Disclosure of Invention
The embodiment of the application solves the problem that the effectiveness of mental health detection and analysis is not high in the prior art by a data visualization mode by providing the mental health detection and analysis method based on the large model, and improves the effectiveness of data visualization in the mental health detection and analysis.
The embodiment of the application provides a mental health detection and analysis method based on a large model, which comprises the following steps of S1, obtaining student data characteristics at preset time by carrying out characteristic extraction on student data corresponding to students to be detected, inputting the student data characteristics at the preset time into the preset model to obtain data characteristic parameters of the student data at the preset time, wherein the preset model is the large data model for analyzing the data characteristic parameters of the student data, the preset time represents the starting date of inquiry, the data characteristic parameters are used for representing the overall situation that the student data is abnormal at the preset time, S2, obtaining a data fluctuation coefficient by analyzing the data characteristic parameters in a historical time period, obtaining a data abnormal evaluation index by combining the student data characteristics and the data characteristic parameters at the preset time, wherein the data fluctuation coefficient is used for describing the abnormal change trend of the student data, the data abnormal evaluation index is used for comprehensively measuring the overall abnormal state of the student data, S3, carrying out abnormal grade division on the students to be detected according to the preset evaluation threshold and the data visual, and the data abnormal group is used for identifying the abnormal degree of the students to be detected.
The method comprises the steps of obtaining a first class emotion judgment result output by a preset model, obtaining a second class emotion judgment result output by the preset model, obtaining a third class emotion judgment result output by the preset model, obtaining a student score and a homework completion rate of the student to be detected, obtaining a real-time heart rate of the student to be detected through wearable equipment, wherein the real-time heart rate is used for describing the heart rate of the student to be detected, the wearable equipment corresponds to the student to be detected one by one, and the student ID and the first class emotion judgment result are orderly stored in the preset database according to the collection time.
Further, the student data is specifically obtained by extracting all original data from a preset database within a detection time period, judging whether to execute missing value processing according to the scanning result of the original data, and if the missing value processing is executed, judging the data type of the obtained missing value and filling, wherein the detection time period represents a date continuously intercepted from the preset time to the past.
Further, the student data characteristics comprise a first characteristic, a second characteristic and a third characteristic, wherein the specific acquisition process of the first characteristic comprises the steps of obtaining a reduced data quantity, a constant data quantity and an increased data quantity by counting first data, obtaining a data change frequency by combining the total quantity of the first data, and meanwhile, accumulating data duty ratio evaluation values obtained by combining preset data duty ratio evaluation values to obtain a data evaluation value characteristic, wherein the data change frequency comprises a data reduction frequency, a data constant frequency and a data increase frequency, and the data duty ratio evaluation value represents the product of the data change frequency and a corresponding preset data duty ratio evaluation value; traversing the first data to find out whether a transition point exists in the first data, if the transition point exists, counting the number of the transition points to obtain the number of the transition data, and combining the total number of the first data to obtain a data transition quantity characteristic, wherein the first characteristic comprises a data evaluation value characteristic and a data transition quantity characteristic, the second characteristic comprises a performance decline characteristic and a work decline characteristic, the performance decline characteristic represents the ratio of the number of times of performance decline to the total number of the second data, the work decline characteristic represents the ratio of the number of times of work decline to the total number of the second data, the third characteristic comprises a wake number characteristic and a sleep ratio characteristic, the wake number characteristic represents the ratio of the number of times of wake to the total number of the third data, and the sleep ratio represents the ratio of average sleep time to the preset ideal sleep time.
Further, the specific obtaining process of the second feature includes traversing the second data to find whether a score-lowering point and an operation-lowering point exist in the second data, if so, counting the number of the score-lowering points and the operation-lowering points to obtain the score-lowering times and the operation-lowering times respectively, and combining the total number of the second data to obtain the second feature, wherein the score-lowering points are used for marking student score-lowering, and the operation-lowering points are used for marking operation completion rate lowering.
Further, the specific acquisition process of the third characteristic comprises the steps of traversing the third data to find whether awake points exist in the third data, counting the number of the awake points to obtain the awake times if the awake points exist, obtaining the awake times characteristic by combining the third data, wherein the awake points are used for marking whether a student to be detected is awake, obtaining a sleep time period according to the acquisition time of adjacent sleep heart rates in the third data, counting the sleep time period to obtain daily sleep time, analyzing all daily sleep time of the detection time period to obtain average sleep time, and obtaining the sleep duty ratio characteristic by combining the preset ideal sleep time.
Further, the specific acquisition process of the data fluctuation coefficient comprises the steps of acquiring a development coefficient weight from a preset database, wherein the development coefficient weight comprises a first coefficient weight, a second coefficient weight and a third coefficient weight, extracting historical data characteristic parameters from the preset database, grouping the historical data characteristic parameters according to the change relation of a data characteristic parameter threshold and the historical data characteristic parameters to obtain a historical parameter set, wherein the data characteristic parameter threshold comprises a first parameter threshold and a second parameter threshold, the historical parameter set comprises a first parameter set, a second parameter set and a third parameter set, obtaining an interval time period according to preset moments of adjacent data characteristic parameters in the historical parameter set, obtaining a data fluctuation coefficient according to the historical data characteristic parameters, the interval time period and the development coefficient weight, and storing the data fluctuation coefficient and the corresponding preset moments in the preset database, wherein the specific calculation formula of the data fluctuation coefficient is as follows:
;
Wherein n is the number of the student to be detected,N is the total number of students to be detected,For the nth student to be tested for data fluctuation coefficient,Is the number of the data characteristic parameter in the first parameter group of the nth student to be detected,,For the total number of data characteristic parameters in the first parameter set of the nth student to be tested,Is the number of the data characteristic parameter in the second parameter set of the nth student to be detected,,For the total number of data characteristic parameters in the second parameter set of the nth student to be tested,Is the number of the data characteristic parameter in the third parameter group of the nth student to be detected,,For the total number of data characteristic parameters in the third parameter set of the nth student to be tested,The nth student to be tested is the nth student in the first parameter setThe data characteristic parameters are used to determine,The nth student to be tested is the nth student in the second parameter setThe data characteristic parameters are used to determine,The nth student to be tested is the third parameter setThe data characteristic parameters are used to determine,The nth student to be tested is the nth student in the first parameter setThe interval period of the data characteristic parameter,The nth student to be tested is the nth student in the second parameter setThe interval period of the data characteristic parameter,The nth student to be tested is the third parameter setThe interval period of the data characteristic parameter,For the average interval period of the first parameter set of the nth student to be tested,For the average interval period of the second parameter set of the nth student to be tested,For the average interval period of the third parameter set of the nth student to be tested,For the first coefficient weight to be the first,For the weight of the second coefficient,And the third coefficient weight.
Further, the specific acquisition process of the data abnormality assessment index comprises the steps of acquiring a data assessment weight from a preset database, wherein the data assessment weight comprises a first weight and a second weight, acquiring the data abnormality assessment index according to a data characteristic parameter, a student data characteristic, a data fluctuation coefficient and the data assessment weight at a preset moment, and storing the data abnormality assessment index and the corresponding preset moment in the preset database, wherein the calculation formula of the data abnormality assessment index is as follows:
;
In the formula,Evaluating an index for the nth data anomaly of the student to be tested,Is the data characteristic parameter of the nth student to be detected,Evaluating the value characteristic for the nth student to be tested,For the nth sleep duty cycle characteristic of the student to be tested,For the nth student to be tested for data transition quantity characteristics,For the nth student to be tested for performance degradation characteristics,Is the homework descending feature of the nth student to be detected,Is the feature of the nth number of wakefulness of the student to be detected,For the reference data fluctuation coefficient,As a first weight to be used,As a result of the second weight being set,Is an adjustment factor.
Further, the grading comprises the specific steps of A1, classifying each student to be detected into corresponding abnormal grades by comparing the data abnormality evaluation index of the student to be detected with a preset evaluation threshold, wherein the abnormal grades comprise a first grade, a second grade and a third grade, A2, sorting the abnormal grades according to the data abnormality evaluation index to obtain a data abnormality group, and storing the data abnormality group in a preset database.
Further, the data visualization comprises the specific processes of acquiring a data anomaly group, a data anomaly evaluation index and a data fluctuation coefficient from a preset database, and drawing a data visualization chart, wherein the specific processes are that a data fluctuation coefficient time sequence chart is drawn according to the change relation between the data fluctuation coefficient and the preset moment; drawing a time sequence diagram of the data abnormality evaluation index according to the change relation between the data abnormality evaluation index and the preset time, and drawing a layout of the data abnormality components according to the change relation between the student ID and the abnormality grade.
One or more technical solutions provided in the embodiments of the present application at least have the following technical effects or advantages:
1. The method comprises the steps of carrying out feature extraction on student data corresponding to students to be detected to obtain student data features at preset time, inputting a preset model to obtain data feature parameters of the student data at the preset time, analyzing the data feature parameters at the historical time to obtain data fluctuation coefficients, combining the student data features and the data feature parameters at the preset time to obtain data abnormality evaluation indexes, and finally combining a preset evaluation threshold to carry out abnormality grading on the students to be detected to obtain data abnormality groups, so that accurate identification and hierarchical management on the student data with abnormal changes are realized, further improvement of effectiveness of data visualization in mental health detection and analysis is realized, and the problem that effectiveness of mental health detection and analysis is not high in the prior art is effectively solved through a data visualization mode.
2. The method comprises the steps of acquiring data to obtain the original data of students to be detected, judging whether to execute missing value processing according to the scanning result of the original data, filling according to the data type of the missing value if the missing value processing is executed, and finally classifying according to the data type to obtain the student data, so that the integrity and the accuracy of the student data are ensured, and the calculation precision of the data fluctuation coefficient and the data abnormality assessment index is improved in subsequent analysis.
3. Historical data characteristic parameters are extracted from a preset database, the historical data characteristic parameters are grouped according to the change relation between the data characteristic parameter threshold and the historical data characteristic parameters to obtain a historical parameter set, then an interval period is obtained according to the preset time of the adjacent data characteristic parameters in the historical parameter set, a data fluctuation coefficient is obtained by combining the development coefficient weight obtained from the preset database, and a data abnormal evaluation index is obtained by combining the data characteristic parameters, the student data characteristic and the data evaluation weight at the preset time, so that the numerical value of the data abnormal evaluation index is realized, and the accurate evaluation of the overall abnormal state of student data is realized.
Detailed Description
According to the embodiment of the application, the problem of low effectiveness of mental health detection and analysis in a data visualization mode in the prior art is solved by providing the mental health detection and analysis method based on the large model, the student data characteristics at the preset moment are obtained by carrying out characteristic extraction on the student data corresponding to the students to be detected, the data characteristic parameters of the student data at the preset moment are obtained by inputting the preset model, the data fluctuation coefficient is obtained by analyzing the data characteristic parameters in the historical time period, the data abnormality evaluation index is obtained by combining the student data characteristics and the data characteristic parameters at the preset moment, and finally the data abnormality group is obtained by carrying out abnormality grading on the students to be detected by combining the preset evaluation threshold, so that the effectiveness of data visualization in mental health detection and analysis is improved.
The technical scheme in the embodiment of the application aims to solve the problem of low effectiveness of psychological health detection and analysis in a data visualization mode, and the general idea is as follows:
The data abnormality evaluation index is obtained through the data characteristics of the students and the data characteristic parameters obtained by inputting the data characteristics of the students into the preset model, and the data abnormality group is obtained by carrying out abnormality grading on the students to be detected by combining with the preset evaluation threshold value, so that the effect of improving the data visualization effectiveness in mental health detection and analysis is achieved.
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
The method for detecting and analyzing the mental health based on the large model comprises the following steps of S1, obtaining student data characteristics at preset time through characteristic extraction of student data corresponding to students to be detected, inputting the student data characteristics at the preset time into a preset model to obtain data characteristic parameters of the student data at the preset time, wherein the student data are behavior data extracted from a preset database and used for representing students to be detected, the preset model is a large data model used for analyzing the data characteristic parameters of the student data, the preset time represents the starting date of inquiry, the data characteristic parameters are used for representing the overall situation that the student data are abnormal at the preset time, S2, obtaining data abnormality evaluation indexes by analyzing the data characteristic parameters at historical time periods, combining the student data characteristics and the data characteristic parameters at the preset time, wherein the data abnormality evaluation indexes are used for describing abnormal change trends of the student data, and are used for comprehensively measuring the overall abnormal states of the student data, S3, carrying out abnormality grade classification on the students to be detected according to preset evaluation thresholds and carrying out abnormality grade classification on the preset evaluation thresholds, and carrying out specific abnormality data identification on the students to be detected by the students to be detected are different from the preset evaluation thresholds, and determining the abnormality degree of the students to be detected is different from the students to be detected.
In this embodiment, in order to ensure accurate calculation of data feature parameters, a universal thousands of questions Plus model (Qwen-Plus) is used as a preset model, which can solve more diversified tasks, and has wide application prospects in multiple fields such as natural language processing, intelligent customer service, content creation and other scenes, firstly, the universal thousands of questions Plus model is set through a setting statement, then an inquiry statement is obtained through student data features, and the inquiry statement is input into the universal thousands of questions Plus model to obtain the data feature parameters; for example, the setup sentence may be "i input a student's data evaluation value feature, sleep duty feature, data transition amount feature, performance decline feature, homework decline feature, and awake number feature in order," please output the student's data feature parameter, the data feature parameter represents the overall score of the student's feature, wherein the first two features are positive correlation and the rest are negative correlation, the score range is 0-10", if the data evaluation value feature of a student to be detected is 0.7, sleep duty feature is 0.8, data transition amount feature is 0.2, performance decline feature is 0.3, homework decline feature is 0.4, awake number feature is 0.2," the inquiry sentence may be "0.7 0.8 0.2 0.3 0.4 0.2," and the final data feature parameter output by the thousands of Plus model is "7.33," the value is the average value calculated by the thousands of Plus model taking the above feature and its positive and negative correlation into consideration, the specific formula is calculated asThe method has the advantages that the data characteristic parameters are finally obtained to be 7.33, the improvement of the effectiveness of data visualization in the psychological health detection and analysis is realized, and the problem that the effectiveness of the psychological health detection and analysis is not high in the prior art in a data visualization mode is effectively solved.
Further, the method comprises the steps of acquiring original data of students to be detected by carrying out characteristic extraction on student data corresponding to the students to be detected to obtain student data characteristics at preset time, and storing the original data into a preset database, wherein the specific process is that image acquisition is carried out through image acquisition equipment to obtain classroom images, denoising images obtained by preprocessing the classroom images are input into a preset image model to obtain data to be analyzed, the classroom images are images obtained by acquiring classrooms of the students to be detected at preset time by the image acquisition equipment, preprocessing comprises denoising, gaussian filtering, graying and face segmentation, the denoising images represent face images obtained after noise in the classroom images is removed, the preset image model represents a big data model for a face emotion recognition task, the data to be analyzed comprises first analysis data, second analysis data and third analysis data, the first analysis data represent first emotion judgment results output by the preset model, the second analysis data represent second emotion judgment results output by the preset model, and the third analysis data represent third emotion judgment results output by the preset model; obtaining the real-time heart rate of the student to be detected through a wearable device, wherein the real-time heart rate is used for describing the heart rate of the student to be detected, the wearable device is in one-to-one correspondence with the student to be detected, the student IDs and the original data are orderly stored in a preset database according to the acquisition time, the student IDs are the ordering of the student to be detected, the acquisition time is the time for data acquisition, and the original data comprise data to be analyzed, student achievements, homework completion rate and real-time heart rate.
In the embodiment, equipment in a data acquisition process is critical to the accuracy of original data, the image acquisition equipment is a monitoring camera of a classroom, the wearable equipment is an intelligent bracelet, the acquisition time of data to be analyzed is 5 seconds once, the acquisition time of real-time heart rate is 1 second once, the acquisition time of student achievements and work completion rates is one time a day, the embodiment uses a universal emotion answer Plus model as a preset image model, firstly acquires classroom images through the monitoring camera, then performs denoising, gaussian filtering and graying on the classroom images, and uses a Haar cascade classifier to divide the classroom images to obtain face images of students to be detected, then inputs the face images and inquiry sentences into the universal emotion answer Plus model to obtain data to be analyzed, classifies the face images and inquiry sentences according to the values of the data to be analyzed to obtain first analysis data, second analysis data and third analysis data, for example, the inquiry sentences can be emotion answer 3, passive emotion answer 2, passive emotion answer 1', if the face images are crying, the face images are the face images to be detected, and the face images to be detected are all the face images to be analyzed, and the face data to be analyzed are all the face data to be analyzed if the face images to be analyzed are the face data to be read, and the face data to be analyzed are all the face data to be analyzed.
It should be understood that the data to be analyzed output by the preset image model is not only used for directly judging the emotion state of the student, but also provides a more comprehensive reference for analyzing the fluctuation trend of the psychological health of the student, the psychological health of the student may be related to the long-term expression, the physiological state or the dynamic change in the teaching activities of the student, the analysis of the fluctuation data is helpful for deeply understanding the variation trend of the psychological state of the student, and not limited to the emotion expression at a single moment, and the purpose of the preset image model is not to accurately analyze the details of each emotion, but to explore the law of the overall psychological health state through data induction and summarization, so that the fluctuation of the data to be analyzed has important influence on the finally obtained abnormal evaluation index of the data and is helpful for further improving the reliability of the psychological health detection and analysis method.
The method comprises the steps of acquiring student data, wherein a detection time period is defined by a user, acquiring the student data, and the method comprises the steps of extracting all original data in the detection time period from a preset database through a written database query statement, judging whether to execute missing value processing through scanning results of the original data, judging the data type of the obtained missing value and filling the missing value if the missing value processing is executed, and the specific process comprises the steps of filling the data to be analyzed of adjacent original data by using the data to be analyzed of the corresponding ID if the data type of the missing value is the data to be analyzed, filling the average value of the student score in the original data if the data type of the missing value is the student score, filling the average value of the work completion rate in the original data by using the median of the work completion rate in the original data if the type of the missing value is the work completion rate, and filling the average value of the real-time heart rate in the adjacent two original data if the missing value type is the real-time heart rate, classifying the data to obtain the student data according to the data after the filling, wherein the student data comprises first data, second data and third data, the first data corresponds to the student ID to the data to be analyzed, and the second data correspond to the student ID, and the work completion data, the second data correspond to the data, and the data of the data to be analyzed by the ID, and the data to be continuously transmitted to the preset data, and the data are continuously represent the failure time period, and the failure time is generated from the preset data, and the data.
In the embodiment, the data to be analyzed and the real-time heart rate are changed along with time, the front data to be analyzed and the rear data to be analyzed and the real-time heart rate generally have certain continuity and correlation, so that the missing data to be analyzed are filled by the data to be analyzed of adjacent original data, the missing real-time heart rate is filled by the average value of the two adjacent real-time heart rates, the continuity of the original data is improved, the student achievements are comprehensive reflection of learning achievements in a detection period, the missing student achievements are filled by the average value of the student achievements in the original data, so that deviation of feature extraction caused by the missing student achievements is avoided, extreme numerical values possibly exist in the work completion rate, the reasonability of the original data is maintained by the fact that the missing value of the median of the work completion rate in the original data is filled, the first data are used for describing emotional state information of students to be detected, the academic performance information of the students to be detected is described by the second data, the accuracy of the student data is guaranteed through missing value processing, and the reliability of feature extraction of the health detection and analysis method is further improved.
Further, the student data characteristics comprise a first characteristic, a second characteristic and a third characteristic, the specific acquisition process of the first characteristic is as follows, the first data is counted to obtain a reduced data quantity, a constant data quantity and a raised data quantity, the data change frequency is obtained by combining the first data quantity, meanwhile, the data duty ratio evaluation value obtained by combining the preset data duty ratio evaluation value is accumulated to obtain a data evaluation value characteristic, the reduced data quantity refers to the occurrence number of the first analysis data in the first data, the constant data quantity refers to the occurrence number of the second analysis data in the first data, the raised data quantity refers to the occurrence number of the third analysis data in the first data, the data change frequency comprises a data reduction frequency, a data constant frequency and a data increase frequency, the data reduction frequency represents the ratio of the reduced data quantity to the total quantity of the first data, the data constant frequency represents the ratio of the constant data quantity to the total quantity of the first data, the data increase frequency represents the ratio of the raised data quantity to the total quantity of the first data, the preset data duty ratio evaluation value is determined by personnel according to the type of the data to be analyzed, the data duty ratio evaluation value represents the number of the data to be analyzed, the data duty ratio represents the occurrence number of the data change frequency to the corresponding preset data duty ratio evaluation value occurs in the first data, the number of the first analysis data is calculated by the preset data duty ratio evaluation value represents the type of the data, the number of the data to represent the transition point to be compared with the corresponding preset data duty ratio evaluation value, the first data is used for obtaining a transition point after the first data is compared with the first data to be analyzed to be compared with the first data to have the total data to be compared with the first data to have the total value to be analyzed data, the first data is used to represent the first data to have the state data is used to have the state data is compared to represent the data is compared with the first data is compared to the data is obtained to represent the data is converted data is compared with the first data is used to the data is obtained to have the data is compared to have, the data transfer variable feature represents a ratio of a transition data amount to a first data total amount, the first feature comprises a data evaluation value feature and a data transfer variable feature, the second feature comprises a performance decline feature and a work decline feature, the performance decline feature represents a ratio of a performance decline number to the second data total amount, the work decline feature represents a ratio of a work decline number to the second data total amount, the third feature comprises an awake number feature and a sleep duty ratio feature, the awake number feature represents a ratio of an awake number to the third data total amount, and the sleep duty ratio feature represents a ratio of an average sleep time to a preset ideal sleep time.
In this embodiment, the preset data duty ratio evaluation value should be set by a preset person evaluating the possible influence of different data to be analyzed, for example, the third analysis data may indicate that a student to be detected has a positive influence, so the preset data duty ratio evaluation value of the third analysis data is set to 1, the first analysis data may have a negative influence, so the preset data duty ratio evaluation value of the first analysis data is set to-1, the second analysis data may indicate no significant influence, so the preset data duty ratio evaluation value of the second analysis data is set to 0, by calculating the occurrence frequency of different data to be analyzed and evaluating the data duty ratio evaluation value in combination with the preset data duty ratio evaluation value, the data change frequency is used for measuring the occurrence frequency of specific data to be analyzed in a detection time period, the higher the data change frequency indicates that the data to be analyzed has more frequent occurrence, the preset data duty ratio evaluation value reflects the influence degree of the data to be analyzed on the whole emotion state, the higher the preset data duty ratio evaluation value indicates that the influence of the data to the whole state is larger, the total number of the data to be analyzed on the change over the state is higher, the total number of the data to be analyzed is used for measuring the change frequency of the data to be detected in the detection time period, the data to be used for measuring the characteristic change value is more than the value to be used for the change the data to be measured, the characteristic change value is used for the data to be compared with the data to be measured, and the characteristic change value is used for the data to be measured, and the change value is more has more than the total number to be measured, and the characteristic value is used, and the value is used to be compared, the higher the data transition quantity characteristic is, the more severe the emotion fluctuation of student data in the detection time period is, if the transition data quantity is 0, the condition that the emotion fluctuation of the student does not occur in the detection time period is indicated to be 0, the dynamic change condition of the emotion of the student to be detected can be accurately estimated through the data transition quantity characteristic, and the accuracy of the feature extraction of the mental health detection and analysis method is improved.
It should be understood that, in the above analysis method, in order to extract the trend and the fluctuation situation of the first data, the fluctuation of the first data may reflect the emotion fluctuation of the student, so the method focuses on capturing these trend, and thus focuses on the possible emotion problems, for example, the data evaluation value features are not only used to simply quantify the positive and negative effects of the emotion, but also may provide valuable references for evaluating the emotion fluctuation, and this processing manner is not used to precisely locate each emotion detail, but is used to quickly identify possible abnormal situations through the overall analysis of the first data, improve the data processing efficiency, in terms of the data transition quantity features, focus on the frequency and the stability of emotion change, reflect the fluctuation situation of the emotion state through statistics of the transition points, the student with frequent emotion fluctuation may have the emotion problem that needs further focus, and the student with relatively stable emotion state indicates that the emotion of the student has relatively balanced emotion during the detection period, and it needs to be noted that these features are not directly used to judge the severity of the mental health detection problem, but are used to help analyze the fluctuation of the first data, and further support the evaluation of the abnormal data.
Further, the specific acquisition process of the second feature comprises the steps of traversing the second data to find whether a score-falling point and a job-falling point exist in the second data, if so, counting the number of the score-falling point and the job-falling point to obtain the score-falling times and the job-falling times respectively, and combining the total number of the second data to obtain the second feature, wherein the second feature comprises the score-falling feature and the job-falling feature, the score-falling times represent the total number of the score-falling points, the job-falling times represent the total number of the job-falling points, the score-falling feature represents the ratio of the score-falling times to the total number of the second data, the job-falling feature represents the ratio of the job-falling times to the total number of the second data, the score-falling point represents the smaller student score when one student score in the second data is smaller than the previous student score, and the job-falling point represents the smaller job-completion rate when the one job-completion rate in the second data is smaller than the previous job-completion rate, and the score-falling points are used for marking the descending of the students.
In this embodiment, the second data is traversed to find the score-lowering points and the job-lowering points, where the score-lowering points and the job-lowering points are used to mark the lowering trends of the student score and the job completion rate, and by counting the number of the score-lowering points and the job-lowering points, the condition of fluctuation of the student data in the detection period can be quantified, the larger the score-lowering feature is, the larger the score fluctuation of the student data in the detection period is, the higher the instability of the job completion rate is, if the score-lowering number or the job-lowering number is 0, it is indicated that the score or the job completion rate of the student in the detection period is kept stable, the corresponding lowering feature is 0, and the condition of change of the student data in the detection period is accurately evaluated through analysis of the second feature, thereby improving the accuracy of feature extraction of the mental health detection and analysis method.
Further, the specific acquisition process of the third characteristic comprises the steps of traversing to find whether a wake point exists in the third data, counting the number of the wake points to obtain the wake times if the wake point exists, combining the third data to obtain the wake times characteristic, wherein the wake point represents the real-time heart rate of one student to be detected in the third data, which is not the sleep heart rate, and the real-time heart rate when the previous real-time heart rate is the sleep heart rate, the sleep heart rate represents the heart rate of which the real-time heart rate is lower than a preset heart rate threshold value, the preset heart rate threshold value is determined by a preset person according to the age of the student to be detected, the sleep heart rate represents the real-time heart rate of the student to be detected in a sleep state when the data acquisition is performed, the wake point is used for marking whether the student to be detected is awake, the wake times represents the total number of the wake points, and the wake times characteristic represents the ratio of the wake times to the total number of the third data; obtaining a sleep time period according to the acquisition time of the adjacent sleep heart rate in the third data, counting the sleep time period to obtain a daily sleep time, analyzing all the daily sleep time of the detection time period to obtain an average sleep time, obtaining a sleep duty ratio characteristic by combining with a preset ideal sleep time, wherein the sleep time period represents the difference between the acquisition time of each sleep heart rate and the acquisition time of the last sleep heart rate, the daily sleep time represents the sum of all the sleep time periods in one natural day, the average sleep time represents the average value of all the third data daily sleep time in the detection time period, the preset ideal sleep time is determined by preset personnel according to the age of the student to be detected, the sleep occupancy rate feature represents the ratio of the average sleep time to the preset ideal sleep time, and the third feature comprises a wake number feature and a sleep occupancy rate feature.
In this embodiment, the setting of the preset heart rate threshold and the preset ideal sleep time should be determined according to the age bracket of the student to be detected; for example, the preset heart rate threshold may be set to a value lower than the resting heart rate, because the metabolic level of the human body during sleep is reduced, resulting in a natural decrease in heart rate, resulting in a heart rate during sleep that is generally lower than the resting heart rate, for example, a typical resting heart rate range of a child aged 6-12 years is 70-110 times/min, the preset heart rate threshold may be set to 60 times/min, a resting heart rate of a teenager aged 13-18 years is generally between 60-100 times/min, the preset heart rate threshold may be set to 55 times/min, the child needs more sleep time, the preset ideal sleep time should be set to 9-12 hours, the preset ideal sleep time of the teenager should be set to 8-10 hours, the number of wakefulness points may be used to mark the change from the sleep state to the awake state by traversing the third data, the number of wakefulness times may reflect whether the student data in the detection period is frequently wakefulness, the characteristics of the number of wakefulness times represent the student is more than the number of wakefulness, the student is about the number of times of wakefulness to be detected during sleep, the time is calculated by counting the number of times of wakefulness of the student is 60-100 times/min, the time is calculated by calculating the preset ideal sleep time and the average time to represent the overall quality of the sleep characteristics is about the sleep quality of the sleep to be more than the ideal by the sleep time is calculated by the average time to be compared to the ideal sleep time to the sleep time is compared to the ideal time to the sleep time, the sleep time is calculated to be compared to the sleep time is required to be compared to the sleep time to be measured by the sleep time is about the ideal time to 12 is compared to 12 and the sleep time. Thereby improving the accuracy of overall monitoring and analysis of the psychological health status of students.
Further, the specific acquisition process of the data fluctuation coefficient comprises the steps of acquiring a development coefficient weight from a preset database, wherein the development coefficient weight comprises a first coefficient weight, a second coefficient weight and a third coefficient weight, the first coefficient weight is used for describing the influence degree of a first parameter group on the data fluctuation coefficient, the second coefficient weight is used for describing the influence degree of a second parameter group on the data fluctuation coefficient, and the third coefficient weight is used for describing the influence degree of a third parameter group on the data fluctuation coefficient; extracting historical data characteristic parameters from a preset database, grouping the historical data characteristic parameters according to a change relation between a data characteristic parameter threshold and the historical data characteristic parameters to obtain a historical parameter set, wherein the historical data characteristic parameters represent data characteristic parameters of a historical time period, the data characteristic parameter threshold comprises a first parameter set, a second parameter set and a third parameter set, the historical parameter set comprises the first parameter set, the second parameter set represents the data characteristic parameters which are not more than the first parameter threshold in the historical time period, the second parameter set represents the data characteristic parameters which are not more than the second parameter threshold and are more than the first parameter threshold in the historical time period, the third parameter set represents the data characteristic parameters which are more than the second parameter threshold in the historical time period, an interval time period is obtained according to preset moments of adjacent data characteristic parameters in the historical parameter set, the interval time period represents a difference between preset moment of each data characteristic parameter and preset moment of one data characteristic parameter in the same set, the data fluctuation coefficient is obtained according to the historical data characteristic parameters, the interval time period and the development coefficient weight, the data fluctuation coefficient is stored in the preset database, the specific calculation formula of the data fluctuation coefficient is as follows:
;
Wherein n is the number of the student to be detected,N is the total number of students to be detected,For the nth student to be tested for data fluctuation coefficient,Is the number of the data characteristic parameter in the first parameter group of the nth student to be detected,,For the total number of data characteristic parameters in the first parameter set of the nth student to be tested,Is the number of the data characteristic parameter in the second parameter set of the nth student to be detected,,For the total number of data characteristic parameters in the second parameter set of the nth student to be tested,Is the number of the data characteristic parameter in the third parameter group of the nth student to be detected,,For the total number of data characteristic parameters in the third parameter set of the nth student to be tested,The nth student to be tested is the nth student in the first parameter setThe data characteristic parameters are used to determine,The nth student to be tested is the nth student in the second parameter setThe data characteristic parameters are used to determine,The nth student to be tested is the third parameter setThe data characteristic parameters are used to determine,The nth student to be tested is the nth student in the first parameter setThe interval period of the data characteristic parameter,The nth student to be tested is the nth student in the second parameter setThe interval period of the data characteristic parameter,The nth student to be tested is the third parameter setThe interval period of the data characteristic parameter,For the average interval period of the first parameter set of the nth student to be tested,For the average interval period of the second parameter set of the nth student to be tested,For the average interval period of the third parameter set of the nth student to be tested,For the first coefficient weight to be the first,For the weight of the second coefficient,And the third coefficient weight.
In this embodiment, the setting of the data characteristic parameter threshold should be set according to the historical data characteristic parameter, for example, the first parameter threshold may be set as the average value of the historical data characteristic parameter, and the second parameter threshold may be set as the sum of the average value and the standard deviation of the historical data characteristic parameter, where the data characteristic parameter is used to describe the overall situation that the student data is abnormal at the preset time, the higher the data characteristic parameter is, the more the student data to be detected is abnormal at the preset time, the higher the data fluctuation coefficient is, the interval period is used to represent the time length of the change of the data characteristic parameter, the longer the interval period is, the lower the rate of change of the data characteristic parameter is, the lower the data fluctuation coefficient is, the greater the development coefficient weight is, the more significant the influence of the data characteristic parameter of the group on the data fluctuation coefficient is, and the data fluctuation coefficient is higher, and the higher the data fluctuation coefficient is the overall consideration of the historical data characteristic parameter, the interval period and the development coefficient weight is, the higher the data fluctuation coefficient is used to describe the abnormal change of the data, and the higher the student fluctuation coefficient is the greater the abnormal trend.
、、Respectively in the history parameter sets、AndTo simplify the description of the independent variables of the data fluctuation coefficients, a first fluctuation index is defined as,Defining a second fluctuation index as the first fluctuation index of the nth student to be detected,Defining a third fluctuation index as the second fluctuation index of the nth student to be detected,The third fluctuation index of the nth student to be detected is represented as a simplified formulaToThe number of the components is 2.5,The number of the particles is 3.4,The number of the components is 3.1,The total number of the components is 0.3,The total number of the components is 0.3,For example, 0.4, the statistical table of variation of the data fluctuation coefficient is shown in table 1:
table 1 table of variation statistics of data fluctuation coefficients
According to the first column data and the second column data of the table 1, when the first fluctuation index is increased, the data fluctuation coefficient and the first fluctuation index are positively correlated, according to the first column data and the third column data, when the second fluctuation index is increased, the data fluctuation coefficient is increased, the second fluctuation index and the data fluctuation coefficient are positively correlated, according to the first column data and the fourth column data, when the third fluctuation index is increased, the data fluctuation coefficient is increased, the third fluctuation index and the data fluctuation coefficient are positively correlated, and the data fluctuation coefficient is obtained by combining the first fluctuation index, the second fluctuation index and the third fluctuation index, so that the interference of subjective factors can be effectively avoided, and the accuracy of the data fluctuation coefficient of the mental health detection and analysis method is improved.
Specifically, the first coefficient weight is a weight corresponding to a preset first parameter set in a preset database, and represents a value of the influence degree of the first parameter set on the data fluctuation coefficient, when in use, the weight corresponding to the preset first parameter set can be directly obtained from the preset database, and the corresponding relationship can be a preset mapping relationship, for example, a mapping set is formed by an average value of data characteristic parameters in the first parameter set of the history parameter set and a weight corresponding to an average value of data characteristic parameters in the preset first parameter set in the preset database, and the average value of the data characteristic parameters in the real-time first parameter set is input into the mapping set to obtain the corresponding weight, wherein the mapping relationship can be a one-to-one or a many-to-one relationship. The range of values in this example is [0,1].
Specifically, the second coefficient weight is a weight corresponding to a second parameter set preset in the preset database, and represents a value of the influence degree of the second parameter set on the data fluctuation coefficient, when in use, the weight corresponding to the second parameter set can be directly obtained from the preset database, and the corresponding relationship can be a preset mapping relationship, for example, a mapping set is formed by an average value of data characteristic parameters in the second parameter set of the history parameter set and a weight corresponding to an average value of data characteristic parameters in the second parameter set preset in the preset database, and the average value of the data characteristic parameters in the second parameter set in real time is input into the mapping set to obtain the corresponding weight, wherein the mapping relationship can be a one-to-one or a many-to-one relationship. In this example, the range of values is [0,1], and the sum of the third coefficient weight and the first and second coefficient weights is 1.
Further, the specific acquisition process of the data abnormality assessment index comprises the steps of acquiring data assessment weights from a preset database, wherein the data assessment weights comprise a first weight and a second weight, the first weight is used for describing the influence degree of data characteristic parameters on the data abnormality assessment index, the second weight is used for describing the influence degree of student data characteristics on the data abnormality assessment index, the data abnormality assessment index is obtained according to the data characteristic parameters, the student data characteristics, the data fluctuation coefficients and the data assessment weights at preset moments, the data abnormality assessment index and the corresponding preset moments are stored in the preset database, and the calculation formula of the data abnormality assessment index is as follows:
;
In the formula,Evaluating an index for the nth data anomaly of the student to be tested,Is the data characteristic parameter of the nth student to be detected,Evaluating the value characteristic for the nth student to be tested,For the nth sleep duty cycle characteristic of the student to be tested,For the nth student to be tested for data transition quantity characteristics,For the nth student to be tested for performance degradation characteristics,Is the homework descending feature of the nth student to be detected,Is the feature of the nth number of wakefulness of the student to be detected,For the reference data fluctuation coefficient,As a first weight to be used,As a result of the second weight being set,Is an adjustment factor.
In this embodiment, the data characteristic parameter is used to describe the overall situation that student data is abnormal at a preset time, and the higher the data characteristic parameter is, the more serious the abnormal situation of the student to be detected at a specific time is, so that the data abnormality evaluation index is higher; the data evaluation value features are used for reflecting the emotion states of student data in the detection time period, and the higher the data evaluation value features are, the better the emotion states of students to be detected in the detection time period are, so that the data abnormality evaluation index is lower; the data transition quantity characteristic is used for measuring the stability of the emotion state of student data in a detection time period, the higher the data transition quantity characteristic is, the worse the stability of the emotion state of the student to be detected in the detection time period is, the higher the data abnormality evaluation index is, the higher the performance decline characteristic is used for reflecting the performance fluctuation condition of the student data in the detection time period, the bigger the performance fluctuation condition of the student data in the detection time period is, the higher the data abnormality evaluation index is, the lower the homework decline characteristic is used for reflecting the homework fluctuation condition of the student data in the detection time period, the higher the homework decline characteristic is, the higher the instability of the homework completion rate of the student data in the detection time period is, the higher the data abnormality evaluation index is, the higher the wakefulness number characteristic is used for reflecting the wakefulness number of the student data in the detection time period is, the more frequent the wakefulness of the student data in the detection time period is, the higher the data abnormality evaluation index is, the sleep occupation bit is used for measuring whether the actual sleep time to be detected accords with the ideal sleep time period of the student in the detection time period is required by the sleep time period, the higher the sleep occupation bit is the higher the ideal time is, the sleep time to be detected, the sleep time is the higher the sleep time is more in the ideal state is the sleep time is detected, the data anomaly evaluation index is used for describing the anomaly change trend of the student data, the higher the data fluctuation coefficient is, the greater the anomaly change trend of the student data is, the higher the data anomaly evaluation index is, the data evaluation weight is used for measuring the contribution degree of the data characteristic parameters and the student data characteristics to the data anomaly evaluation index, the greater the data evaluation weight is, the more obvious the influence of the corresponding data on the data anomaly evaluation index is, the higher the data anomaly evaluation index is, the comprehensive consideration of the data characteristic parameters, the student data characteristics, the data fluctuation coefficient and the data evaluation weight is used for comprehensively measuring the overall anomaly state of the student data, and the higher the data anomaly evaluation index is, the worse the overall anomaly state of the student data is.
For the average value of the historical time period, for simplifying the independent variable description of the data abnormality evaluation index, the abnormality index is defined as,The n-th abnormal index of the student to be detected is the simplified formulaToThe number of the components is 1.3,The total number of the components is 0.4,The total number of the components is 0.6,For example, 0.5, the statistical table of variation of the data anomaly evaluation index is shown in table 2:
Table 2 variation statistics of data anomaly evaluation index
According to the first, second and third column data of table 2, it can be seen that the data abnormality evaluation index increases when the data fluctuation coefficient increases, and the data abnormality evaluation index decreases when the data fluctuation coefficient decreases, the data fluctuation coefficient and the data abnormality evaluation index are positively correlated, according to the first and fourth column data, it can be seen that the data abnormality evaluation index increases when the data characteristic parameter increases, the data characteristic parameter and the data abnormality evaluation index are positively correlated, according to the first and fifth column data, the data abnormality evaluation index increases, and the abnormality index and the data abnormality evaluation index are positively correlated, and the overall abnormality state of the student data can be intuitively reflected by integrating the data fluctuation coefficient, the data characteristic parameter and the abnormality index, thereby improving the accuracy of the data abnormality evaluation index of the mental health detection and analysis method.
Specifically, the first weight is a weight corresponding to a preset data characteristic parameter in a preset database, and represents a value of the influence degree of the data characteristic parameter on the data abnormality evaluation index, when the method is used, the weight corresponding to the preset data characteristic parameter can be directly obtained from the preset database, the corresponding relation can be a preset mapping relation, for example, the weight corresponding to the data characteristic parameter value in the history period and the data characteristic parameter value in the preset database form a mapping set, the real-time data characteristic parameter value is input into the mapping set to obtain the corresponding weight, and the mapping relation can be in one-to-one correspondence or in many-to-one relation. In this example, the value range is [0,1], and the sum of the second weight and the first weight is 1.
Specifically, the adjustment factor is a factor corresponding to a preset data fluctuation coefficient in a preset database, and represents a value of the adjustment degree of the data fluctuation coefficient to the data abnormality evaluation index, when in use, a weight corresponding to the preset data fluctuation coefficient can be directly obtained from the preset database, and the corresponding relationship can be a preset mapping relationship, for example, a mapping set is formed by the data fluctuation coefficient value of the historical time period and the weight corresponding to the preset data fluctuation coefficient value in the preset database, and the real-time data fluctuation coefficient value is input into the mapping set to obtain the corresponding weight, wherein the mapping relationship can be in one-to-one correspondence or in many-to-one correspondence. The range of values in this example is [0,1].
Further, the grading comprises the specific steps of A1, classifying each student to be detected into corresponding abnormal grades by comparing the data abnormality evaluation index of the student to be detected with a preset evaluation threshold, wherein the preset evaluation threshold comprises a first threshold and a second threshold, the abnormal grades comprise a first grade, a second grade and a third grade, the first grade represents the student to be detected with the data abnormality evaluation index not larger than the first threshold, the second grade represents the student to be detected with the data abnormality evaluation index larger than the first threshold and not larger than the second threshold, the third grade represents the student to be detected with the data abnormality evaluation index larger than the second threshold, the data abnormality groups are obtained by sorting according to the data abnormality evaluation index in the abnormal grades, the data abnormality groups are stored in a preset database, and the sorting represents the sorting from low to high according to the data abnormality evaluation index of the student in each abnormal grade.
In the embodiment, the setting of the preset evaluation threshold value is set according to the data abnormality evaluation index of the historical time period, for example, firstly, the data abnormality evaluation index of the historical time period is obtained from a preset database, the average value and the standard deviation of the data abnormality evaluation index are obtained through analysis, the first threshold value is set as the average value of the data abnormality evaluation index of the historical time period, and the second threshold value is set as the sum of the average value and the standard deviation of the data abnormality evaluation index of the historical time period;
The data visualization method comprises the specific processes of obtaining a data anomaly group, a data anomaly evaluation index and a data fluctuation coefficient from a preset database, drawing a data visualization chart, wherein the data visualization chart comprises a data fluctuation coefficient time sequence chart, a data anomaly evaluation index time sequence chart and a data anomaly component chart, and concretely comprises the steps of drawing the data fluctuation coefficient time sequence chart according to the change relation between the data fluctuation coefficient and the preset moment, drawing the data anomaly evaluation index time sequence chart according to the change relation between the data anomaly evaluation index, a second threshold and the preset moment, drawing the data anomaly component chart according to the change relation between student IDs and anomaly grades, and visually displaying the change trend of the abnormal condition of student data.
In this embodiment, the data visualization chart shows the relationship among the data fluctuation coefficient, the data anomaly evaluation index and the data anomaly group, as shown in fig. 2, which is a time series chart of the data fluctuation coefficient provided by the embodiment of the present application, it can be seen from the chart that, with the lapse of time, the data fluctuation coefficient shows a certain fluctuation, especially, a significant rise occurs at the 6 th preset moment, which indicates that the student data may be affected by some factors and needs to pay further attention, and as shown in fig. 3, which is a time series chart of the data anomaly evaluation index provided by the embodiment of the present application, wherein 0.8 is set as the second threshold value in the preset evaluation threshold values. As can be seen from the figure, when the data abnormality evaluation index is higher than 0.8, the overall abnormality state of the student data is high, particularly at the 8 th preset time, the data abnormality evaluation index is higher than the second threshold, and the related personnel should take measures in time, wherein the measures comprise checking whether the student data is abnormal or not and improving the attention of the student, and as shown in fig. 4, the data abnormality composition diagram provided by the embodiment of the application reflects the abnormality grade conditions of different students. According to the method, the abnormal grades of students to be detected with the student IDs of 1,3 and 9 are higher, related personnel should pay attention to emotion changes of the students preferentially, necessary psychological support and coaching are provided, and the related personnel can identify abnormal student data in a short time by visually displaying the data fluctuation coefficient, the data abnormality evaluation index and the data abnormality group, and timely take targeted measures, so that the efficiency of the psychological health detection and analysis method is effectively improved.
It should be clear that the data fluctuation coefficient and the data abnormality assessment index are not directly equivalent to the poor psychological health condition of the student to be detected, but reflect that abnormality or instability may exist in the student data, and the abnormality may be caused by various factors, such as errors in the data acquisition process, short-term fluctuation of emotion of the student or other non-persistent factors, so when the data fluctuation coefficient and the fluctuation index are found to be high, relevant personnel should intervene in time, firstly check the accuracy and the validity of the student data to ensure that the student data is free of errors, and if the data is truly free of errors, further communicate with the student, understand the emotion state or behavior change of the student, and take appropriate psychological health intervention measures to ensure that the student is supported timely and effectively.
In summary, according to the embodiment of the application, the characteristic extraction is performed on the student data corresponding to the student to be detected to obtain the student data characteristic at the preset moment, the preset model is input to obtain the data characteristic parameter of the student data at the preset moment, then the data characteristic parameter of the historical time period is analyzed to obtain the data fluctuation coefficient, the data abnormality evaluation index is obtained by combining the student data characteristic and the data characteristic parameter at the preset moment, and finally the abnormality classification is performed on the student to be detected by combining the preset evaluation threshold to obtain the data abnormality group, so that the accurate identification and the hierarchical management of the student data with abnormal change are realized, the improvement of the effectiveness of the data visualization in the psychological health detection and analysis is further realized, and the problem that the effectiveness of the psychological health detection and analysis is not high in the data visualization mode in the prior art is effectively solved.
It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as a method, system, or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. It is therefore intended that the following claims be interpreted as including the preferred embodiments and all such alterations and modifications as fall within the scope of the invention.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.