connections within and between sentences.

[0078] From the above content, it can be seen that in this way, the data analysis system can quantify the role of keywords in the text, provide numerical representation for subsequent analysis and evaluation, and facilitate subsequent model analysis and processing.

[0079] In an optional embodiment, the data analysis system determines facial key point information, limb key point information and sound information from the behavioral information, and then determines the facial features of the target user when answering preset questions based on the facial gestures composed of the facial key point information, determines the limb features of the target user when answering preset questions based on the limb gestures composed of the limb key point information, and determines the sound features of the target user when answering preset questions based on the sound information, wherein the sound features include at least the target user's voice loudness, speaking speed information, rhythm change information, pause patterns and frequency of occurrence of different types of modal particles, and then converts the facial features, limb features and sound features into feature vectors to obtain behavioral feature vectors.

[0080] Optionally, facial expression recognition and body recognition mainly rely on computer vision technology and deep learning algorithms, especially convolutional neural networks. The data analysis system captures the user's facial and body images in real time through an image acquisition device, and uses CNN technology to identify key feature points of the face and body, such as eyebrows, eyes, nose, mouth and surrounding muscle changes, body posture, hand movements, etc. These feature points are the basis of expression analysis and body analysis, and can reflect the user's emotional changes. CNN can recognize a variety of basic emotions such as happiness, sadness, surprise, anger, fear and disgust by training and learning facial expression patterns. The data analysis system analyzes the change sequence of these feature points to capture subtle changes in expression, so as to more accurately infer the user's emotional state.

[0081] Optionally, the data analysis system mainly uses signal processing technology and machine learning models to perform voice and intonation analysis. The system uses signal processing technology to extract acoustic parameters such as pitch, rhythm, intensity, timbre and pauses from the user's voice signal. These parameters can reflect the speaker's emotional state and tone, and then the extracted acoustic parameters are input into a recurrent neural network model such as LSTM. The LSTM network can maintain long-term memory and is particularly effective in processing time series data. It can capture the emotional patterns that change over time in the voice signal and identify the speaker's emotional tendencies.

[0082] From the above content, we can see that the data analysis system combines facial, body and voice information to enable the system to perform multimodal emotion analysis, which is not limited to understanding direct information in language, but also able to gain insight into the emotional state in non-verbal signals. This comprehensive analysis method improves the accuracy and credibility of emotion recognition.

[0083] In an optional embodiment, the data analysis system implements acquisition of a training data set, wherein the training data set includes communication case information of L historical users, wherein L is an integer greater than 1, and the communication case information of each historical user is used to record the text record information, behavior information of the historical user for the target demand and the matching degree of the historical user with the target demand, then the information amount of the communication case information of each historical user is obtained, and then the quality score corresponding to the communication case information of each historical user is determined based on the information amount of the communication case information of each historical user, and the communication case information of M target historical users is determined from the training data set based on the quality score corresponding to the communication case information of each historical user, wherein M is an integer less than or equal to L and greater than 1, and then the text record information and behavior information of the target historical user for the target demand recorded in the communication case information of each target historical user is used as training sample data, and then the matching degree of each target historical user with the target demand is used as a training sample label, and finally the training sample data and the training sample label are input into the neural network, and after multiple iterations of training, the target model is obtained.

[0084] Optionally, the data analysis system first collects communication case information of multiple historical users, professional knowledge related to the target needs, etc. Their communication case information includes text record information for a certain target need (such as text answering questions), behavior information (such as voice intonation, facial expressions and other non-verbal behaviors), and a quantitative matching score, which indicates the degree of fit between the user and the target need. Then, the quality score corresponding to the communication case information of each historical user is determined. This score determines the importance and reference value of the case in the model training process. Then, according to the quality score of the historical user communication case information, some of the most representative and training-worthy cases are selected as target training samples. These target historical users' Communication case information will be used for model training to ensure that the model can learn the most relevant evaluation patterns from high-quality data. Then, the text record information and behavior information of each target historical user in response to the target needs recorded in the communication case information of the target historical user will be used as training sample data, and the matching degree between each target historical user and the target needs will be used as the training sample label. Finally, based on the above high-quality data set, the supervised learning method is used to train the model, and multiple iterations of training are performed. During the training process, the model will continuously adjust its internal parameters to minimize the difference between the predicted results and the actual labels, so as to learn how to evaluate the matching degree between the user and the target needs based on the user's performance, thereby ensuring the accuracy and objectivity of data analysis.

[0085] Optionally, communication cases can be scored by professionals relevant to the target need, ensuring the quality and relevance of each sample.

[0086] Optionally, the data analysis system will periodically update the training data and models to adapt to new requirements and technological advances.

[0087] Optionally, in an application scenario, it can be an interview scenario, a personnel training scenario, or a personnel assessment scenario. For example, a user registers and logs into the system, and at the same time informs the user that the system will use his or her facial expressions, body shapes, and voice data, and obtains his or her consent. Then, the system sets up corresponding test scenarios according to the user's corresponding position, including simulated customer interaction, professional knowledge Q&A, etc., and then determines the key ability indicators that need to be measured, such as communication skills, professional knowledge mastery, and emotional management. After that, the user interacts with the system through a microphone, and the system uses automatic speech recognition technology to convert speech into text for analysis, and simultaneously performs emotional analysis of facial expressions and voice intonation to capture the user's emotional changes. The system will understand the user's questions and answers based on the LSTM and Transformer architectures, and generate appropriate feedback. Finally, based on the user's answer content, emotional expression, and other multi-dimensional data, the model comprehensively calculates an evaluation score, and automatically generates a detailed evaluation report, including scores, areas of strength, and suggestions for improvement. It can also provide personalized training and development suggestions based on the evaluation results. For training scenarios, for employees who have received training, the system can set up regular follow-up evaluations to monitor their growth.

[0088] Optionally, the data analysis system only collects the minimum amount of data required to complete the analysis. For video and audio information that does not require long-term storage of video clips, the system will immediately delete the original video data after the analysis is completed, and will anonymize or de-identify the collected data to reduce the risk of exposure of personally identifiable information.

[0089] From the above content, it can be seen that the data analysis system ensures the efficiency and effectiveness of model training by screening out high-quality training samples, avoiding the negative impact of low-quality data on model performance, thereby improving the accuracy and reliability of the final model. The trained model can quickly and accurately predict the match between users and target needs, thereby improving the efficiency and accuracy of user data analysis.

[0090] The data analysis method provided in the embodiment of the present application obtains multi-dimensional data of the user, including facial information, body information, and text information, converts and extracts features of the multi-dimensional information so that the system can process data more quickly and accurately, and then inputs the processed user data into a trained model to obtain a matching degree between the final user and the target demand, thereby achieving the purpose of automatically, comprehensively, and accurately analyzing the user data to obtain matching results, solving the technical problem of low analysis efficiency in the prior art for user data analysis, and thereby achieving the technical effect of improving the accuracy and efficiency of user data analysis.

[0091] Example 2

[0092] The embodiment of the present application also provides a data analysis device. It should be noted that the data analysis device of the embodiment of the present application can be used to execute the data analysis method provided in the embodiment of the present application. The data analysis device provided in the embodiment of the present application is introduced below.

[0093] According to an embodiment of the present application, a device for implementing the above data analysis method is also provided. Figure 3 is a schematic diagram of an optional data analysis device according to an embodiment of the present application, such as Figure 3 As shown, the device includes: an acquisition unit 301, a conversion unit 302, an extraction unit 303 and a determination unit 304.

[0094] Optionally, an acquisition unit 301 acquires initial data information of a target user through a target system, wherein the initial data information includes text record information and behavior information of the target user, wherein the text record information is text information provided by the target user for target needs, and the behavior information is behavior data of the target user when answering preset questions; a conversion unit 302 converts the text record information into a text vector recognizable by the target system; an extraction unit 303 extracts a behavior feature vector from the behavior information, wherein the behavior feature vector is used to characterize the facial features, body features, and voice features of the target user when answering preset questions; a determination unit 304 inputs the text vector and the behavior feature vector into a target model, and determines the degree of match between the target user and the target needs based on the text vector and the behavior feature vector through the prior knowledge learned by the target model during model training, wherein the target model is a neural network model trained using the text record information and behavior information of historical users as training samples and the degree of match between historical users and target needs as labels.

[0095] Optionally, the data analysis device further includes: a setting unit and a selection unit. The setting unit is used to set a target test scenario corresponding to the target user, wherein the target test scenario is used to simulate the behavior performance of the target user for the target demand; the selection unit is used to select the analysis index corresponding to the target user according to the target test scenario, wherein the analysis index represents the analysis method for the text record information and the analysis method for the behavior information.

[0096] Optionally, the acquisition unit 301 includes: a first acquisition subunit, a first conversion subunit, a second acquisition subunit, and a first processing subunit. The first acquisition subunit is used to acquire the voice information of the target user when answering the preset question; the first conversion subunit is used to convert the voice information into first text information; the second acquisition subunit is used to acquire the text input information of the target user from the user input interface of the target system; and the first processing subunit is used to use the text input information and the first text information as the text record information of the target user.

[0097] Optionally, the acquisition unit 301 includes: a third acquisition subunit, a first determination subunit, a second determination subunit, and a second processing subunit. The third acquisition subunit is used to acquire the voice information and image information of the target user when answering the preset question; the first determination subunit is used to determine the language expression behavior information of the target user according to the voice information; the second determination subunit is used to determine the body behavior information and facial behavior information of the target user according to the image information; and the second processing subunit is used to use the language expression behavior information, body behavior information, and facial behavior information as the behavior information of the target user.

[0098] Optionally, the conversion unit 302 includes: a first recognition subunit and a second conversion subunit. The first recognition subunit is used to recognize at least one keyword in the text record information, wherein the keyword at least includes a word reflecting the tone of the target user and a word associated with a preset question; the second conversion subunit is used to convert the at least one keyword into a text vector according to the contextual relationship of the at least one keyword in the text record information.

[0099] Optionally, the extraction unit 303 includes: a third determination subunit, a fourth determination subunit, a fifth determination subunit, a sixth determination subunit, and a third conversion subunit. The third determination subunit is used to determine facial key point information, body key point information, and sound information from the behavior information; the fourth determination subunit is used to determine the facial features of the target user when answering the preset question based on the facial gesture composed of the facial key point information; the fifth determination subunit is used to determine the body features of the target user when answering the preset question based on the body gesture composed of the body key point information; the sixth determination subunit is used to determine the sound features of the target user when answering the preset question based on the sound information, wherein the sound features at least include the loudness of the target user's voice, speech speed information, rhythm change information, pause rules, and the frequency of occurrence of different types of modal particles; the third conversion subunit is used to convert the facial features, body features, and sound features into feature vectors to obtain the behavior feature vector.

[0100] Optionally, the determination unit 304 includes: a fourth acquisition subunit, a fifth acquisition subunit, a seventh determination subunit, an eighth determination subunit, a third processing subunit, a fourth processing subunit and a first training subunit. Among them, the fourth acquisition subunit is used to obtain the training data set, wherein the training data set includes the communication case information of L historical users, wherein L is an integer greater than 1, and the communication case information of each historical user is used to record the text record information, behavior information of the historical user for the target demand and the matching degree of the historical user with the target demand; the fifth acquisition subunit is used to obtain the information volume of the communication case information of each historical user; the seventh determination subunit is used to determine the quality score corresponding to the communication case information of each historical user according to the information volume of the communication case information of each historical user; the eighth determination subunit determines the communication case information of M target historical users from the training data set according to the quality score corresponding to the communication case information of each historical user, wherein M is an integer less than or equal to L and greater than 1; the third processing subunit is used to use the text record information and behavior information of the target historical user for the target demand recorded in the communication case information of each target historical user as training sample data; the fourth processing subunit is used to use the matching degree of each target historical user with the target demand as a training sample label; the first training subunit is used to input the training sample data and the training sample label into the neural network, and obtain the target model after multiple iterative training.

[0101] The data analysis device provided in the embodiment of the present application obtains multi-dimensional data of the user, including facial information, body information, and text information, and converts and extracts features of the multi-dimensional information so that the system can process data more quickly and accurately. The processed user data is then input into a trained model to obtain a matching degree between the final user and the target demand, thereby achieving the purpose of automatically, comprehensively, and accurately analyzing the user data to obtain matching results, solving the technical problem of low analysis efficiency in the prior art for analyzing user data, and thereby achieving the technical effect of improving the accuracy and efficiency of user data analysis.

[0102] Example 3

[0103] An embodiment of the present application may provide an electronic device, Figure 4 is a structural block diagram of an electronic device according to an embodiment of the present application. Figure 4 As shown, the electronic device may include: one or more ( Figure 4 (only one is shown) processor 402, memory 404, storage controller, and peripheral interface, wherein the peripheral interface is connected to the radio frequency module, audio module and display.

[0104] Among them, the memory can be used to store software programs and modules, such as program instructions / modules corresponding to the methods and devices in the embodiments of the present application, and the processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, realizing the above-mentioned method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0105] The processor can call the information and application programs stored in the memory through the transmission device to perform the following steps: obtain the initial data information of the target user through the target system, wherein the initial data information includes the text record information and behavior information of the target user, the text record information is the text information provided by the target user for the target demand, and the behavior information is the behavior data of the target user when answering the preset questions; convert the text record information into a text vector that can be recognized by the target system; extract the behavior feature vector from the behavior information, wherein the behavior feature vector is used to characterize the facial features, body features and voice features of the target user when answering the preset questions; input the text vector and the behavior feature vector into the target model, and determine the matching degree between the target user and the target demand based on the text vector and the behavior feature vector through the prior knowledge learned by the target model during model training, wherein the target model is a neural network model trained using the text record information and behavior information of historical users as training samples and the matching degree between historical users and target demands as labels.

[0106] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: setting a target test scenario corresponding to the target user, wherein the target test scenario is used to simulate the target user's behavioral performance for the target needs; selecting analysis indicators corresponding to the target user according to the target test scenario, wherein the analysis indicators represent the analysis method for text record information and the analysis method for behavioral information.

[0107] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain the voice information of the target user when answering the preset questions; convert the voice information into first text information; obtain the text input information of the target user from the user input interface of the target system; and use the text input information and the first text information as text record information of the target user.

[0108] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain the voice information and image information of the target user when answering preset questions; determine the language expression behavior information of the target user based on the voice information; determine the body behavior information and facial behavior information of the target user based on the image information; and use the language expression behavior information, body behavior information and facial behavior information as the behavior information of the target user.

[0109] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: identify at least one keyword in the text record information, wherein the keywords at least include words that reflect the tone of the target user and words that are associated with preset questions; and convert at least one keyword into a text vector based on the contextual relationship of the at least one keyword in the text record information.

[0110] The processor can call the information and application programs stored in the memory through the transmission device to perform the following steps: determine the facial key point information, the limb key point information and the sound information from the behavior information; determine the facial features of the target user when answering the preset questions based on the facial posture composed of the facial key point information; determine the limb features of the target user when answering the preset questions based on the limb posture composed of the limb key point information; determine the sound features of the target user when answering the preset questions based on the sound information, wherein the sound features at least include the loudness of the target user's voice, speech speed information, rhythm change information, pause rules and the frequency of occurrence of different types of modal particles; convert the facial features, limb features and sound features into feature vectors to obtain behavior feature vectors.

[0111] By adopting the embodiment of the present application, a data analysis solution is provided, which obtains multi-dimensional data of the user, including facial information, body information, and text information, converts and extracts features of the multi-dimensional information so that the system can process the data more quickly and accurately, and then inputs the processed user data into the trained model to obtain the matching degree between the final user and the target demand, thereby achieving the purpose of automatically, comprehensively and accurately analyzing the user data to obtain the matching result, solving the technical problem of low analysis efficiency in the prior art for user data analysis, and thus achieving the technical effect of improving the accuracy and efficiency of user data analysis.

[0112] It can be understood by those skilled in the art that Figure 4 The structure shown is for illustration only, and the electronic device may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a PDA, a mobile Internet device (Mobile Internet Devices, MID), a PAD, or other terminal devices. Figure 4The structure of the electronic device is not limited. Figure 4 More or fewer components (such as network interfaces, display devices, etc.) shown in, or having Figure 4 Different configurations shown.

[0113] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0114] Example 4

[0115] The embodiment of the present application further provides a storage medium. Optionally, in this embodiment, the storage medium can be used to store the program code executed by the data analysis method provided in the first embodiment.

[0116] Optionally, in this embodiment, the above storage medium may be located in any computer terminal in a computer terminal group in a computer network, or in any mobile terminal in a mobile terminal group.

[0117] The present application also provides a computer program product, which, when executed on a data processing device, is suitable for executing the steps of the data analysis method.

[0118] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0119] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0120] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of the units can be a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, 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 units or modules, which can be electrical or other forms.

[0121] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0122] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0123] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions to enable a computer device (which can be a personal computer, a server or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk and other media that can store program codes.

[0124] The above is only a preferred embodiment of the present application. 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 application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A data analysis method, characterized in that: include: Acquire initial data information of the target user through the target system, wherein the initial data information includes text record information and behavior information of the target user, the text record information is text information provided by the target user for target needs, and the behavior information is behavior data of the target user in answering preset questions; Converting the text record information into a text vector recognizable by the target system; Extracting a behavior feature vector from the behavior information, wherein the behavior feature vector is used to characterize facial features, body features, and voice features of the target user when answering the preset question; The text vector and the behavior feature vector are input into the target model, and the matching degree between the target user and the target demand is determined according to the text vector and the behavior feature vector through the prior knowledge learned by the target model during model training, wherein the target model is a neural network model trained using the text record information and behavior information of historical users as training samples and the matching degree between the historical users and the target demand as labels.

2. The data analysis method according to claim 1, characterized in that: Before obtaining the initial data information of the target user through the target system, the data analysis method further includes: Setting a target test scenario corresponding to the target user, wherein the target test scenario is used to simulate the behavior performance of the target user in response to the target demand; An analysis indicator corresponding to the target user is selected according to the target test scenario, wherein the analysis indicator represents an analysis method for the text record information and an analysis method for the behavior information.

3. The data analysis method according to claim 1, characterized in that: Get the target user's text record information, including: Acquiring voice information of the target user when answering the preset question; Converting the voice information into first text information; Acquire text input information of the target user from the user input interface of the target system; The text input information and the first text information are used as text record information of the target user.

4. The data analysis method according to claim 1, characterized in that: Obtain the target user's behavior information, including: Acquire voice information and image information of the target user when answering the preset question; Determining language expression behavior information of the target user according to the voice information; Determining the body behavior information and facial behavior information of the target user according to the image information; The language expression behavior information, the body behavior information and the facial behavior information are used as the behavior information of the target user.

5. The data analysis method according to claim 1, characterized in that: Converting the text record information into a text vector recognizable by the target system includes: Identify at least one keyword in the text record information, wherein the keyword includes at least a word reflecting the tone of the target user and a word associated with the preset question; The at least one keyword is converted into the text vector according to a contextual relationship of the at least one keyword in the text record information.

6. The data analysis method according to claim 1, characterized in that: Extracting a behavior feature vector from the behavior information includes: Determine facial key point information, body key point information and sound information from the behavior information; Determining facial features of the target user when answering the preset question based on the facial gesture composed of the facial key point information; Determining the limb features of the target user when answering the preset question according to the limb posture composed of the limb key point information; Determining, based on the sound information, the sound features of the target user when answering the preset question, wherein the sound features at least include the target user's voice loudness, speech speed information, rhythm change information, pause rules, and the frequency of occurrence of different types of modal particles; The facial features, the body features, and the sound features are converted into feature vectors to obtain the behavior feature vector.

7. The data analysis method according to claim 1, characterized in that: The target model is obtained by the following steps: Acquire a training data set, wherein the training data set includes communication case information of L historical users, wherein L is an integer greater than 1, and the communication case information of each historical user is used to record text record information and behavior information of the historical user for the target demand and the matching degree between the historical user and the target demand; Obtaining the amount of communication case information of each historical user; Determining a quality score corresponding to the communication case information of each historical user according to the amount of information of the communication case information of each historical user; Determine, from the training data set, communication case information of M target historical users according to the quality score corresponding to the communication case information of each historical user, where M is an integer less than or equal to L and greater than 1; Using the text record information and behavior information of each target historical user for the target demand recorded in the communication case information of the target historical user as training sample data; The matching degree between each target historical user and the target requirement is used as a training sample label; The training sample data and the training sample labels are input into a neural network, and after multiple iterations of training, the target model is obtained.

8. A data analysis device, characterized in that: include: An acquisition unit, which acquires initial data information of a target user through a target system, wherein the initial data information includes text record information and behavior information of the target user, the text record information is text information provided by the target user for target needs, and the behavior information is behavior data of the target user when answering preset questions; A conversion unit, converting the text record information into a text vector recognizable by the target system; An extraction unit, extracting a behavior feature vector from the behavior information, wherein the behavior feature vector is used to characterize the facial features, body features, and voice features of the target user when answering the preset question; A determination unit inputs the text vector and the behavior feature vector into a target model, and determines the degree of matching between the target user and the target demand based on the text vector and the behavior feature vector through the prior knowledge learned by the target model during model training, wherein the target model is a neural network model trained using the text record information and behavior information of historical users as training samples and the degree of matching between the historical users and the target demand as labels.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored executable program, wherein when the executable program is executed, the device where the computer-readable storage medium is located is controlled to execute the data analysis method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: include: A memory storing an executable program; A processor, used to run the program, wherein the program executes the data analysis method described in any one of claims 1 to 7 when running.

11. A computer program product comprising computer instructions, characterized in that: When the computer instructions are executed by a processor, the steps of the data analysis method according to any one of claims 1 to 7 are implemented.

Images