Psychological consultation method, device and equipment based on AI and storage medium

Abstract

The invention relates to an AI-based psychological counseling method, and aims to provide personalized psychological counseling service through an intelligent technology. The method comprises the following steps: firstly, acquiring voice data of a user, and converting the voice data into text data through an automatic voice recognition technology; then, voice and text data of the user are analyzed, and user feature information including gender, age, emotional features and the like is extracted; and based on the analysis results, generating a cue word template for guiding the target psychological counseling model to generate personalized text psychological counseling suggestions meeting user requirements. Through automatic and intelligent analysis processes, the real-time performance and pertinence of psychological consultation are remarkably improved, and more accurate psychological health support can be provided for users.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to an AI-based psychological consultation method, device, equipment and storage medium. Background Art

[0002] In the existing technology, psychological counseling services are mainly carried out through traditional face-to-face communication. Although this method can effectively deal with various psychological problems, it also has many shortcomings, especially when dealing with the growing demand for mental health.

[0003] Traditional psychological counseling relies on face-to-face communication between offline counselors and users, which severely limits the coverage of services. Users usually need to make an appointment and go to a specific location for counseling, which has great time and geographical limitations, especially in remote areas or places with scarce resources. It is difficult for users to obtain timely and effective psychological support. In addition, due to limited counselor resources, users often have to wait for a long time, which to some extent reduces the timeliness and effectiveness of psychological intervention.

[0004] Traditional psychological counseling methods also have limitations in terms of personalized services. Counselors’ understanding of users often relies on communication in a few meetings, and it is difficult to gain an in-depth understanding of users’ long-term emotional changes and psychological patterns in a short period of time, thereby providing highly personalized advice. In addition, the subjective judgment of counselors may lead to uneven quality of advice, making it difficult to ensure consistency and professionalism.

[0005] In the traditional psychological counseling process, data such as users’ emotional changes and feedback information cannot usually be systematically collected and analyzed. This lack of data makes it difficult for counselors to track users’ emotional trends and adjust and optimize suggestions in subsequent counseling. The lack of an effective data feedback mechanism makes it difficult to continuously improve counseling results.

[0006] Face-to-face psychological counseling services are difficult to achieve 24 / 7 real-time response. When users encounter urgent psychological problems during non-working hours, it is often difficult to get help in time. This deficiency may cause users to miss the best time for intervention, which in turn aggravates psychological problems. Summary of the invention

[0007] The main purpose of the present invention is to provide an AI-based psychological counseling method, device, equipment and storage medium, aiming to solve the technical problem that existing psychological counseling methods are unable to automatically acquire and analyze users' voice data and emotional characteristics in real time to generate personalized psychological counseling suggestions.

[0008] To achieve the above object, the present invention provides an AI-based psychological consultation method, comprising:

[0009] Acquire user voice data, convert the user voice data into text data, and analyze the user voice data to identify user feature information, wherein the feature information includes gender, age, and emotional features;

[0010] Generate a prompt word template according to the text data and the feature information, wherein the prompt word template is used to guide the target psychological counseling model to generate psychological counseling suggestions;

[0011] The text data and the prompt word template are input into a target psychological counseling model to generate personalized text psychological counseling suggestions.

[0012] In one embodiment, before the text data and the prompt word template are input into the target psychological counseling model, the method further includes:

[0013] Obtain psychological data and perform structured processing to generate knowledge base documents that can be used by language models;

[0014] The knowledge base document is input into the artificial intelligence framework and combined with the preset language model to generate the target psychological counseling model.

[0015] In one embodiment, after generating the target psychological consultation model, the method further includes:

[0016] Record the user's emotional changes, user feedback information, and the actual effect of consulting suggestions for each consultation;

[0017] Generate new knowledge entries based on the recorded information, and use the new knowledge entries to update the knowledge base document;

[0018] Use the updated knowledge base document to adjust the target psychological counseling model.

[0019] In one embodiment, after generating personalized text psychological counseling suggestions, the method further includes:

[0020] The text psychological counseling suggestions are converted into voice psychological counseling suggestions through text-to-speech technology, and the voice psychological counseling suggestions are used to provide feedback on the user's psychological counseling;

[0021] Adjusting the timbre and tone of the voice psychological counseling suggestion according to the characteristic information of the user;

[0022] The voice psychological counseling suggestions are played to the user via a playback device.

[0023] In one embodiment, the user voice data is analyzed to identify the user's characteristic information, including:

[0024] Extracting basic frequency components from the user's voice data, the basic frequency components including high-frequency and pitch, and inferring the user's gender and age characteristics based on the basic frequency components;

[0025] The pitch fluctuation of the user's voice data is analyzed to identify the user's emotional characteristics and optimize gender characteristics and age characteristics.

[0026] In one embodiment, the text data and the prompt word template are input into a target psychological counseling model to generate personalized text psychological counseling suggestions, including:

[0027] Obtaining the user's historical consultation records, wherein the historical consultation records include the user's historical text data, historical text psychological consultation suggestions and user feedback information;

[0028] The historical consultation records are analyzed to generate user behavior and emotion pattern analysis results, and the text data is combined with the prompt word template to input the target psychological consultation model to generate the personalized text psychological consultation suggestions.

[0029] In one embodiment, after generating personalized text psychological counseling suggestions, the method further includes:

[0030] Collect data statistics on the user's consultation records within a preset period, including statistics on the content of each consultation suggestion, user feedback and emotional change data;

[0031] Generate a consulting report containing emotional change trends and consulting effect evaluation based on statistical data, and generate personalized mental health management suggestions based on the consulting report;

[0032] The personalized mental health management suggestions are fed back to the user in text and / or voice form.

[0033] Furthermore, to achieve the above-mentioned purpose, the present invention also provides an AI-based psychological consultation device, comprising:

[0034] A voice processing module, which obtains user voice data, converts the user voice data into text data, and analyzes the user voice data to identify user feature information, including gender, age, and emotional features;

[0035] A feature extraction module generates a prompt word template according to the text data and the feature information, wherein the prompt word template is used to guide the target psychological counseling model to generate psychological counseling suggestions;

[0036] The consultation generation module inputs the text data and the prompt word template into the target psychological consultation model to generate personalized text psychological consultation suggestions.

[0037] Furthermore, to achieve the above-mentioned purpose, the present invention also provides an AI-based psychological counseling device, which includes a memory, a processor, and an AI-based psychological counseling program stored in the memory and executable on the processor, and the AI-based psychological counseling program, when executed by the processor, implements the steps of the AI-based psychological counseling method as described above.

[0038] Furthermore, to achieve the above-mentioned purpose, the present invention also provides a computer storage medium, on which an AI-based psychological counseling program is stored, and when the AI-based psychological counseling program is executed by a processor, the steps of the AI-based psychological counseling method as described above are implemented.

[0039] Beneficial effects: The present invention relates to an AI-based psychological counseling method, which aims to provide personalized psychological counseling services through intelligent technology. The present invention first obtains the user's voice data and converts it into text data through automatic speech recognition technology. Then the user's voice and text data are analyzed to extract user feature information including gender, age, emotional characteristics, etc. Based on these analysis results, a prompt word template is generated to guide the target psychological counseling model to generate personalized text psychological counseling suggestions that meet user needs. The present invention significantly improves the real-time and pertinence of psychological counseling through automated and intelligent analysis processes, and can provide users with more accurate mental health support. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0041] Figure 1 This is a flow chart of an embodiment of an AI-based psychological consultation method of the present invention;

[0042] Figure 2 This is a functional module diagram of a preferred embodiment of an AI-based psychological counseling device of the present invention;

[0043] Figure 3 This is a schematic diagram of the structure of the device hardware operating environment involved in the AI-based psychological counseling device implementation scheme of the present invention. DETAILED DESCRIPTION

[0044] It should be understood that the specific embodiments described herein are only used to explain the present invention, and are not used to limit the present invention.

[0045] It should be noted that, in the existing technology, psychological counseling services are mainly carried out through traditional face-to-face communication. Although this method can effectively deal with various psychological problems, it also has many shortcomings, especially when responding to the growing demand for mental health.

[0046] Traditional psychological counseling relies on face-to-face communication between offline counselors and users, which severely limits the coverage of services. Users usually need to make an appointment and go to a specific location for counseling, which has great time and geographical limitations, especially in remote areas or places with scarce resources. It is difficult for users to obtain timely and effective psychological support. In addition, due to limited counselor resources, users often have to wait for a long time, which to some extent reduces the timeliness and effectiveness of psychological intervention.

[0047] Traditional psychological counseling methods also have limitations in terms of personalized services. Counselors’ understanding of users often relies on communication in a few meetings, and it is difficult to gain an in-depth understanding of users’ long-term emotional changes and psychological patterns in a short period of time, thereby providing highly personalized advice. In addition, the subjective judgment of counselors may lead to uneven quality of advice, making it difficult to ensure consistency and professionalism.

[0048] In the traditional psychological counseling process, data such as users’ emotional changes and feedback information cannot usually be systematically collected and analyzed. This lack of data makes it difficult for counselors to track users’ emotional trends and adjust and optimize suggestions in subsequent counseling. The lack of an effective data feedback mechanism makes it difficult to continuously improve counseling results.

[0049] Face-to-face psychological counseling services are difficult to achieve 24 / 7 real-time response. When users encounter urgent psychological problems during non-working hours, it is often difficult to get help in time. This deficiency may cause users to miss the best time for intervention, which in turn aggravates psychological problems.

[0050] See also Figure 1 , Figure 1 This is a flowchart of an embodiment of an AI-based psychological counseling method provided by the present invention. It should be noted that although a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than that shown here.

[0051] like Figure 1 As shown, the AI-based psychological consultation method proposed by the present invention includes the following steps:

[0052] S10, acquiring user voice data, converting the user voice data into text data, and analyzing the user voice data to identify user feature information, where the feature information includes gender, age, and emotional features;

[0053] In this embodiment, the user's voice input can be converted into machine-processable text data through automatic speech recognition technology (Automatic Speech Recognition, ASR). This process involves multiple core technical links, including speech signal processing, noise elimination, voiceprint recognition, etc.

[0054] The user's voice features are analyzed at multiple levels to extract the user's gender, age and emotional characteristics. These characteristic information are the key input for the subsequent generation of personalized psychological counseling suggestions. Specifically, gender and age characteristics can be inferred through voiceprint features (such as frequency and pitch), while emotional characteristics can be identified through voice pitch fluctuations, speaking speed, tone and other parameters.

[0055] In one specific implementation, the basic automatic speech recognition and feature extraction method includes:

[0056] Acquire voice data: Users input voice data into the system through a microphone or smart device, and the data is captured in real time through the voice acquisition module.

[0057] Voice data conversion: The system uses automatic speech recognition technology to convert captured voice signals into text data. ASR technology achieves efficient and accurate transcription by matching the sound wave signal of the voice with the preset voice model.

[0058] Feature extraction: After the text is generated, the system performs feature analysis on the original voice data, including frequency analysis, pitch analysis, etc., to extract the user's gender and age characteristics. In addition, the system detects the emotional state in the voice, such as anxiety, anger, calmness, etc., through the sentiment analysis module.

[0059] In another specific implementation, the method based on enhanced speech feature analysis includes:

[0060] Acquiring voice data: Users input voice through the high-quality microphone of the smart device, and the system will pre-process it according to the environmental noise conditions to reduce the interference of background noise.

[0061] Voice data conversion: The system not only uses ASR technology to convert speech to text, but also uses multimodal fusion technology to combine speech with other perceptual information (such as facial expressions, heart rate, etc.) to improve the accuracy of text transcription.

[0062] Multi-level feature extraction: The system uses a deep learning model to comprehensively analyze the user's voice data and text data to further refine the classification of gender, age and emotional features. For example, through the specific frequency components and tone characteristics in the voice, the system can identify the user's emotional changes in a specific situation and dynamically adjust the analysis strategy.

[0063] In other specific implementations, the methods based on real-time emotional state monitoring and feedback include:

[0064] Acquire voice data: Users input information through voice, and the system will monitor the emotional state in the voice in real time and mark the emotional fluctuations when capturing the voice.

[0065] Voice data conversion: While converting voice to text, the system performs real-time analysis of the emotional fluctuations in the voice and processes the analysis results synchronously with the text data.

[0066] Dynamic feature extraction: The system combines the user's historical emotional data and current voice input, and continuously adjusts the weight of feature extraction through a real-time updated model to identify more accurate emotional features. This method is particularly suitable for scenarios where the user's emotional state changes rapidly.

[0067] Example description:

[0068] Users interact with the system through a psychological counseling application on their smartphones. The user expresses their current anxiety, and the system collects the user's voice data through a microphone and immediately converts it into text data. During the speech recognition process, ASR technology accurately transcribes each word in the speech into text while eliminating background noise.

[0069] While generating text, the system analyzes the user's voice, identifies higher-frequency voice components, and combines pitch fluctuations to infer that the user is a young woman who shows obvious anxiety and uneasiness.

[0070] Based on these characteristic information, the system generates a special prompt word template, which contains specific coping strategies and encouraging sentences for anxiety. The system then inputs the template and the user's text data into the target psychological counseling model to generate a personalized text psychological counseling suggestion, which recommends deep breathing exercises and provides several feasible relaxation methods.

[0071] Finally, the system converts the generated suggestions into speech through text-to-speech technology, gently feeds back to the user, and displays the corresponding text suggestions on the application interface.

[0072] Through ASR technology and voice data analysis, it is possible to capture and process the user's emotional state in real time, generating timely and effective psychological counseling suggestions. This solves the problems of slow response time and insufficient personalization in traditional psychological counseling. By analyzing the user's gender, age and emotional characteristics, the system can provide highly personalized psychological counseling services, ensuring that the suggestions are more in line with the user's actual needs, and increasing the effectiveness of psychological counseling.

[0073] S20, generating a prompt word template according to the text data and the feature information, wherein the prompt word template is used to guide the target psychological counseling model to generate psychological counseling suggestions;

[0074] In this embodiment, the user's text data is combined with the extracted feature information (such as gender, age, and emotional characteristics) to generate a template for guiding the target psychological counseling model to generate personalized suggestions. The prompt word template is a structured data input that generates corresponding guiding content based on the user's specific situation (such as current emotional state, personalized needs, etc.), ensuring that the psychological counseling model can generate more accurate and user-friendly suggestions.

[0075] In one specific implementation, the method of generating a prompt word template based on a rule includes:

[0076] Feature information matching: The system first matches the preset rule library based on the user's text data and feature information (such as gender, age, and emotional characteristics). For example, if it is detected that the user is depressed and is a young or middle-aged woman, the system may select a template related to anxiety relief.

[0077] Template generation: The system selects entries corresponding to the matching rules from the preset prompt word library and generates a prompt word template. The template may include "anxiety management skills", "emotional support suggestions", etc., which are used to guide the model to generate corresponding psychological counseling suggestions.

[0078] In another specific implementation, the method of generating the prompt word template based on machine learning includes:

[0079] Data input and model training: The system uses a large amount of user data to train the prompt word generation model, which can learn how to automatically generate prompt word templates based on the user's text and feature information. Through deep learning technology, the model can identify more complex user emotions and needs and generate more customized prompt words.

[0080] Dynamic adjustment: The system dynamically adjusts the prompt word template based on real-time feedback and new data to ensure that the generated psychological counseling suggestions can reflect the user's latest emotional state and needs.

[0081] In other specific implementations, the method of generating a prompt word template based on multimodal information fusion includes:

[0082] Multimodal information acquisition: In addition to text data and voice features, the system can also combine multimodal information such as the user's expression and heart rate to generate a more comprehensive prompt word template.

[0083] Template optimization: The system integrates multimodal information and generates prompt word templates through comprehensive analysis, which can guide the psychological counseling model to generate more detailed and accurate psychological counseling suggestions. For example, the system can combine the nervousness detected in the voice and the anxiety characteristics in the facial expressions to generate targeted counseling prompt words.

[0084] Example description:

[0085] The user inputs voice data through a smart device, and the system converts it into text data, and detects that the user shows anxiety during the consultation. The user's characteristic information shows that she is a woman in her 30s. Based on this information, the system generates a prompt word template, including "anxiety management skills", "mindfulness practice guidance" and "positive emotion reinforcement".

[0086] These prompt word templates were input into the target psychological counseling model, which generated a personalized psychological counseling recommendation, suggesting that the user relieve anxiety through deep breathing exercises and mindfulness meditation, and provided some practical methods. In addition, the system also recommended that the user keep a daily emotional journal to help her better manage her emotions.

[0087] By generating prompt word templates, highly customized psychological counseling suggestions can be generated according to the user's specific situation, significantly improving the personalization of services. The process of model-generated suggestions is simplified, and the accuracy and efficiency of generated suggestions are improved through structured input.

[0088] S30, inputting the text data and the prompt word template into a target psychological counseling model to generate personalized text psychological counseling suggestions.

[0089] In this embodiment, the target psychological counseling model is a large language model based on artificial intelligence (AI) technology. It has learned a lot of psychological knowledge and counseling cases through training, and can generate counseling suggestions that meet specific user needs based on input text data and prompt word templates. This process not only includes the analysis of the user's current input, but also involves the understanding and application of the prompt word template to ensure that the generated suggestions are highly consistent with the user's emotional state and specific needs.

[0090] In one specific implementation, the method of generating recommendations based on the standard model includes:

[0091] Data input: The system inputs the user's text data and the prompt word template generated according to the user's characteristics into the target psychological counseling model. This model may be a specially optimized large-scale pre-trained language model such as GPT, BERT, etc., which already has strong language understanding and generation capabilities.

[0092] Suggestion generation: The model performs semantic understanding based on the input data and prompt word templates, and combines the content of the internal knowledge base to generate personalized text psychological counseling suggestions that meet the user's current needs. The generated suggestions may include emotional support, emotion management methods, daily mental health maintenance strategies, etc.

[0093] In another specific implementation, the method of generating suggestions based on the context-aware model includes:

[0094] Context-aware input: The system inputs not only the current text data and prompt word templates, but also the user's historical interaction records, allowing the model to understand the context of the current conversation. This helps generate more consistent and coherent psychological counseling suggestions, especially in multiple consecutive consultations.

[0095] Emotion matching generation: Based on the prompt word template, the model focuses on the user's current emotional state and generates more empathetic suggestions. For example, if the user shows anxiety, the model may generate encouraging or soothing suggestions.

[0096] In other specific implementations, the method of generating suggestions based on the real-time feedback model includes:

[0097] Dynamically adjust input: In the process of generating suggestions, the system can dynamically adjust the weights of text data and prompt word templates based on real-time feedback from users. For example, if a user shows emotional changes during the consultation process, the system can adjust the input data to prompt the model to generate suggestions that are more in line with the user's current state.

[0098] Generation and feedback loop: After the model generates preliminary suggestions, the system can immediately feed them back to the user and further adjust the suggestions based on the user's response. This real-time interaction makes the generated psychological counseling suggestions more targeted and timely.

[0099] Example description:

[0100] The user inputs "I've been under a lot of pressure lately and I feel like I can't breathe" through voice. The system inputs this text data into the target psychological counseling model along with the user's anxiety emotions detected previously. The prompt word templates include "anxiety management skills" and "breathing regulation methods."

[0101] The model combines the user's text data and the prompt word template to generate a personalized suggestion, which reads, "When you feel overly stressed, you can try deep breathing exercises. Find a quiet place, inhale and exhale slowly, and try to focus on your breathing. At the same time, you can organize your thoughts by writing in a diary to help yourself better cope with stress."

[0102] The generated text suggestions are presented to the user through the application interface and suggest that the user can reduce anxiety through daily exercises.

[0103] By combining text data with prompt word templates, the target psychological counseling model can generate highly personalized suggestions to ensure that the user's emotional and psychological needs are accurately responded to. Using the powerful computing power of the AI ​​model, the system can generate psychological counseling suggestions that meet user needs in a very short time, significantly improving the efficiency and response speed of counseling. The ability to generate and feedback in real time enables the system to provide more intimate and real-time psychological support, enhancing the user experience and satisfaction.

[0104] The present invention relates to an AI-based psychological counseling method, which aims to provide personalized psychological counseling services through intelligent technology. The present invention first obtains the user's voice data and converts it into text data through automatic speech recognition technology. Then the user's voice and text data are analyzed to extract user feature information including gender, age, emotional characteristics, etc. Based on these analysis results, a prompt word template is generated to guide the target psychological counseling model to generate personalized text psychological counseling suggestions that meet the user's needs. The present invention significantly improves the real-time and pertinence of psychological counseling through an automated and intelligent analysis process, and can provide users with more accurate mental health support.

[0105] In one embodiment, in the above S30, before the text data and the prompt word template are input into the target psychological counseling model, the method further includes:

[0106] a1, obtain psychology data and perform structured processing to generate knowledge base documents that can be used by the language model;

[0107] a2, input the knowledge base document into the artificial intelligence framework and combine it with the preset language model to generate the target psychological counseling model.

[0108] In this embodiment, before inputting text data and prompt word templates into the target psychological counseling model, it is first necessary to obtain psychological data and structure it to generate a knowledge base document that can be used by the language model. This step is the basis for ensuring that the target psychological counseling model can accurately generate personalized suggestions. By building a rich and structured knowledge base, the system can provide the model with professional and reliable psychological background knowledge, and enhance the quality and scientificity of generated suggestions.

[0109] The acquisition of psychology data includes collecting a large number of psychology books, research papers, case studies, consultation conversation records, etc. After these data are cleaned, classified, annotated and structured, the generated knowledge base documents can provide accurate references for the language model in the artificial intelligence framework.

[0110] The purpose of inputting knowledge base documents into the artificial intelligence framework and combining them with the preset language model is to train or fine-tune the target psychological counseling model so that it has the ability to understand and apply psychological knowledge. In this way, the model can not only generate fluent suggestions, but also ensure that the suggestions are consistent with psychological theory and practice.

[0111] In one specific implementation, the method of building and generating a model based on a basic knowledge base includes:

[0112] Psychological data collection: The system extracts important data from public psychology literature, books, and research papers to ensure that the knowledge base contains rich psychological theories and practical cases.

[0113] Data cleaning and structured processing: The system removes noise, duplicates, classifies and labels the collected psychology data. The processed data is structured into knowledge base documents to adapt to the input format of the language model, such as knowledge graphs or structured databases.

[0114] Model training: Input the structured knowledge base documents into the artificial intelligence framework, and train or fine-tune them in combination with the preset language model (such as GPT, BERT, etc.) to enable the model to generate psychological counseling advice.

[0115] In another specific implementation, the method of constructing a knowledge base based on review by domain experts includes:

[0116] Experts participate in knowledge base construction: During the data cleaning and structuring stage, experts in the field of psychology are introduced to review and ensure the professionalism and reliability of the knowledge base content. Experts can annotate the data or add additional knowledge items to make the knowledge base more complete.

[0117] Model fine-tuning: After the knowledge base documents are input into the AI ​​framework, the system fine-tunes the language model based on feedback from domain experts, so that the model can more accurately apply psychological knowledge when generating psychological counseling recommendations.

[0118] In other specific implementations, the methods based on dynamic knowledge base updating and model generation include:

[0119] Real-time data update: The system extracts data from the latest psychology research and consulting cases regularly or in real time to dynamically update the knowledge base. The updated knowledge base is restructured and used to fine-tune the existing language model.

[0120] Model adaptive training: After each knowledge base update, the system will perform adaptive training on the target psychological counseling model to ensure that the model can apply the latest psychological knowledge and generate psychological counseling suggestions that are in line with the development of the times and new discoveries.

[0121] Example description:

[0122] The system collects a large amount of psychological data from multiple authoritative psychology journals, research databases, and clinical case reports. These data include the latest research results in anxiety management, depression treatment, cognitive behavioral therapy, etc. The system first cleans and structures these data and converts them into a knowledge graph format. After the knowledge base document is completed, the system inputs it into the GPT-based artificial intelligence framework and fine-tunes it in combination with the preset language model.

[0123] The trained target psychological counseling model can generate more scientific and personalized suggestions. For example, when a user expresses anxiety, the model can recommend deep breathing and mindfulness exercises based on the latest psychological research, while providing specific strategies for coping with anxiety in daily life.

[0124] This embodiment acquires and structures psychological data, and the system constructs a scientific and professional knowledge base to ensure that the generated psychological counseling suggestions are in line with psychological theory. Combining the structured knowledge base and the preset language model, the target psychological counseling model can generate more accurate and personalized psychological counseling suggestions to meet the actual needs of users. By regularly updating the knowledge base and adaptively training the model, the system can maintain the timeliness and cutting-edge nature of psychological counseling suggestions, ensuring that users receive the best quality mental health support.

[0125] In one embodiment, in a2 above, after generating the target psychological counseling model, the method further includes:

[0126] b1, record the user's emotional changes, user feedback information, and the actual effect of consultation suggestions for each consultation;

[0127] b2, generating new knowledge entries according to the recorded information, and using the new knowledge entries to update the knowledge base document;

[0128] b3. Use the updated knowledge base document to adjust the target psychological counseling model.

[0129] In this embodiment, after generating the target psychological counseling model, the system further records the user's emotional changes, user feedback information, and the actual effect of counseling suggestions for each consultation. These recorded information are used to generate new knowledge items and update the knowledge base documents. This process is the key to dynamically adjusting and optimizing the target psychological counseling model, which enables the model to continuously learn and adapt to changes in user needs and improve the accuracy and effectiveness of psychological counseling suggestions.

[0130] After each psychological consultation, the user's emotional changes during the consultation process (such as the transition from anxiety to relaxation) and the user's feedback on the consultation suggestions (such as whether the user adopts the suggestions and what the effect is) are recorded. These data can provide a basis for generating new knowledge items.

[0131] Based on these recorded data, the system generates new knowledge items that reflect the user's emotional changes and behavioral responses in specific situations. These new knowledge items are then integrated into the existing knowledge base documents, ensuring that the knowledge base can continue to reflect the latest user behavior and psychological developments.

[0132] Using the updated knowledge base documents, the system can adjust and optimize the target psychological counseling model so that it can generate more accurate and personalized psychological counseling suggestions in future counseling processes.

[0133] In one specific implementation, the method based on regular recording and knowledge base updating includes:

[0134] Emotion and feedback records: After each psychological consultation, the system records the user's emotional changes and feedback information. The recorded information includes the user's emotional fluctuations during the consultation process and the degree of acceptance of the generated suggestions.

[0135] Knowledge entry generation: The system generates new knowledge entries based on the recorded data. These entries may include new emotional response patterns, user response methods in specific situations, etc.

[0136] Knowledge base update: Integrate the generated knowledge entries into the existing knowledge base documents to ensure that the knowledge base always contains the latest user behavior and feedback information.

[0137] Model adjustment: After the knowledge base is updated, the system will fine-tune the target psychological counseling model so that it can adapt to the content of the new knowledge items and improve the quality of counseling advice.

[0138] In another specific implementation, the method based on real-time dynamic update and model optimization includes:

[0139] Real-time data recording: During the psychological consultation process, the system records the user's emotional changes and feedback information in real time. These data will be used to generate new knowledge entries immediately after the consultation.

[0140] Automated knowledge entry generation: The system uses machine learning algorithms to automatically convert real-time recorded data into new knowledge entries and dynamically update the knowledge base.

[0141] Continuous model optimization: The system continuously monitors the performance of the model, and after the knowledge base is updated, it adjusts and optimizes the reasoning and generation logic of the target psychological counseling model in real time to ensure that the model can quickly adapt to the user's new needs.

[0142] In other specific implementations, the methods based on personalized feedback loop and knowledge base adaptation include:

[0143] Personalized feedback loop: The system generates personalized feedback mechanisms for different users, and records detailed feedback information based on the user's long-term behavior patterns and emotional changes. This information will serve as the core data source for generating new knowledge items.

[0144] Adaptive knowledge base update: The system adaptively updates the knowledge base based on personalized feedback and data analysis results. This approach enables the knowledge base to more accurately reflect the psychological state and needs of individual users.

[0145] Model refinement: Based on the updated knowledge base, the system makes refinement adjustments to the target psychological counseling model to ensure that the generated counseling suggestions accurately match the user's current status and long-term needs.

[0146] Example description:

[0147] The user showed slight anxiety during a psychological consultation and gave feedback after the consultation that the breathing adjustment method in the suggestion was very helpful. The system recorded the user's emotional changes during the consultation, from initial anxiety to final relaxation, and recorded the user's positive feedback on the suggestion.

[0148] Based on this data, the system generated new knowledge items, including "Anxious users' positive response to breathing regulation" and "The effectiveness of guiding users to perform breathing regulation under mild anxiety." These knowledge items were integrated into the existing knowledge base and used to fine-tune the target psychological counseling model.

[0149] In the next consultation, when the system detects similar anxiety, the model is more inclined to generate suggestions related to breathing regulation and can further optimize the suggestions based on user feedback.

[0150] By recording the data of each consultation and updating the knowledge base, this embodiment enables the system to continuously learn the user's behavior patterns and emotional reactions, thereby continuously adapting and optimizing the psychological consultation model and generating more accurate suggestions. The new knowledge items reflect the user's individual psychological state and needs, enabling the system to provide more personalized suggestions in future consultations and improve user satisfaction. The update of the knowledge base and the fine-tuning of the model enable the system to maintain its scientificity and cutting-edge nature, ensuring that high-quality mental health support can be provided for different users and situations.

[0151] In one embodiment, in the above S30, after generating personalized text psychological counseling suggestions, the following is further included:

[0152] c1, converting the text psychological counseling suggestions into voice psychological counseling suggestions through text-to-speech technology, and the voice psychological counseling suggestions are used to feedback the user's psychological counseling;

[0153] c2, adjusting the timbre and tone of the voice psychological counseling suggestion according to the characteristic information of the user;

[0154] c3, playing the voice psychological counseling suggestions to the user through a playback device.

[0155] In this embodiment, after generating personalized text psychological counseling suggestions, the system also includes converting these text suggestions into voice suggestions through text-to-speech (TTS) technology. The core of this step is to make the psychological counseling suggestions fed back to the user in a more natural and friendly way, especially when the user needs to get advice through hearing. In this way, the system can not only provide counseling suggestions in text form, but also respond to user needs in real time and flexibly in voice, thereby improving the user experience.

[0156] Text-to-speech technology refers to converting the generated text psychological counseling suggestions into natural language voice output through speech synthesis technology. TTS technology can adjust the voice timbre, speaking speed and emotional expression according to the user's gender, age, emotional state and other information, making the voice suggestions more in line with the user's psychological state and acceptance habits.

[0157] In one specific implementation, the standard TTS conversion and feedback method includes:

[0158] TTS conversion: The system inputs the generated text psychological counseling suggestions into the TTS module, and the TTS module converts the text content into speech according to the preset speech synthesis parameters (such as speaking speed, timbre, and pitch).

[0159] Voice feedback: The system plays the generated voice psychological counseling suggestions to the user through speakers or headphones, ensuring that the user can receive the suggestions in the form of voice. This method is especially suitable for users when they are on the move or when it is inconvenient to read.

[0160] In another specific implementation, the method based on emotion-enhanced TTS conversion includes:

[0161] Emotional adaptation conversion: During the TTS conversion process, the system dynamically adjusts the emotional expression of the voice according to the user's current emotional state and psychological characteristics. For example, when the user is in an anxious state, the system can choose a calmer and more soothing voice style.

[0162] Emotion matching feedback: The generated voice suggestions are not only accurate in content, but also can further enhance the user's psychological acceptance and comfort through the tone and emotion of the voice.

[0163] In other specific implementations, the multi-language TTS support method includes:

[0164] Multi-language conversion: The system supports TTS conversion in multiple languages ​​and can generate corresponding voice suggestions based on the user's language preferences. For example, the system can automatically select the corresponding TTS language package based on the user's voice input language and generate voice suggestions that suit the user's language habits.

[0165] Personalized language feedback: The system can provide personalized voice suggestion services for users with different language backgrounds through multi-language support, improving the applicability of the system and user satisfaction

[0166] Example description:

[0167] The user generated a text psychological counseling suggestion on coping with anxiety through the system, including "deep breathing exercises" and "emotional journal writing". The system inputs this text suggestion into the TTS module, which selects a gentle female voice suitable for the user's gender and age, and adjusts the speaking speed to match the user's current anxiety.

[0168] The generated voice suggestions are played in real time through the user's headphones, and the user is advised to do deep breathing exercises in a gentle tone, while encouraging the user to sort out his emotions by writing an emotional diary. The user can not only hear the clear guidance, but also feel the soothing emotions conveyed in the voice, which greatly enhances the acceptance and effectiveness of the suggestions.

[0169] By converting text psychological counseling suggestions into voice form, this embodiment enables the system to interact with users in a more natural and friendly way, especially when users are not convenient to read or prefer auditory information, which improves user acceptance and satisfaction. Through emotionally enhanced TTS conversion, the system can adjust the emotional expression of voice output according to the user's current psychological state, making the suggestions more in line with the user's psychological needs and further improving the effect of psychological counseling. Multilingual support ensures that the system can provide high-quality voice advice services to users with different language backgrounds, expanding the application scope and market potential of the system.

[0170] In one embodiment, in the above S10, the user voice data is analyzed to identify the user's characteristic information, including:

[0171] d1, extracting the basic frequency components in the user's voice data, wherein the basic frequency components include high-frequency and pitch, and inferring the user's gender and age characteristics according to the basic frequency components;

[0172] d2, analyzing the pitch fluctuation of the user's voice data, identifying the user's emotional characteristics and optimizing gender characteristics and age characteristics.

[0173] In this embodiment, when analyzing the user's voice data, the system extracts the basic frequency components of the voice, including high-pitched frequencies and pitches, to infer the user's gender and age characteristics. In addition, the system also analyzes the pitch fluctuations in the voice to identify the user's emotional characteristics, and further optimizes the recognition of gender and age based on these emotional characteristics.

[0174] Treble frequencies usually refer to the relatively high fundamental frequency part of the speech signal. For human speech, the fundamental frequency (F0) is the frequency produced by the vibration of the vocal cords and reflects the pitch of the sound. Generally speaking:

[0175] The fundamental frequency in men is usually in the range of 85Hz to 180Hz.

[0176] Women's fundamental frequencies are usually in the range of 165Hz to 255Hz.

[0177] In children, the fundamental frequency is usually above 250 Hz and can sometimes reach 400 Hz or higher.

[0178] Therefore, in gender and age recognition, high frequencies are usually defined as the frequency portion greater than 165 Hz.

[0179] In gender recognition, fundamental frequency is one of the most significant distinguishing features. Specifically:

[0180] Male voice: Usually has a lower fundamental frequency. Due to the length and thickness of the vocal cords, the fundamental frequency of men is lower, and the frequency is mainly concentrated between 85Hz and 180Hz.

[0181] Female voice: Due to shorter and thinner vocal cords, females have a higher fundamental frequency, mainly concentrated between 165Hz and 255Hz. Therefore, if a higher fundamental frequency (such as over 165Hz) is detected in the analysis, the system is more likely to identify it as a female.

[0182] Fundamental frequency is also related to age:

[0183] Children: Because their vocal cords are not yet fully developed, children's fundamental frequencies are usually higher, usually above 250Hz.

[0184] Adults: The fundamental frequency decreases with age, especially after puberty, and then tends to stabilize.

[0185] Elderly people: During old age, relaxation of the vocal cords may cause a slight decrease in fundamental frequency, but the change is not as significant as before and after puberty.

[0186] Therefore, the fundamental frequency can reflect age characteristics to a certain extent:

[0187] If a fundamental frequency above 250 Hz is detected, the system may identify it as a child.

[0188] If the fundamental frequency is between 165 Hz and 250 Hz, it may be identified as a female or a late adolescent male.

[0189] The following frequency ranges can be preset for gender and age recognition:

[0190] 85Hz-180Hz: Identified as male (usually an adult male).

[0191] 165Hz-255Hz: Identified as female.

[0192] Above 250Hz: Identified as a child or prepubescent individual.

[0193] For example, in voice analysis, the system captures the fundamental frequency of the user's voice as 220Hz. According to the preset frequency range, this frequency falls between 165Hz and 255Hz, so the system initially determines that the user may be a female. Further analysis of the voice's pitch fluctuations, combined with the emotional characteristics it exhibits, the system confirms the user's gender and generates corresponding personalized psychological counseling suggestions based on this information.

[0194] Extraction of basic frequency components involves frequency analysis of the speech signal. By analyzing the fundamental frequency (F0) and harmonic components in the speech signal, the system can identify the user's pitch range. These frequency components can not only reflect the user's gender (usually men have lower fundamental frequencies, while women have higher fundamental frequencies), but also provide clues about the user's age (for example, changes in the vocal cords may cause changes in pitch as age increases).

[0195] Pitch fluctuation analysis focuses on the dynamic changes of speech, especially the fluctuations in intonation, speech speed, and volume. These factors are often closely related to emotions, and by analyzing these fluctuations, the system can identify the user's current emotional state, such as anxiety, anger, or calmness. In addition, by combining these emotional features, the system can further optimize the recognition of gender and age features to ensure the accuracy and reliability of the analysis results.

[0196] In one specific implementation, the method based on spectrum analysis and feature extraction includes:

[0197] Fundamental frequency extraction: The system performs spectrum analysis on the user's voice signal to extract the fundamental frequency and harmonic components. By analyzing these frequency components, the system can infer the user's gender characteristics (such as the different frequency distributions of men and women) and age characteristics (such as the different frequency distributions of children, adults and the elderly).

[0198] Pitch fluctuation analysis: The system further performs time series analysis on the pitch fluctuations in the voice signal. By analyzing the continuous changes in the voice, it identifies the user's emotional state. For example, anger may be manifested as a faster speaking speed and higher pitch fluctuations.

[0199] In another specific implementation, the method based on multimodal information fusion analysis includes:

[0200] Combining voice frequency with facial expressions: In addition to the frequency analysis of voice signals, the system can also combine the user's facial expression data to make more accurate inferences about gender and age characteristics. For example, by analyzing the user's voice frequency and facial features, the system can further optimize the prediction of age.

[0201] Emotion recognition optimization: By combining the fluctuations in pitch in speech and changes in facial expressions, the system can more accurately identify the user's emotional state and use this emotional information to optimize the recognition results for gender and age characteristics.

[0202] In other specific implementations, the methods based on dynamic adjustment and feedback mechanisms include:

[0203] Real-time adjustment of frequency analysis: When analyzing voice data, the system can adjust the parameters of frequency analysis in real time to adapt to the voice characteristics of different users. For example, when noise interference is detected in the voice signal, the system can improve the accuracy of gender and age recognition by adjusting the frequency analysis algorithm.

[0204] Feedback-based optimization: The system verifies and adjusts the recognition results of gender, age and emotional characteristics through the user feedback mechanism. Users can provide feedback on the system's recognition results, and the system further optimizes the frequency analysis and pitch fluctuation analysis based on this.

[0205] Example description:

[0206] When a user uses a voice assistant for psychological counseling, the system first extracts the fundamental frequency and harmonic components of the voice signal through voice frequency analysis. The system detects that the user's fundamental frequency is around 220Hz, which falls within the preset range of 165Hz to 255Hz, and preliminarily infers that the user is a female. When further analyzing the pitch fluctuations, the system finds that there are obvious emotional fluctuations in the user's voice, manifested as faster speech speed and significantly higher pitch, indicating that the user may be in an anxious state.

[0207] Based on these preset high-frequency characteristics and detection results, the system generates a personalized psychological counseling suggestion, suggesting that the user try deep breathing exercises and providing some specific relaxation techniques to relieve their anxiety. The system also feeds back the suggestions to the user through voice, and encourages the user to feedback the actual effect of these suggestions later, so that the system can further optimize the suggestion generation process in the future.

[0208] By extracting the basic frequency components in the voice and analyzing the pitch fluctuations, the system can more accurately identify the user's gender, age and emotional characteristics, ensuring that the generated psychological counseling suggestions meet the user's actual needs. It has the ability of real-time analysis and dynamic adjustment, and can instantly adjust the analysis parameters according to the changes in the voice signal to improve the accuracy and reliability of the recognition results. Combining voice frequency with other modal data (such as facial expressions), the system can further optimize the recognition of gender, age and emotional characteristics, and provide more comprehensive and personalized psychological counseling services.

[0209] In one embodiment, in the above S30, the text data and the prompt word template are input into the target psychological counseling model to generate personalized text psychological counseling suggestions, including:

[0210] e1, obtaining the user's historical consultation records, which include the user's historical text data, historical text psychological consultation suggestions and user feedback information;

[0211] e2, analyzing the historical consultation records to generate user behavior and emotion pattern analysis results, combining the text data with the prompt word template to input the target psychological consultation model, and generating the personalized text psychological consultation suggestions.

[0212] In this embodiment, when generating personalized text psychological counseling suggestions, the system not only relies on the current text data and prompt word template, but also obtains the user's historical counseling records. The historical records contain the user's past text data, the generated text psychological counseling suggestions, and the user's feedback information. After these historical records are analyzed by the system, the user's behavior and emotional pattern analysis results can be generated. Combined with these analysis results, they are input into the target psychological counseling model together with the current text data and prompt word template to generate more accurate and personalized psychological counseling suggestions.

[0213] By deeply analyzing the user's historical data, the system can identify the user's behavior patterns and emotional trends. For example, the system can detect the user's recurring emotional reactions or behavioral habits in similar situations, and apply these patterns to the current consulting advice generation to improve the accuracy and pertinence of the advice.

[0214] User behavior and emotional pattern analysis involves using machine learning algorithms to perform cluster analysis on user history records, plot emotional curves, etc., so as to derive the user's emotional patterns (such as the pattern of emotional fluctuations, preferences for coping strategies) and behavioral patterns (such as the high adoption rate of certain specific suggestions). In this way, it is possible to better predict the user's needs in the current situation and generate suggestions that are more in line with the user's long-term psychological state.

[0215] In one specific implementation, the method of analyzing historical records based on time series includes:

[0216] Historical record acquisition: The system acquires the user's historical consultation records, including text data, generated suggestions, and user feedback. The record data is arranged in chronological order to form time series data.

[0217] Behavioral pattern analysis: The system uses time series analysis algorithms to detect users’ emotional trends and behavioral patterns over different time periods. For example, the system may identify that users show higher anxiety levels on stressful workdays and prefer certain mitigation strategies.

[0218] Suggestion generation: Combining the current text data and prompt word templates, the system inputs the user's behavior and emotional pattern analysis results into the target psychological counseling model to generate highly personalized text psychological counseling suggestions.

[0219] In another specific implementation, the behavior and emotion pattern recognition based on machine learning includes:

[0220] Historical record data classification: The system uses a clustering algorithm to classify historical record data, grouping similar emotional reactions and behavior patterns together. These classification results help the system better understand the user's emotional patterns.

[0221] Emotional curve drawing: By analyzing the emotional tags in historical data, the system draws the user's emotional curve to reflect the law of user emotional changes. The system uses this curve to predict the user's current emotional state and generate corresponding suggestions.

[0222] Model optimization generation: Based on the input of current text data and prompt word templates, the system combines the analysis results of historical behavior and emotional patterns to further optimize the generated psychological counseling suggestions to make them more in line with the user's long-term needs.

[0223] In other specific implementations, the method based on personalized feedback loop and adaptive optimization includes:

[0224] Dynamic feedback and record update: The system not only analyzes historical data, but also records the user's response in the current consultation in real time. The system can dynamically update the user's behavior and emotional patterns based on the latest feedback.

[0225] Personalized analysis and generation: The system quickly analyzes newly recorded data, integrates it with historical records, and generates more personalized psychological counseling suggestions. This cycle enables the system to further optimize the accuracy of the suggestions after each consultation.

[0226] Example description:

[0227] The user has repeatedly expressed anxiety about social situations in past consultations and has expressed positive feedback on deep breathing exercises in multiple recommendations. The system captures this history and identifies through time series analysis that the user's mood swings are particularly significant on Monday mornings and that deep breathing exercises are most effective during these time periods.

[0228] In the current consultation, the user again expressed anxiety. The system combined the high acceptance of deep breathing exercises in the historical records with the current text data and the prompt word template to generate a personalized suggestion that the user should practice deep breathing exercises for a few minutes before social situations. At the same time, the system provides a psychological tip to help the user stay calm in social situations. This suggestion not only combines the user's current needs, but also takes into account their long-term emotional patterns and historical feedback.

[0229] This embodiment can generate highly personalized psychological counseling suggestions by combining historical records and current data, ensuring that the suggestions not only meet the user's current needs, but also fit the user's long-term behavior and emotional patterns. Through continuous feedback and updating of historical records, the generated suggestions can be continuously optimized to improve the accuracy of psychological counseling and user satisfaction. Through in-depth analysis of historical data, the user's behavior and emotional patterns can be understood, thereby providing more targeted psychological counseling services in different situations.

[0230] In one embodiment, in the above S30, after generating personalized text psychological counseling suggestions, the following is further included:

[0231] f1, collect data statistics on the user's consultation records within a preset period, including the content of each consultation suggestion, user feedback, and emotional change data;

[0232] f2, generating a consulting report including emotional change trends and consulting effect evaluation based on statistical data, and generating personalized mental health management suggestions based on the consulting report;

[0233] f3, providing the personalized mental health management suggestions to the user in the form of text and / or voice.

[0234] In this embodiment, after generating personalized text psychological counseling suggestions, the system will also collect data statistics on the user's counseling records within a preset period. This statistical process includes the content of each consultation suggestion, user feedback, and emotional change data. By analyzing these data, the system can generate a counseling report containing emotional change trends and counseling effect evaluation. Based on the counseling report, the system further generates personalized mental health management suggestions and feeds them back to the user in text or voice form. This process enables the system to continuously monitor and optimize the user's mental health management over a longer period of time.

[0235] Regularly summarize and analyze the user's consultation records, including the suggestions received by the user each time, the user's feedback on these suggestions (such as satisfaction, effectiveness), and the user's emotional changes (such as mood swings, improvement or deterioration of emotional state). These statistical data provide a basis for subsequent emotional change trend analysis and consultation effect evaluation.

[0236] By analyzing statistical data, we can identify the trend of users’ emotions over a certain period and evaluate the actual effect of each consultation suggestion. The report will show the curve of users’ emotional state changing over time, as well as the effect of each suggestion in different situations.

[0237] Generate personalized mental health management recommendations based on the generated consultation report. These recommendations may include emotion management strategies, mental health maintenance plans, and recommended frequencies for future consultations.

[0238] In one specific implementation, the method based on periodic data statistics and report generation includes:

[0239] Data statistics: At the end of a preset period (such as weekly or monthly), the system automatically counts all consultation records of the user during that period, including each psychological consultation suggestion generated, user feedback information, and the user's emotional changes before and after the consultation.

[0240] Report generation: The system generates a detailed consulting report based on statistical data, including user emotional change trend charts, recommendation adoption status, effect evaluation, etc. The report is presented in the form of charts and text to help users understand their mental health status during the cycle.

[0241] In another specific implementation, the method based on dynamic trend analysis and personalized management suggestions includes:

[0242] Emotional change trend analysis: The system identifies the periodicity and trend of user emotional changes through time series analysis. For example, the system can identify that users have greater emotional fluctuations on Monday mornings and show fatigue or anxiety at the end of the month. Through these analyses, the system can predict possible emotional fluctuations in the future.

[0243] Management suggestion generation: Based on the trend of emotional changes and the evaluation of consultation effects, the system generates personalized mental health management suggestions, such as suggesting that users increase the frequency of consultations during a specific time period, or try new methods of emotional regulation.

[0244] In other specific implementations, the method based on combining comprehensive feedback with multimodal data includes:

[0245] Multimodal data statistics: In addition to text and voice data, the system can also combine user behavior data (such as activity level, sleep quality) for comprehensive statistics and analysis to provide more comprehensive consulting reports.

[0246] Comprehensive feedback and optimization: After providing personalized management suggestions, the system allows users to provide feedback on the effectiveness of the suggestions. Based on these feedbacks, the system continuously optimizes subsequent management suggestions to ensure that users can get the best mental health support.

[0247] Example description:

[0248] The user received psychological counseling through the system for one month. After each counseling session, the system recorded the generated suggestions, the user's feedback, and the user's emotional changes. At the end of the month, the system conducted a statistical analysis of these data and generated a detailed report. The report showed that the user's emotional state was relatively stable at the beginning of the month, but showed significant anxiety in the middle and end of the month, especially during periods of high work pressure.

[0249] Based on the trend of emotional changes in the report, the system recommends that users increase the frequency of deep breathing exercises during high-stress periods in the next month and meditate before bed. In addition, the system recommends that users keep an emotional journal once a week to better manage their emotions. These personalized mental health management suggestions are fed back to users in the form of voice and text to help users better cope with emotional fluctuations in the future.

[0250] This embodiment generates a user's emotional change trend report through regular data statistics and analysis, and provides users with mental health management suggestions based on long-term observation. Personalized management suggestions generated based on consulting reports can help users better cope with emotional fluctuations and psychological challenges, and provide more targeted mental health support. The introduction of user feedback enables the system to continuously optimize management suggestions, ensure their effectiveness and practicality, and thus improve users' mental health level and satisfaction.

[0251] The present invention also provides an AI-based psychological consultation device, referring to Figure 2 , Figure 2 The following is a functional module diagram of a preferred embodiment of an AI-based psychological counseling device of the present invention. The AI-based psychological counseling device includes:

[0252] A voice processing module, which obtains user voice data, converts the user voice data into text data, and analyzes the user voice data to identify user feature information, including gender, age, and emotional features;

[0253] A feature extraction module generates a prompt word template according to the text data and the feature information, wherein the prompt word template is used to guide the target psychological counseling model to generate psychological counseling suggestions;

[0254] The consultation generation module inputs the text data and the prompt word template into the target psychological consultation model to generate personalized text psychological consultation suggestions.

[0255] The specific implementation of the AI-based psychological counseling device of the present invention is basically the same as the above-mentioned embodiments of the AI-based psychological counseling method, and will not be repeated here.

[0256] The present invention also provides an AI-based psychological consultation device, such as Figure 3 As shown, the AI-based psychological counseling device may include: a processor 1001, such as a CPU, a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005. Among them, the communication bus 1002 is used to realize the connection and communication between these components. The user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the optional user interface 1003 may also include a standard wired interface and a wireless interface. The network interface 1004 may optionally include a standard wired interface and a wireless interface (such as a WI-FI interface). The memory 1005 may be a high-speed RAM memory, or a stable memory (non-volatile memory), such as a disk memory. The memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0257] Those skilled in the art will understand that Figure 3 The hardware structure of the AI-based psychological counseling device shown in the figure does not constitute a limitation of the AI-based psychological counseling device, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.

[0258] like Figure 3As shown, the memory 1005 as a storage medium may include an operating system, a network communication module, a user interface module, and an AI-based psychological counseling program. The operating system is a program for managing and controlling AI-based psychological counseling equipment and software resources, and supports the operation of the network communication module, the user interface module, the AI-based psychological counseling program, and other programs or software; the network communication module is used to manage and control the network interface 1004; and the user interface module is used to manage and control the user interface 1003.

[0259] exist Figure 3 In the hardware structure of the AI-based psychological counseling device shown, the network interface 1004 is mainly used to connect to the background server and communicate data with the background server; the user interface 1003 is mainly used to connect to the client and communicate data with the client; the processor 1001 can call the AI-based psychological counseling program stored in the memory 1005 and perform the same operations as the AI-based psychological counseling method.

[0260] The specific implementation of the AI-based psychological counseling device of the present invention is basically the same as the above-mentioned embodiments of the AI-based psychological counseling method, and will not be repeated here.

[0261] In addition, an embodiment of the present invention further proposes a computer storage medium on which an AI-based psychological counseling program is stored. When the AI-based psychological counseling program is executed by a processor, the steps of the AI-based psychological counseling method described above are implemented.

[0262] The specific implementation methods of the storage medium of the present invention are basically the same as the above-mentioned embodiments of the AI-based psychological counseling method, and will not be repeated here.

[0263] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the enlightenment of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims. All equivalent structures or equivalent process changes made using the contents of the specification and drawings of the present invention, or directly or indirectly used in other related technical fields, are protected by the present invention.

[0264] It should be noted that if software tools or components other than those of our company appear in the embodiments of the present application, they are only used for illustrative purposes and do not represent actual use.

Claims

1. A psychological consultation method based on AI, characterized in that: The following steps are involved: Acquire user voice data, convert the user voice data into text data, and analyze the user voice data to identify user feature information, wherein the feature information includes gender, age, and emotional features; Generate a prompt word template according to the text data and the feature information, wherein the prompt word template is used to guide the target psychological counseling model to generate psychological counseling suggestions; The text data and the prompt word template are input into a target psychological counseling model to generate personalized text psychological counseling suggestions.

2. The AI-based psychological counseling method according to claim 1, characterized in that: Before inputting the text data and the prompt word template into the target psychological counseling model, the method further includes: Obtain psychological data and perform structured processing to generate knowledge base documents that can be used by language models; The knowledge base document is input into the artificial intelligence framework and combined with the preset language model to generate the target psychological counseling model.

3. The AI-based psychological consultation method according to claim 2, characterized in that: After generating the target psychological counseling model, it also includes: Record the user's emotional changes, user feedback information, and the actual effect of consulting suggestions for each consultation; Generate new knowledge entries based on the recorded information, and use the new knowledge entries to update the knowledge base document; Use the updated knowledge base document to adjust the target psychological counseling model.

4. The AI-based psychological counseling method according to claim 1, characterized in that: After generating personalized text psychological counseling suggestions, it also includes: The text psychological counseling suggestions are converted into voice psychological counseling suggestions through text-to-speech technology, and the voice psychological counseling suggestions are used to provide feedback on the user's psychological counseling; Adjusting the timbre and tone of the voice psychological counseling suggestion according to the characteristic information of the user; The voice psychological counseling suggestions are played to the user via a playback device.

5. The AI-based psychological consultation method according to claim 1, characterized in that: And analyze the user voice data to identify the user's characteristic information, including: Extracting basic frequency components from the user's voice data, the basic frequency components including high-frequency and pitch, and inferring the user's gender and age characteristics based on the basic frequency components; The pitch fluctuation of the user's voice data is analyzed to identify the user's emotional characteristics and optimize gender characteristics and age characteristics.

6. The AI-based psychological consultation method according to claim 1, characterized in that: Inputting the text data and the prompt word template into the target psychological counseling model to generate personalized text psychological counseling suggestions, including: Obtaining the user's historical consultation records, wherein the historical consultation records include the user's historical text data, historical text psychological consultation suggestions and user feedback information; The historical consultation records are analyzed to generate user behavior and emotion pattern analysis results, and the text data is combined with the prompt word template to input the target psychological consultation model to generate the personalized text psychological consultation suggestions.

7. The AI-based psychological counseling method as claimed in claim 1, characterized in that: After generating personalized text psychological counseling suggestions, it also includes: Collect data statistics on the user's consultation records within a preset period, including the content of each consultation suggestion, user feedback, and emotional change data; Generate a consulting report containing emotional change trends and consulting effect evaluation based on statistical data, and generate personalized mental health management suggestions based on the consulting report; The personalized mental health management suggestions are fed back to the user in text and / or voice form.

8. An AI-based psychological counseling device, characterized in that: The AI-based psychological counseling device includes: A voice processing module, which obtains user voice data, converts the user voice data into text data, and analyzes the user voice data to identify user feature information, including gender, age, and emotional features; A feature extraction module generates a prompt word template according to the text data and the feature information, wherein the prompt word template is used to guide the target psychological counseling model to generate psychological counseling suggestions; The consultation generation module inputs the text data and the prompt word template into the target psychological consultation model to generate personalized text psychological consultation suggestions.

9. An AI-based psychological counseling device, characterized in that: The AI-based psychological counseling device includes a memory, a processor, and an AI-based psychological counseling program stored in the memory and executable on the processor. When the AI-based psychological counseling program is executed by the processor, the steps of the AI-based psychological counseling method as described in any one of claims 1 to 7 are implemented.

10. A computer storage medium, characterized in that: The storage medium stores an AI-based psychological counseling program, which, when executed by a processor, implements the steps of the AI-based psychological counseling method as described in any one of claims 1-7.

Images