Mental health monitoring and early warning method and system based on big data
Through multi-source data collection and analysis, a mental health warning model is built, which solves the timeliness and personalized problems of traditional mental health monitoring methods, realizes more accurate mental health status monitoring and personalized intervention plans, and improves mental health Accuracy and effectiveness of early warning.
Patent Information
- Application Number
- CN202411047830.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-01
- Publication Date
- 2024-11-22
AI Technical Summary
Traditional mental health monitoring and early warning methods are limited by manpower, time and region, making it difficult to timely and effectively detect and intervene in multiple psychological problems in young groups such as college students. Moreover, the mental health monitoring and early warning methods based on big data have low data quality and data. The problem of insufficient accuracy of the treatment and analysis methods and insufficient personalization of the early warning model.
Collect multi-source data through multiple collection channels, perform data preprocessing and analysis, identify mental health risk factors and abnormal behavior patterns, build a mental health warning model, monitor the user's mental health status in real time and issue warning information, and evaluate and Push personalized intervention plans.
It has achieved a more accurate judgment of individual mental health status, reduced the probability of false alarms and missed reports, timely discovered abnormal situations, provided the possibility for early intervention, and formulated a personalized early warning intervention plan, which has improved the effectiveness of early warning intervention.
Smart Images

Figure CN119007998A_ABST
Abstract
Description
Technical Field
[0001] The embodiments of the present invention relate to the field of psychological health early warning technology, and in particular to a psychological health monitoring and early warning method and system based on big data. Background Art
[0002] In today's high-pressure social environment, mental health issues are becoming increasingly prominent, especially among young groups such as college students who face multiple pressures and are prone to various psychological problems. Traditional mental health monitoring and early warning methods are limited by manpower, time and location, making it difficult to detect and intervene in a timely and effective manner. The rapid development of big data technology has provided new possibilities for mental health monitoring and early warning. Summary of the invention
[0003] The embodiment of the present invention provides a big data-based mental health monitoring and early warning method and system, which can comprehensively analyze multi-source data, monitor and accurately judge the individual's mental health status in real time, provide personalized early warning intervention plans, and improve the effect of early warning intervention.
[0004] To achieve the above objectives, in a first aspect, the present invention provides a mental health monitoring and early warning method based on big data, comprising:
[0005] Collect user data through multiple collection channels to generate multi-source data;
[0006] Preprocessing the multi-source data to generate a preprocessed multi-source data set;
[0007] Analyzing the preprocessed multi-source data to identify mental health risk factors and abnormal behavior patterns;
[0008] Based on the mental health risk factors and the abnormal behavior patterns, construct a mental health early warning model;
[0009] Based on the mental health early warning model, real-time monitoring of the user's mental health status data is performed, wherein when the user's mental health status data is monitored to be abnormal, an early warning message is issued;
[0010] Based on the warning information, an assessment of mental health status is performed to generate an assessment result;
[0011] Based on the assessment results, a personalized mental health early warning intervention plan is generated and pushed.
[0012] In one embodiment of the present invention, collecting user data through multiple collection channels to generate multi-source data includes:
[0013] Collecting user data through a first collection channel, a second collection channel, and a third collection channel to obtain first user data, second user data, and third user data;
[0014] The multi-source data is generated based on the first user data, the second user data, and the third user data.
[0015] In one embodiment of the present invention, preprocessing the multi-source data to generate a preprocessed multi-source data set includes:
[0016] Performing data cleaning, data standardization and normalization, and data integration processing on the multi-source data to generate integrated multi-source data;
[0017] Annotating the integrated multi-source data to generate normal psychological state data and abnormal psychological state data;
[0018] The preprocessed multi-source data set is generated based on the normal psychological state data and the abnormal psychological state data.
[0019] In one embodiment of the present invention, the analysis based on preprocessing multi-source data to identify mental health risk factors and abnormal behavior patterns includes:
[0020] Based on the preprocessed multi-source data set, obtaining psychological characteristics, behavioral characteristics and social characteristics;
[0021] Based on the psychological characteristics, the behavioral characteristics and the social characteristics, a preset method is used to identify the mental health risk factors and the abnormal behavior patterns.
[0022] In one embodiment of the present invention, the real-time monitoring of the user's mental health status data based on the mental health early warning model, wherein when the user's mental health status data is monitored to be abnormal, issuing early warning information includes:
[0023] Based on the mental health early warning model, real-time monitoring of the user's mental health status data is performed, and monitoring results are generated;
[0024] When the monitoring result exceeds a preset threshold, it is determined that an abnormal situation has occurred and an early warning message is issued.
[0025] In one embodiment of the present invention, the mental health status is evaluated based on the early warning information, and the evaluation result is generated, including:
[0026] Based on the warning information, identifying the warning factors causing the warning and the severity of the warning;
[0027] Acquire multi-dimensional information, and based on the multi-dimensional information, the warning factors and the warning severity, evaluate the mental health status and generate an evaluation result.
[0028] In a second aspect, the present invention provides a mental health monitoring and early warning system based on big data, comprising: a first generation module, a second generation module, an identification module, a construction module, an issuance module, a third generation module and a fourth generation module. The first generation module is used to collect user data through multiple collection channels to generate multi-source data. The second generation module is used to pre-process the multi-source data to generate a pre-processed multi-source data set. The identification module is used to analyze based on the pre-processed multi-source data to identify mental health risk factors and abnormal behavior patterns. The construction module is used to construct a mental health early warning model based on the mental health risk factors and the abnormal behavior patterns. The issuance module is used to monitor the user's mental health status data in real time based on the mental health early warning model, wherein when the user's mental health status data is monitored to be abnormal, an early warning message is issued. The third generation module is used to evaluate the mental health status based on the early warning information and generate an evaluation result. The fourth generation module is used to generate and push a personalized mental health early warning intervention plan based on the evaluation result.
[0029] In one embodiment of the present invention, the first generating module includes: an obtaining unit and a first generating unit. The obtaining unit is used to collect user data through a first collecting channel, a second collecting channel and a third collecting channel to obtain first user data, second user data and third user data. The first generating unit is used to generate the multi-source data based on the first user data, the second user data and the third user data.
[0030] In a third aspect, the present invention provides an electronic device, comprising:
[0031] at least one processor; and
[0032] a memory communicatively coupled to the at least one processor;
[0033] Among them, the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the big data-based mental health monitoring and early warning method as described above.
[0034] In a fourth aspect, the present invention provides a computer-readable storage medium, comprising a computer program and instructions. When the computer program or the instructions are run on a computer, the computer executes the big data-based mental health monitoring and early warning method as described above.
[0035] Compared with the prior art, the big data-based mental health monitoring and early warning method and system according to the present invention has the following beneficial effects:
[0036] 1. Ability to comprehensively analyze multi-source data, more accurately judge the mental health status of individuals, and reduce the probability of false positives and false negatives;
[0037] 2. It can monitor the mental health status of individuals in real time, detect abnormal situations in time, and provide the possibility for early intervention;
[0038] 3. Based on the individual's specific situation and data analysis results, a personalized mental health early warning intervention plan is formulated to improve the effectiveness of early warning intervention. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1 It is a flowchart of a method for monitoring and early warning of mental health based on big data in Embodiment 1 of the present invention;
[0040] Figure 2 It is a structural diagram of a big data-based mental health monitoring and early warning system in Embodiment 2 of the present invention;
[0041] Figure 3 It is a structural schematic diagram of an electronic device in Embodiment 3 of the present invention. DETAILED DESCRIPTION
[0042] The embodiments of the present invention are further described in detail below in conjunction with the accompanying drawings and embodiments. It is to be understood that the specific embodiments described herein are only used to explain the embodiments of the present invention, rather than to limit the embodiments of the present invention. It is also necessary to explain that, for ease of description, only parts related to the embodiments of the present invention are shown in the accompanying drawings, rather than all structures.
[0043] To facilitate understanding, the main implementation concepts of the embodiments of the present invention are first briefly described.
[0044] In today's society, with the accelerated pace of life and intensified social competition, mental health issues have gradually become an important factor affecting the quality of life of individuals and social stability. In particular, young groups such as college students are more likely to have psychological problems such as anxiety and depression when facing multiple pressures such as academic, employment, and interpersonal relationships. However, traditional mental health monitoring and early warning methods mainly rely on questionnaires, face-to-face consultations and other means. Due to the limitations of manpower, time and location, it is difficult to achieve real-time monitoring and effective intervention for a large number of individuals.
[0045] In recent years, the rapid development of big data technology has provided new ideas and methods for mental health monitoring and early warning. Big data technology can process and analyze massive and diverse data to extract valuable information and patterns. In the field of mental health, by collecting and analyzing multi-source data such as individual behavioral data, social data, and physiological data, we can have a more comprehensive understanding of the individual's mental state and changing trends, thereby achieving more accurate mental health monitoring and early warning.
[0046] However, there are still some challenges in the current mental health monitoring and early warning methods based on big data, such as low data quality, inaccurate data processing and analysis methods, and insufficient personalization of early warning models.
[0047] By discovering the above-mentioned defects in the prior art, the inventors have provided a mental health monitoring and early warning method and system based on big data. By comprehensively analyzing multi-source data, real-time monitoring of individual mental health status, and formulating personalized early warning intervention plans, the accuracy and effectiveness of mental health monitoring and early warning are improved, thereby safeguarding the mental health of individuals.
[0048] Embodiment 1
[0049] Figure 1 is a flow chart of a method for monitoring and early warning of mental health based on big data in Embodiment 1 of the present invention, such as Figure 1 As shown, embodiment 1 provides a mental health monitoring and early warning method based on big data, including:
[0050] Step S100, collecting user data through multiple collection channels to generate multi-source data;
[0051] Specifically, step S100 is the starting step of this mental health monitoring and early warning method, the core of which is to collect user data extensively through multiple channels to form comprehensive, multi-dimensional multi-source data. Specifically, this step includes collecting relevant data of users through different collection channels (such as social media, online surveys, physiological monitoring equipment, etc.), and then generating multi-source data containing various information of users. The implementation of this step provides a rich and comprehensive data foundation for the subsequent mental health status analysis and early warning model construction.
[0052] Step S200, preprocessing the multi-source data to generate a preprocessed multi-source data set;
[0053] Specifically, first of all, it is necessary to clean the multi-source data to remove noise, outliers and duplicate data to ensure the purity and consistency of the data. Then, data standardization and normalization are performed to convert data from different sources and scales into a unified numerical range, eliminate the dimensional differences between the data, and make the subsequent analysis and comparison more fair and objective. After that, data integration is performed to organize and integrate the cleaned and standardized data according to a specific structure and format to form an integrated multi-source data set, which facilitates the subsequent feature extraction and model construction. Finally, the integrated multi-source data is labeled and divided into normal psychological state data and abnormal psychological state data according to the characteristics and labels of the data, so as to facilitate the subsequent identification of mental health risk factors and abnormal behavior patterns. Through this series of preprocessing operations, the quality and availability of multi-source data can be effectively improved, laying a solid foundation for the subsequent mental health status analysis and the construction of early warning models. At the same time, this step is also an indispensable part of the entire mental health monitoring and early warning method. It ensures the accuracy and effectiveness of subsequent analysis, improves the sensitivity and specificity of the early warning model, and protects the individual's mental health.
[0054] Step S300, analyzing the pre-processed multi-source data to identify mental health risk factors and abnormal behavior patterns;
[0055] Specifically, step S300 is a key link for in-depth analysis based on the pre-processed multi-source data set to identify mental health risk factors and abnormal behavior patterns. The core of this step is to extract valuable features and patterns from multi-source data, and then reveal risk factors and abnormal behaviors related to mental health status. Specifically, first, multi-dimensional information such as psychological characteristics, behavioral characteristics and social characteristics will be obtained from the pre-processed multi-source data set, which can fully reflect the individual's mental health status and behavior patterns. Then, using preset methods and techniques, such as machine learning algorithms, statistical analysis, etc., these features are deeply analyzed and mined to identify risk factors and abnormal behavior patterns related to mental health problems. These risk factors and abnormal behavior patterns may include, for example, individual mood swings, changes in social behavior, abnormalities in physiological indicators, etc., which can serve as early warning signals to indicate possible mental health problems in individuals. Through the analysis and identification of step S300, the individual's mental health status and behavior patterns can be understood more accurately, providing strong support for the subsequent construction of early warning models and the formulation of personalized intervention plans. At the same time, this step is also the core link in the entire mental health monitoring and early warning method. It is directly related to the accuracy and effectiveness of the early warning model, and is of great significance for improving the individual's mental health level and preventing mental illness.
[0056] Step S400, constructing a mental health early warning model based on the mental health risk factors and the abnormal behavior patterns;
[0057] Specifically, step S400 is a key stage in building a mental health early warning model. Based on the mental health risk factors and abnormal behavior patterns extracted in step S300, it uses advanced machine learning algorithms or statistical models to train and verify the data to establish an early warning system that can accurately predict individual mental health risks. The core of this stage is the selection and optimization of the model to ensure that the model can capture subtle changes in the data while avoiding overfitting, thereby improving the accuracy and generalization of predictions. Through repeated iterations and adjustment of model parameters, step S400 aims to build a highly sensitive and specific early warning model to achieve real-time monitoring and early warning of individual mental health status, and provide a scientific basis for timely intervention measures. The successful implementation of this step is of great significance to improving the efficiency and quality of mental health services and promoting the overall well-being of individuals and society.
[0058] In step S400, building a mental health early warning model is the core of the entire mental health monitoring and early warning method. This step is based on the mental health risk factors and abnormal behavior patterns identified in step S300, and uses advanced machine learning algorithms or statistical models to establish a model that can accurately predict individual mental health risks. Specifically:
[0059] Choosing a suitable machine learning algorithm is the key to building a mental health early warning model. This embodiment uses the long short-term memory network (LSTM) in the deep learning model to build an early warning model. Before building the model, it is necessary to ensure that the preprocessed multi-source data set (including mental health risk factors and abnormal behavior patterns) is complete and high-quality. These data should be divided into training sets, validation sets, and test sets for use in the model training, validation, and testing stages.
[0060] Feature selection is an important step in model building. It aims to select the most influential features for the prediction target from the preprocessed data set. In the mental health warning model, these features may include the user's psychological characteristics (such as emotional stability, stress level), behavioral characteristics (such as social activity, living habits), physiological characteristics (such as heart rate, sleep quality), etc. Through feature selection, the complexity of the model can be reduced and the generalization ability of the model can be improved.
[0061] Use LSTM network to build early warning model. LSTM is particularly suitable for processing time series data because it can capture long-term dependencies. Define the parameters of LSTM layer such as input size, number of hidden units, number of layers, etc. Add necessary layers, such as fully connected layers (Dense layers), to output the final early warning results. Train the LSTM model using training set data. Calculate the error between the predicted value and the actual value through forward propagation, and then adjust the network weights using the backpropagation algorithm to minimize the error. Monitor the loss value and accuracy during training to prevent overfitting. Use validation set data to evaluate the generalization ability of the model. Adjust model parameters, optimizer settings, or feature selection based on the validation results to improve model performance. Use test set data to evaluate the performance of the final model. If the model performance meets the requirements, deploy it to actual applications for real-time early warning. For example, it is necessary to build a mental health early warning model based on LSTM to monitor the user's mental health status in real time and issue early warnings when potential risks occur. Collect users' psychological test results (such as depression scale scores), social media activity records (such as posting frequency, emotional expression), and physiological indicators monitored by wearable devices (such as heart rate variability, sleep quality), etc. Extract statistical features (such as mean score and standard deviation) from psychological test results. Extract text features and time series features (such as emotional word frequency and posting time interval) from social media activity records. Extract time series features (such as daily trend of heart rate variability) from physiological indicators. Define an LSTM model, where the input layer accepts time series features (such as time series of physiological indicators) and captures long-term dependencies through multiple LSTM layers. Add a fully connected layer (Dense layer) as the output layer to output warning results (such as normal, mild risk, moderate risk, etc.). Train the LSTM model using the labeled training set data. Use the validation set data to monitor the generalization ability of the model and adjust the model parameters to optimize performance. Use the test set data to evaluate the accuracy, recall rate, and F1 score of the final model. If the model performance meets the requirements, deploy it to the mental health monitoring system to monitor the user's mental health status in real time and issue warning information when potential risks are detected. Through the above steps, an effective mental health warning model can be built using the LSTM network to provide users with timely mental health support.
[0062] Step S500, monitoring the user's mental health status data in real time based on the mental health early warning model, wherein when the user's mental health status data is monitored to be abnormal, issuing an early warning message;
[0063] Specifically, step S500 is the final link in the mental health monitoring and early warning method, and is also a key step in converting the results of the previous analysis and modeling into practical applications. This stage mainly focuses on the generation and transmission of early warning signals, as well as the formulation and implementation of personalized intervention plans. Specifically, step S500 will evaluate the individual's mental health status in real time according to the prediction results of the early warning model, and generate early warning signals in a timely manner when potential risks are identified. These signals will be transmitted to the individual and his designated supporters or professionals through appropriate means (such as mobile phone application notifications, emails, etc.) so that necessary intervention measures can be taken in a timely manner. At the same time, step S500 will also formulate personalized intervention plans based on the individual's specific risk factors and abnormal behavior patterns. These plans may include psychological counseling, cognitive behavioral therapy, relaxation training and other forms of mental health services, aiming to help individuals effectively cope with mental health problems and improve psychological resilience. Through the implementation of step S500, the mental health monitoring and early warning method not only realizes real-time monitoring and early warning of the individual's mental health status, but also provides targeted intervention support, thereby forming a complete mental health service system. The effective implementation of this step is of vital importance for promoting individual mental health and preventing the occurrence and development of mental illness.
[0064] Step S600, based on the warning information, evaluating the mental health status and generating an evaluation result;
[0065] Specifically, step S600 is a process of further evaluating the user's mental health based on the early warning information. The purpose of this step is to more accurately understand the user's current mental health so as to provide more targeted suggestions for the subsequent early warning intervention plan. Specifically, the early warning factors and the severity of the early warning that lead to the early warning are first identified. Early warning factors may include, for example, certain behavioral patterns, emotional states, social activities, etc. of the user, which may indicate that there is a potential risk in the user's mental health. The severity of the early warning is used to assess the magnitude of this risk in order to determine whether immediate intervention measures are needed. Next, multi-dimensional information is obtained, which may include, for example, the user's personal background, historical mental health records, current living environment, etc. This information is essential for a comprehensive assessment of the user's mental health because it can provide more context and details about the user's mental health. Finally, based on the multi-dimensional information, early warning factors, and the severity of the early warning, an assessment of the mental health state is performed, and an assessment result is generated. This assessment result will serve as an important basis for the subsequent formulation of a personalized mental health early warning intervention plan. Through this step, the user's mental health can be understood more accurately, and more effective help and support can be provided to the user.
[0066] Step S700, generating and pushing a personalized mental health early warning intervention plan based on the evaluation results;
[0067] Specifically, step S700 is a process of generating and pushing a personalized mental health early warning intervention plan based on the evaluation results. The core of this step is to tailor a set of suitable early warning intervention plans for users based on the mental health status evaluation results obtained in the previous step. The plan may, for example, include advising users to undergo specific psychological counseling or treatment, recommending suitable mental health resources or applications, or providing specific strategies for coping with current mental health challenges. Through personalized intervention plans, the unique needs of users can be better met, and the effectiveness and success rate of early warning intervention can be improved. In addition, step S700 may also involve further communication with users to ensure that users understand and accept the proposed intervention plan, and provide necessary support and guidance to help users implement the plan smoothly. Ultimately, the goal of this step is to help users effectively cope with mental health challenges, improve users' mental health levels, and prevent potential psychological problems. Through personalized early warning intervention plans, users can be provided with more intimate and effective mental health services to promote the overall well-being and quality of life of users.
[0068] In this embodiment, step S100 includes:
[0069] Step S101, collecting user data through a first collection channel, a second collection channel, and a third collection channel to obtain first user data, second user data, and third user data;
[0070] Specifically, through professional psychological assessment tools and methods, a comprehensive and detailed examination of the user's psychological state is carried out to determine whether there are potential psychological problems or risks. This step is the basis for the formulation and implementation of subsequent intervention measures, and is crucial to ensuring the pertinence and effectiveness of the intervention. Through scientific and objective evaluation, timely and accurate psychological health feedback can be provided to individuals, laying a solid foundation for the smooth progress of subsequent steps. Among them, the first collection channel can be, for example, a psychological measurement tool (such as a questionnaire, a scale, etc.), the second collection channel can be, for example, a wearable device (such as a smart bracelet, a heart rate monitor, etc.), and the second collection channel can be, for example, social media (such as Weibo, WeChat, Douyin, etc.). Among them, psychological measurement tools are used to conduct regular or irregular psychological state assessments to ensure the timeliness and accuracy of the data. Among them, the user's public information is captured from social media, such as posting content, interaction status, etc., to analyze the user's emotional state and social behavior pattern. Among them, the wearable device can collect the user's physiological data, such as heart rate, blood pressure, sleep quality, etc.
[0071] Step S102, generating the multi-source data based on the first user data, the second user data and the third user data;
[0072] Specifically, based on the first user data, the second user data and the third user data obtained through the first collection channel, the second collection channel and the third collection channel respectively, these data are integrated to generate multi-source data containing multi-dimensional information such as psychological measurement, wearable devices, social media, etc.
[0073] In this embodiment, step S200 includes:
[0074] Step S201, performing data cleaning, data standardization and normalization, and data integration processing on the multi-source data to generate integrated multi-source data;
[0075] Specifically, after generating multi-source data, the preprocessing stage begins. Its core task is to carefully process and organize these raw data to ensure the effectiveness and accuracy of subsequent analysis. Specifically, this step first performs data cleaning to identify and remove noise, outliers, and duplicate records in the data, thereby purifying the data set. Next, data standardization and normalization are performed. This step aims to eliminate the deviations that may be introduced by differences in dimensions and scales between different data sources, so that all features are within the same numerical range, which is convenient for subsequent unified analysis. After that, data integration processing is performed to organize and arrange the cleaned and standardized multi-source data according to a specific logical structure and format, forming a clearly structured and easily accessible integrated multi-source data set. Through this series of preprocessing measures, a solid data foundation has been laid for the subsequent identification of mental health risk factors and abnormal behavior patterns.
[0076] In a specific embodiment, there is a multi-source data set, which includes college students' mental health data from different channels, such as psychological measurement questionnaires (such as SDS self-rating depression scale), social media activity records (such as Weibo posting frequency, interaction status), physiological indicators monitored by wearable devices (such as heart rate variability, sleep quality), etc. These raw data contain noise, outliers, and duplicate records, and the dimensions and scales of different data sources are inconsistent, so preprocessing is required to generate high-quality integrated multi-source data.
[0077] Check all data sources to identify and delete obviously unreasonable data points, such as abnormally high or low heart rate variability values (beyond human physiological limits) or obviously wrong social media posting times (such as timestamps in the future). For extreme scores in psychological measurement questionnaires (such as extremely high or extremely low SDS scores), Winsorizing processing (i.e. setting upper and lower limits, and replacing the excess with the upper and lower limits) is performed to reduce the impact of extreme values on the analysis. Check and delete duplicate rows in the data set to ensure the uniqueness of each record.
[0078] Since the score ranges and meanings of different questionnaires may be different, standardization is required. For example, convert the SDS score to a Z score with a mean of 0 and a standard deviation of 1. For count data such as posting frequency and number of interactions, normalization can be performed to scale the value range to the [0,1] interval. For example, use the minimum-maximum normalization method. For physiological indicators such as heart rate variability and sleep quality, normalization is also performed to ensure that different data sources are analyzed on the same scale.
[0079] Ensure that the timestamp format of all data sources is unified and adjust the time interval for subsequent time series analysis. For example, integrate all data by day or hour. Integrate the cleaned, standardized, and normalized data by user ID and timestamp to form a multidimensional dataset containing psychological characteristics, behavioral characteristics, and social characteristics. For example, for each user, the daily data can include SDS scores, posting frequency, average heart rate variability, etc. For missing values in the integrated dataset, select an appropriate filling strategy based on the data characteristics and missing ratio, such as mean filling, median filling, or using machine learning models to predict filling.
[0080] Through the above steps, an integrated multi-source dataset was generated, which is of high quality and uniform format, suitable for the subsequent identification of mental health risk factors and abnormal behavior patterns. The integrated dataset not only reduces the impact of noise and outliers, but also eliminates the dimension and scale differences between different data sources through standardization and normalization, providing a solid foundation for subsequent machine learning analysis and model building.
[0081] Step S202, annotating the integrated multi-source data to generate normal psychological state data and abnormal psychological state data;
[0082] Specifically, in step S202, labeling the integrated multi-source data is an important part of the mental health monitoring and early warning process. The purpose of labeling is to clearly distinguish between normal and abnormal mental states in the data, and to provide clear standards and basis for subsequent risk factor identification and abnormal behavior pattern analysis. Labeling is to mark each record or data point in the integrated multi-source data as "normal" or "abnormal". The user's mental state is comprehensively judged based on multiple dimensions such as the user's psychological characteristics, behavioral performance, and physiological indicators. For example, users who show positive emotions, are socially active, and have stable physiological indicators will be labeled as normal mental states; on the contrary, users who show negative emotions, social withdrawal, and abnormal physiological indicators may be labeled as abnormal mental states. In the labeling process, in order to improve the accuracy and efficiency of labeling, automated auxiliary tools can be used. These tools can be trained based on existing mental health data sets, learn the characteristic differences between normal and abnormal mental states, and automatically classify new data. The normal mental state data and abnormal mental state data generated by the labeling process provide clear classification standards for subsequent analysis and model building. This will not only help to more accurately identify mental health risk factors and abnormal behavior patterns, but also improve the predictive performance and reliability of early warning models, providing individuals with more personalized and effective mental health early warning intervention plans.
[0083] In a specific embodiment, during the construction of the mental health monitoring and early warning system, after the data cleaning, standardization and normalization, and integration processing in step S201, a high-quality integrated multi-source data set is obtained. Next, the integrated multi-source data is labeled to generate normal mental state data and abnormal mental state data. Specifically:
[0084] Based on professional knowledge and historical data in the field of mental health, clear labeling standards should be established. These standards should include, for example, score thresholds for psychological measurement questionnaires, abnormal performance in social media activities, changes in physiological indicators monitored by wearable devices, etc. For example, the SDS self-rating depression scale score above a certain threshold (such as 53 points) can be set as an abnormal psychological state, and the heart rate variability that is continuously higher than the normal range (such as SDNN below 50ms) can be set as an abnormal physiological indicator.
[0085] Introduce automated labeling technology to improve efficiency. For example, use machine learning models to train existing labeled data, let the model learn the labeling standards and automatically predict unlabeled data. According to the labeling results, the integrated multi-source data set is divided into two parts: normal psychological state data and abnormal psychological state data. These two parts of data will be used for subsequent identification of mental health risk factors, analysis of abnormal behavior patterns, and construction of early warning models (as shown in Table 1).
[0086] Table 1
[0087] User ID date SDS score Posting frequency Heart rate variability Annotation results U001 2024-07-01 45 5 65 normal U002 2024-07-01 60 1 40 abnormal U003 2024-07-02 52 3 55 normal … … … … … …
[0088] Through the above steps, the integrated multi-source data set was successfully annotated, and normal mental state data and abnormal mental state data were generated. These data will provide strong support for subsequent mental health monitoring and early warning work.
[0089] Step S203, generating the pre-processed multi-source data set based on the normal psychological state data and the abnormal psychological state data;
[0090] Specifically, these clearly classified data are integrated into the preprocessing process to form a preprocessed multi-source dataset for final analysis. This process not only ensures the consistency and accuracy of the data, but also provides a solid foundation for the subsequent identification of mental health risk factors and model building. Through this process, the noise and redundancy in the data are further reduced, and the structure and format of the data are unified, making the dataset more suitable for processing by machine learning algorithms. The generation of the preprocessed multi-source dataset marks the completion of the data preparation stage and paves the way for the subsequent intelligent analysis and development of early warning models. The successful implementation of this step is crucial to improving the accuracy and efficiency of the early warning system.
[0091] In a specific embodiment, after data cleaning, standardization and normalization, integration processing and annotation are completed, the annotated integrated multi-source data is further sorted to generate a pre-processed multi-source data set for final analysis. This step is the last link in the data pre-processing process and is crucial to ensure the effectiveness and accuracy of subsequent analysis. Specifically:
[0092] All the labeled data are sorted and integrated according to key fields such as user ID and timestamp to ensure the time sequence and consistency of the data. Missing values and abnormal values in the data are checked to ensure that these values have been properly processed according to the method in step S201 to obtain the pre-processed multi-source data set.
[0093] In this embodiment, step S300 includes:
[0094] Step S301, acquiring psychological characteristics, behavioral characteristics and social characteristics based on the preprocessed multi-source data set;
[0095] Specifically, the preprocessed multi-source data set is mined in depth to extract key psychological, behavioral, and social features. Psychological features may include the user's emotional state, stress level, cognitive function, etc., which are usually monitored through psychometric tools or physiological indicators. Behavioral features cover the user's daily activity patterns, living habits, and ways of coping with stress. This information can be extracted from data sources such as the user's wearable device records and mobile phone usage habits. Social features focus on the user's interaction on social media, the size and composition of the social network, and the performance in the online community. These data mainly come from social media platforms. By comprehensively using a variety of data analysis and mining techniques, the system can comprehensively capture the user's psychological, behavioral, and social status, providing a rich information basis for subsequent risk assessment and abnormal behavior identification.
[0096] Step S302, based on the psychological characteristics, the behavioral characteristics and the social characteristics, using a preset method, identifying the mental health risk factors and the abnormal behavior patterns;
[0097] Specifically, a preset analysis model is used (the model can be, for example, a long short-term memory network (LSTM) in the deep learning model in step S500 to construct a model, and the specific method is similar to step S500 and will not be repeated here), to conduct an in-depth analysis of the extracted psychological characteristics, behavioral characteristics and social characteristics. The analysis model is used to identify potential mental health risk factors and abnormal behavior patterns. The system evaluates whether the user's mental state deviates from the normal range by comparing user characteristics with known risk indicators and abnormal behavior pattern libraries. For example, a continuous high-pressure state, an increase in social isolation behavior, or abnormal mood swings may be identified as risk factors or abnormal patterns. This process depends not only on the accuracy and completeness of the data, but also on the accuracy and adaptability of the algorithm. Through comprehensive analysis and intelligent judgment, the system can promptly discover and report users' mental health problems, providing a scientific basis for subsequent intervention measures.
[0098] For example, there is a data set containing college students' mental health data, which is collected through multiple channels, including psychological measurement tools (such as depression scales, anxiety scales), social media activity records (such as Weibo posting frequency, emotional tendencies), and physiological indicators monitored by wearable devices (such as heart rate variability, sleep quality). These data have been preprocessed to generate a preprocessed multi-source data set. Step S302 extracts psychological characteristics such as emotional stability and stress level from psychological measurement tools, extracts behavioral characteristics such as posting frequency, number of interactions, and emotional tendencies from social media data, and analyzes the user's social network structure, number of friends, interaction frequency and other social characteristics. The extracted features are sorted by time series to ensure that each feature is a sequence data that changes over time. The data set is divided into a training set, a validation set, and a test set for subsequent training and evaluation of the model. A mental health risk factor and abnormal behavior pattern recognition model is constructed using an LSTM network. The LSTM model structure may include multiple LSTM layers to capture long-term dependencies in sequence data, followed by a fully connected layer (Dense layer) for final classification or regression prediction. The LSTM model is trained using the training set data. The error between the predicted value and the actual value is calculated through forward propagation, and then the network weights are adjusted using the backpropagation algorithm to minimize the error. During the training process, the training loss and validation loss are monitored to prevent overfitting, and early stopping or learning rate adjustment is performed when necessary. Use the validation set to evaluate the generalization ability of the model and adjust the model parameters or structure to improve performance. Optimize feature selection, network architecture or hyperparameters, such as the number of LSTM layers, the number of hidden units, and the learning rate, to further improve model accuracy. Use the trained LSTM model to predict the test set data and evaluate the performance of the model in actual applications. If the model performance meets the requirements, it is applied to real-time monitoring of the user's mental health status data to identify mental health risk factors and abnormal behavior patterns.
[0099] For example, the data of a specific user is analyzed through the LSTM model: the user's psychological measurement results in the past month (such as weekly depression scale scores), social media activity records (daily posts and emotional tendencies), and physiological indicators monitored by wearable devices (daily heart rate variability). The user's data is input into the trained LSTM model, and the model outputs the user's current mental health status and potential risk factors. The model prediction results show that the user's recent depression scale score has continued to rise, social media activity has decreased, and heart rate variability has increased. Based on these characteristics, the model identifies that the user is at moderate risk of depression and points out that his social withdrawal and abnormal physiological reactions may be risk factors.
[0100] In this embodiment, step S500 includes:
[0101] Step S501, monitoring the user's mental health status data in real time based on the mental health early warning model, and generating monitoring results;
[0102] Specifically, the system captures information about the user's mood changes, behavior patterns, and social interactions in real time by integrating multiple data sources, such as sensor data on the user's device, social media interaction records, and possible psychological assessment test results. This information is then input into a pre-trained mental health early warning model for analysis. The model uses complex algorithms to conduct a multi-dimensional assessment of the user's mental state, comparing the current data with the preset normal range or historical baseline data to identify any possible abnormalities or signs of risk. Ultimately, the system generates a detailed monitoring results report that not only outlines the user's current mental health status, but may also include, for example, risk level assessment, a list of abnormal indicators, and potential cause analysis, providing strong support for subsequent intervention measures.
[0103] Step S502, when the monitoring result exceeds a preset threshold, it is determined that an abnormal situation has occurred and an early warning message is issued;
[0104] Specifically, when the monitored data on the user's mental health status exceeds the preset safety threshold, the early warning mechanism is immediately activated to determine that the user's current mental state is abnormal. The setting of this threshold is based on a large amount of historical data and psychological expertise, and aims to accurately distinguish normal fluctuations from mental health risks that really need attention. Once an abnormal situation is determined to have occurred, the system quickly generates early warning information, which describes in detail the monitored abnormal indicators, risk levels and possible scope of impact, ensuring that the recipient can quickly understand and respond. The early warning information is sent instantly to relevant parties, including the user himself, guardians, psychological counselors or medical institutions, through the user's preset communication channels (such as mobile phone text messages, emails or in-app notifications), so as to quickly initiate intervention measures to protect the user's mental health safety. Among them, if the monitoring results do not exceed the preset threshold, monitoring will continue.
[0105] In a specific embodiment, the user's data is automatically collected from multiple data sources on a regular basis (such as daily or hourly). These data sources may include mental health applications, social media accounts, and wearable devices used by users. The collected data includes, but is not limited to, the user's self-assessment of emotions, the frequency of interaction with friends, the emotional tendency in the posting content, the sleep quality report, the heart rate change, etc. The collected raw data needs to undergo preprocessing steps such as cleaning, standardization, and formatting to ensure that they meet the input requirements of the mental health early warning model. For example, the text content on social media is converted into an emotional score, and the data of the wearable device is converted into a standardized heart rate variability index. The preprocessed data is input into the mental health early warning model in real time. The model uses an LSTM network to analyze time series data to capture the changing trend of the user's mental health status. The model simultaneously considers data from multiple dimensions, such as emotional stability, social activity, physiological indicators, etc., to comprehensively evaluate the user's mental health status. The model outputs the user's current mental health status assessment result. This may be a specific score, grade, or a probability distribution indicating the possibility of the user being in different mental health states. For example, the model outputs a score between 0 and 100, with the higher the score, the better the mental health; or it outputs a classification label, such as "normal", "mild stress", "moderate depression risk", etc. In addition to the overall assessment, the model may also provide specific risk factor prompts, such as "recent social activities have decreased, and there may be loneliness" or "heart rate variability has increased, and there may be greater pressure." The monitoring results are displayed on the platform in real time for users and their guardians or psychological counselors to view. If the monitoring results show that the user's mental health status is abnormal or has potential risks, the platform will automatically trigger the early warning mechanism and send a notification to the relevant personnel.
[0106] For example, at a certain time of the day, the platform collects the latest data of users: low self-evaluation scores of emotions, reduced social media activities, increased self-evaluation of learning pressure, and wearable device data showing a significant increase in heart rate variability. These data are input into the mental health early warning model in real time. After model analysis, the monitoring results are output, showing that the user is currently in a state of "moderate depression risk", and specifically points out that "recent reduction in social activities and abnormal physiological indicators" may be the main risk factors. The platform immediately sends a warning notification to the user and advises him to seek professional psychological counseling. At the same time, the notification is also sent to the user's guardian or designated emergency contact.
[0107] In this embodiment, step S600 includes:
[0108] Step S601, based on the warning information, identifying the warning factors causing the warning and the severity of the warning;
[0109] Specifically, we first analyze the warning information in depth, and through data mining and pattern recognition technology, we accurately locate the specific factors that trigger the warning, namely the warning factors. These warning factors may include, for example, abnormal behavior patterns of users (such as social withdrawal, excessive use of social media), mood swings (such as persistent depression, increased anxiety), or abnormal physiological indicators (such as decreased sleep quality, heart rate changes). At the same time, the system will also evaluate the severity of the warning, determine the potential impact of the warning on the user's mental health through quantitative analysis, and provide important reference for subsequent intervention measures. This process combines big data analysis and professional knowledge to ensure the accuracy of the warning factors and the rationality of the severity of the warning, laying a solid foundation for the subsequent assessment of mental health status and the formulation of intervention plans.
[0110] Step S602, obtaining multi-dimensional information, and evaluating the mental health status based on the multi-dimensional information, the warning factors and the warning severity, and generating an evaluation result;
[0111] Specifically, comprehensive multi-dimensional information related to the user is collected, including but not limited to the user's personal background information (such as age, gender, occupation, etc.), past mental health records, living habits, family and social environment, etc. By integrating these multi-dimensional information, the system can build a more complete and three-dimensional user portrait, providing rich data support for the assessment of mental health status. Then, using advanced algorithms and models, the collected multi-dimensional information is combined with the previously identified warning factors and warning severity for in-depth analysis and comparison. This process not only considers the direct impact of the warning factors, but also evaluates their potential chain reactions on the user's overall mental health status. For example, if the warning factor is a persistent sleep disorder, the system will further explore its possible negative impact on the user's daytime function, emotional regulation ability, and even social relationships. Finally, based on the above comprehensive evaluation, a detailed evaluation result is generated. This result not only clearly points out the user's current mental health status, but also elaborates on the specific reasons for the status, the potential impact of the warning factors, and possible future development trends. At the same time, the evaluation results can also include targeted suggestions or intervention measures to provide users with personalized mental health improvement plans to promote the overall improvement of their mental health level.
[0112] In a specific embodiment, before step S601, the method further includes: confirming the validity of the warning information, if it is valid, executing step S601, if it is invalid, not continuing to execute step S601;
[0113] Specifically, confirming the effectiveness of early warning information is a crucial link, because it is directly related to the accuracy and timeliness of subsequent intervention measures. Confirming the effectiveness of early warning information can avoid waste of resources and unnecessary panic caused by false alarms or invalid warnings. The specific methods are as follows:
[0114] By comparing user behavior and mental state information collected from different data sources (such as social media data, wearable device data, psychological test data, etc.), check whether there are consistent abnormal performances. If multiple data sources show abnormalities, the effectiveness of the warning information increases. Analyze the trend of changes in user historical data to confirm whether the current warning information represents a continuous or sudden abnormal pattern rather than an isolated event. A rule base based on psychological principles and expert experience is used to verify whether the warning information conforms to known mental health risk factors and abnormal behavior patterns. For example, if the warning information indicates that the user's sleep quality has dropped sharply, while the heart rate has continued to rise, and this change has lasted for several days, according to the rules in the field of mental health, this may be an effective warning signal. In addition, before issuing the warning information, you can interact with the user through a lightweight questionnaire or instant message to confirm the user's current mental state as a reference for the effectiveness of the warning information. User feedback can be combined with the system's preset thresholds or rules to improve the accuracy of the warning. Using machine learning models to classify and predict warning information in real time, the model can learn which warning signals are effective and which are false alarms based on a large amount of historical data. The AI-assisted verification mechanism can continuously optimize itself and improve the accuracy of warning information. For example, when monitoring a college student user Xiao Zhang's social media activity suddenly dropped sharply, and his wearable device records showed that the night sleep quality was lower than normal for a week, and his heart rate also increased slightly. The system confirms the validity of the warning information in the following steps: The system first compares Xiao Zhang's interaction frequency on social media with historical records to confirm whether the decline in activity is significant and continuous. At the same time, check the heart rate and sleep quality data recorded by the wearable device to confirm whether these two indicators also show abnormal trends. The system applies the preset rule library and finds that Xiao Zhang's sleep quality has declined and his heart rate has increased for more than a week, which meets the preset rule of "long-term accumulation of psychological pressure leading to sleep disorders". The system can also send a short questionnaire to Xiao Zhang through application messages, asking him whether he has felt increased pressure or other discomfort recently. For example, Xiao Zhang replied to confirm that he did feel anxious and difficult to fall asleep recently. The machine learning model predicts based on historical data that Xiao Zhang's current state matches the "moderate psychological pressure" category with a match of up to 90%. Based on the above information, it can be confirmed that the warning information about Xiao Zhang is valid, and then step S601 is executed to further identify the warning factors and the severity of the warning, and generate an evaluation result. In this way, the system can not only accurately identify users' mental health risks, but also take effective intervention measures in a timely manner.
[0115] For example, during the actual real-time monitoring process, a warning message was issued to a user, indicating that he or she is at moderate risk of depression. The platform first reviews and analyzes the specific monitoring data that led to the warning, including the user's emotional self-assessment score, social media activity frequency, learning stress self-assessment, heart rate variability, etc. Determine the accuracy and reliability of the warning information and exclude possible false alarm factors. After analysis, the platform identified several major warning factors: the user's recent emotional self-assessment score has continued to be low and fluctuated greatly; the frequency of social media activities has dropped significantly, and the interaction with friends has decreased; the learning stress self-assessment results show that the stress level continues to rise; the wearable device monitoring data shows that the heart rate variability increases, indicating a physiological stress response. Based on the user's current mental health status data and historical comparison, the platform assesses that his or her depression risk is at a moderate level. Further analysis shows that if the user's emotional state continues to deteriorate or the physiological indicators show more obvious abnormalities, the severity of the warning may increase. In addition to real-time monitoring data, the platform also collects the user's personal background information, such as age, gender, profession, family environment, etc. Review the user's historical mental health records to understand whether he or she has similar psychological problems or intervention experiences in the past. Communicate with users to understand their recent life events, interpersonal relationship changes and other factors that may affect their mental health. Combining real-time monitoring data, personal background information, historical records and communication results, the platform conducts a comprehensive assessment of the user's mental health status. Analyze the interactions and potential impacts between warning factors, such as how learning pressure affects users' emotions and social behaviors. Assess the user's current psychological resilience, coping mechanisms and the effectiveness of the social support system. Based on the comprehensive assessment, the platform generates a detailed assessment report, pointing out the user's current main psychological problems (such as moderate depression risk), warning factors, warning severity and possible development trends. The report may also include recommended intervention measures for the user's personal situation, such as seeking professional psychological counseling, participating in mental health support groups, adjusting learning rhythm and social activities, etc. This not only helps to detect problems in a timely manner and provide early warning support, but also provides a scientific basis for subsequent personalized intervention measures.
[0116] Embodiment 2
[0117] Figure 2 is a schematic diagram of the structure of a big data-based mental health monitoring and early warning system in Embodiment 2 of the present invention, such as Figure 2As shown, embodiment 2 provides a mental health monitoring and early warning system based on big data, including: a first generation module 201, a second generation module 202, an identification module 203, a construction module 204, an issuance module 205, a third generation module 206 and a fourth generation module 207. The first generation module 201 is used to collect user data through multiple collection channels to generate multi-source data. The second generation module 202 is used to pre-process the multi-source data to generate a pre-processed multi-source data set. The identification module 203 is used to analyze based on the pre-processed multi-source data to identify mental health risk factors and abnormal behavior patterns. The construction module 204 is used to construct a mental health early warning model based on the mental health risk factors and the abnormal behavior patterns. The issuance module 205 is used to monitor the user's mental health status data in real time based on the mental health early warning model, wherein when the user's mental health status data is monitored to be abnormal, an early warning message is issued. The third generation module 206 is used to evaluate the mental health status based on the early warning information and generate an evaluation result. The fourth generation module 207 is used to generate and push a personalized mental health early warning intervention plan based on the evaluation result.
[0118] In this embodiment, the first generating module 201 includes: an obtaining unit and a first generating unit. The obtaining unit is used to collect user data through a first collecting channel, a second collecting channel and a third collecting channel to obtain first user data, second user data and third user data. The first generating unit is used to generate the multi-source data based on the first user data, the second user data and the third user data.
[0119] In this embodiment, the second generation module 202 includes: a second generation unit, a third generation unit and a fourth generation unit. The second generation unit is used to perform data cleaning, data standardization and normalization, and data integration processing on the multi-source data to generate integrated multi-source data. The third generation unit is used to annotate the integrated multi-source data to generate normal psychological state data and abnormal psychological state data. The fourth generation unit is used to generate the pre-processed multi-source data set based on the normal psychological state data and the abnormal psychological state data.
[0120] In this embodiment, the identification module 203 includes: an acquisition unit and a first identification unit. The acquisition unit is used to acquire psychological characteristics, behavioral characteristics and social characteristics based on the preprocessed multi-source data set. The first identification unit is used to identify the mental health risk factors and the abnormal behavior patterns based on the psychological characteristics, the behavioral characteristics and the social characteristics using a preset method.
[0121] In this embodiment, the issuing module 205 includes: a fifth generating unit and an issuing unit. The fifth generating unit is used to monitor the user's mental health status data in real time based on the mental health early warning model and generate a monitoring result. The issuing unit is used to determine that an abnormal situation occurs and issue an early warning message when the monitoring result exceeds a preset threshold.
[0122] In this embodiment, the third generation module 206 includes: a second identification unit and a sixth generation unit. The second identification unit is used to identify the warning factors and the severity of the warning that cause the warning based on the warning information. The sixth generation unit is used to obtain multi-dimensional information, and based on the multi-dimensional information, the warning factors and the severity of the warning, evaluate the mental health state and generate an evaluation result.
[0123] The various variations and specific examples of the big data-based mental health monitoring and early warning method provided in Example 1 are also applicable to the big data-based mental health monitoring and early warning system provided in this embodiment. Through the above detailed description of a big data-based mental health monitoring and early warning method, those skilled in the art can clearly know the implementation method of a big data-based mental health monitoring and early warning system in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here.
[0124] Embodiment 3
[0125] Figure 3 is a schematic diagram of the structure of an electronic device in Embodiment 3 of the present invention, such as Figure 3 As shown, the third embodiment further provides an electronic device 300 , which may include: a processor 301 and a memory 302 .
[0126] The memory 302 is used to store programs; the memory 302 may include volatile memory (English: volatile memory), such as random-access memory (English: random-access memory, abbreviated: RAM), such as static random-access memory (English: static random-access memory, abbreviated: SRAM), double data rate synchronous dynamic random access memory (English: Double Data Rate Synchronous Dynamic Random Access Memory, abbreviated: DDR SDRAM), etc.; the memory may also include non-volatile memory (English: non-volatile memory), such as flash memory (English: flash memory). The memory 302 is used to store computer programs (such as applications, functional modules, etc. that implement the above method), computer instructions, etc., and the above computer programs, computer instructions, etc. can be partitioned and stored in one or more memories 302. And the above computer programs, computer instructions, data, etc. can be called by the processor 301.
[0127] The above-mentioned computer programs, computer instructions, etc. may be stored in partitions in one or more memories 302 . And the above-mentioned computer programs, computer instructions, etc. may be called by the processor 301 .
[0128] The processor 301 is used to execute the computer program stored in the memory 302 to implement each step of the method involved in the above embodiment.
[0129] For details, please refer to the relevant description in the previous method embodiment.
[0130] The processor 301 and the memory 302 may be independent structures or integrated structures. When the processor 301 and the memory 302 are independent structures, the memory 302 and the processor 301 may be coupled and connected via a bus 303 .
[0131] The electronic device of this embodiment can execute the technical solution in the above method, and its specific implementation process and technical principle are the same, which will not be repeated here.
[0132] Embodiment 4
[0133] Embodiment 4 also provides a computer-readable storage medium, including a computer program and instructions. When the computer program or instructions are executed on a computer, the computer executes the big data-based mental health monitoring and early warning method of any embodiment of the present invention.
[0134] Computer-readable storage media include: USB flash drives, mobile hard disks, ROM, RAM, magnetic disks, or optical disks, and other media that can store program codes.
[0135] This embodiment also provides a computer program product, which includes: a computer program, which is stored in a readable storage medium. At least one processor of an electronic device can read the computer program from the readable storage medium, and at least one processor executes the computer program so that the electronic device executes the solution provided by any of the above embodiments.
[0136] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the disclosure of the present invention can be performed in parallel, sequentially or in different orders, as long as the desired results of the technical solution disclosed in the present invention can be achieved, and this document does not limit this.
[0137] In summary, the big data-based mental health monitoring and early warning method and system of the present invention can comprehensively analyze multi-source data, more accurately judge the individual's mental health status, and reduce the probability of false alarms and missed alarms; it can monitor the individual's mental health status in real time, discover abnormal situations in time, and provide possibilities for early intervention; it can formulate personalized mental health early warning intervention plans based on the individual's specific situation and data analysis results, thereby improving the effectiveness of early warning intervention.
[0138] Note that the above are only preferred embodiments of the present invention and the technical principles used. Those skilled in the art will understand that the present invention is not limited to the specific embodiments described herein, and that various obvious changes, readjustments and substitutions can be made by those skilled in the art without departing from the scope of protection of the present invention. Therefore, although the present invention has been described in more detail through the above embodiments, the present invention is not limited to the above embodiments, and may include more other equivalent embodiments without departing from the concept of the present invention, and the scope of the present invention is determined by the scope of the appended claims.
Claims
1. A mental health monitoring and early warning method based on big data, characterized in that: include: Collect user data through multiple collection channels to generate multi-source data; Preprocessing the multi-source data to generate a preprocessed multi-source data set; Analyzing the preprocessed multi-source data to identify mental health risk factors and abnormal behavior patterns; Based on the mental health risk factors and the abnormal behavior patterns, construct a mental health early warning model; Based on the mental health early warning model, real-time monitoring of the user's mental health status data is performed, wherein when the user's mental health status data is monitored to be abnormal, an early warning message is issued; Based on the warning information, an assessment of mental health status is performed to generate an assessment result; Based on the assessment results, a personalized mental health early warning intervention plan is generated and pushed.
2. The big data-based mental health monitoring and early warning method according to claim 1, characterized in that: The collecting of user data through multiple collection channels to generate multi-source data includes: Collecting user data through a first collection channel, a second collection channel, and a third collection channel to obtain first user data, second user data, and third user data; The multi-source data is generated based on the first user data, the second user data, and the third user data.
3. The big data-based mental health monitoring and early warning method according to claim 1, characterized in that: Preprocessing the multi-source data to generate a preprocessed multi-source data set includes: Performing data cleaning, data standardization and normalization, and data integration processing on the multi-source data to generate integrated multi-source data; Annotating the integrated multi-source data to generate normal psychological state data and abnormal psychological state data; The preprocessed multi-source data set is generated based on the normal psychological state data and the abnormal psychological state data.
4. The big data-based mental health monitoring and early warning method according to claim 1, characterized in that: The analysis based on preprocessed multi-source data to identify mental health risk factors and abnormal behavior patterns includes: Based on the preprocessed multi-source data set, obtaining psychological characteristics, behavioral characteristics and social characteristics; Based on the psychological characteristics, the behavioral characteristics and the social characteristics, a preset method is used to identify the mental health risk factors and the abnormal behavior patterns.
5. The big data-based mental health monitoring and early warning method according to claim 1, characterized in that: The real-time monitoring of the user's mental health status data based on the mental health early warning model, wherein when the user's mental health status data is monitored to be abnormal, issuing early warning information includes: Based on the mental health early warning model, real-time monitoring of the user's mental health status data is performed, and monitoring results are generated; When the monitoring result exceeds a preset threshold, it is determined that an abnormal situation has occurred and an early warning message is issued.
6. The big data-based mental health monitoring and early warning method according to claim 1, characterized in that: The mental health status is evaluated based on the early warning information, and the evaluation results are generated including: Based on the warning information, identifying the warning factors causing the warning and the severity of the warning; Acquire multi-dimensional information, and based on the multi-dimensional information, the warning factors and the warning severity, evaluate the mental health status and generate an evaluation result.
7. A mental health monitoring and early warning system based on big data, characterized in that: include: The first generation module is used to collect user data through multiple collection channels to generate multi-source data; A second generating module is used to preprocess the multi-source data to generate a preprocessed multi-source data set; An identification module, used to analyze and identify mental health risk factors and abnormal behavior patterns based on the pre-processed multi-source data; A construction module, for constructing a mental health early warning model based on the mental health risk factors and the abnormal behavior patterns; An issuing module is used to monitor the user's mental health status data in real time based on the mental health early warning model, wherein when the user's mental health status data is monitored to be abnormal, an early warning message is issued; A third generating module is used to evaluate the mental health status based on the warning information and generate an evaluation result; The fourth generation module is used to generate and push a personalized mental health early warning intervention plan based on the evaluation results.
8. The big data-based mental health monitoring and early warning system according to claim 7, characterized in that: The first generation module comprises: an obtaining unit, configured to collect user data through a first collection channel, a second collection channel, and a third collection channel to obtain first user data, second user data, and third user data; The first generating unit is configured to generate the multi-source data based on the first user data, the second user data and the third user data.
9. An electronic device, characterized in that: include: at least one processor; as well as a memory communicatively coupled to the at least one processor; In which, the memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the big data-based mental health monitoring and early warning method described in any one of claims 1-6.
10. A computer-readable storage medium, characterized in that: It includes computer programs and instructions. When the computer program or the instructions are run on a computer, the computer executes the big data-based mental health monitoring and early warning method as described in any one of claims 1 to 6.
Citation Information
Cited By
Student mental health early warning method and system based on multi-modal data fusion
CN119207804A
Risk behavior identification method, system and device based on decision tree and medium
CN120048511A
Mental health assessment system based on big data
CN120260938A
Psychological risk assessment early warning model construction method
CN120527004A
Psychological crisis comprehensive perception system and method based on AI and multi-source data
CN120636801A
