Method for dynamically monitoring psychological health conditions of college students and performing early warning based on side writing data

Abstract

The invention discloses a method for dynamically monitoring the psychological health state of college students through side-writing data, and the method comprises the following steps: firstly, carrying out the feature extraction and word vector conversion of the side-writing data and common college student psychological health problems, measuring the similarity of quantized feature words through a clustering method and a cosine distance, and determining a clustering center point; and a K value of clustering is determined. Secondly, according to factors such as social criticality, disease incidence, importance degree and the like, sequencing and weighting the quantized feature words, finally, assigning scores to the extracted features in combination with weights, and setting a quantification standard in a threshold theory;

Description

Technical Field

[0001] The present invention belongs to the cross-technical field of educational assessment and mental health monitoring, and specifically relates to a method for dynamically monitoring the mental health status of college students using profiling data and providing early warning. Background Art

[0002] With the development of higher education and the increasing attention paid to the mental health of college students, timely monitoring of students' mental conditions has become an important task in college management. College students are in a critical period of psychological and social adaptation. Many factors such as academic pressure, interpersonal relationships, and living environment may have a profound impact on their mental health. In order to prevent and intervene in potential psychological problems in a timely manner, colleges and universities are in urgent need of establishing a scientific and real-time mental health monitoring system. Traditional monitoring methods, such as regular psychological questionnaires and individual counseling, have the defects of long feedback cycles and limited coverage, and because students tend to hide their true emotions, these methods are difficult to fully and accurately reflect their psychological state. The present invention proposes a method for dynamically monitoring and early warning the mental health of college students through profiling data, combining data analysis and machine learning technology to construct a real-time monitoring model, which can track students' mental health more comprehensively and objectively, help colleges and universities accurately grasp students' psychological dynamics, discover potential problems early and take effective intervention measures, thereby optimizing teaching design and campus management, and promoting harmonious campus development. Summary of the invention

[0003] 1. Technical issues to be solved

[0004] In view of this, in order to achieve the above-mentioned purpose, the purpose of the present invention is to provide a method for dynamically monitoring the mental health status of college students and providing early warning based on profiling data.

[0005] (II) Technical solution

[0006] In order to achieve the above object, the present invention adopts the following technical solutions:

[0007] 1. A method for dynamically monitoring the mental health status of college students and providing early warnings using profile data, comprising the following steps:

[0008] Step 1: Extract features from profile data and common college students’ mental health problems;

[0009] Step 2: Use Word2vec to convert the extracted features into word vectors through the vector space model;

[0010] Step 3: In the framework of k-means clustering method, the metric distance is modified to cosine distance to measure the similarity of feature words and determine the cluster center point, and the K value of clustering is determined according to the silhouette coefficient graph and elbow rule graph;

[0011] Step 4: Sort and weight the quantified characteristic words of common psychological problems according to factors such as social harm, disease incidence, degree of impact on patients' quality of life, and difficulty of treatment;

[0012] Step 5: Weighted values ​​are assigned to the quantified feature words of the profile data according to the importance of the factors reflected by the profile data;

[0013] Step 6: Modify and weight the mental health assessment value, set the specific quantitative standards in the threshold theory, input the data to be tested, obtain the mental health assessment results in five dimensions: behavior, sense of security, cognition, sociality and emotion, and output the tendencies of different psychological problems.

[0014] 2. Preferably, the questionnaire questions for the profile data in step 1 specifically include: (1) The teaching objectives of this course are clear and the arrangement is clear and reasonable. (2) The teaching content of this course keeps up with the forefront of the subject and social development, which can stimulate interest in learning. (3) The teacher pays attention to the integration of theoretical knowledge and positive values ​​in the teaching process, which has cultivated and improved my moral character. (4) Auxiliary teaching is very helpful for my study. (5) The teacher's words and deeds are appropriate in the teaching process, actively spread excellent culture and positive energy, the classroom order is good, and the students are treated fairly and caringly. (6) The lecturer's lectures are clear and organized, the examples are appropriate, and the key points are highlighted. (7) The lecturer inspired my learning ideas and stimulated my learning enthusiasm through methods such as cases, scenario simulations, and problem discussions. (8) Homework can be corrected by the lecturer in a timely manner, and questions in learning can be answered and guided by the lecturer in a timely manner. (9) The classroom activities are rich in form, and there are opportunities to interact and communicate with teachers and classmates. (10) Through the study of this course, I have mastered the basic theoretical knowledge of the course. (11) Through the study of this course, I have improved my ability to learn independently and to analyze and solve problems by applying knowledge comprehensively. (12) Through the study of this course, I have improved my scientific literacy and humanistic literacy. The options for each question are divided into five levels according to the degree of satisfaction.

[0015] 3. Preferably, the Word2vec model in step 2 is trained mainly through two architectures: CBOW and Skip-Gram. Word2Vec converts words in the text into vector representations with semantic information to complete word quantization.

[0016] 4. Preferably, in step 3, the metric distance is modified to be the cosine distance under the framework of the k-means clustering method to measure the similarity of the feature words and determine the cluster center point.

[0017] The formula for cosine similarity is:

[0018]

[0019] Among them, A and B are vectors representing text, A·B is the dot product of vectors, and ‖A‖×‖B‖ is a vector.

[0020] 5. Preferably, the K value of k-means clustering in step 3 is selected based on the silhouette coefficient diagram and the elbow rule diagram. The K value of the large cluster is selected as 5, and the five clusters are respectively determined as abnormal behavior, abnormal sense of security, abnormal cognition, abnormal social interaction and abnormal emotions. There are still many synonyms obtained, so a secondary cluster analysis is further performed. The results of the secondary cluster analysis are as follows: abnormal behavior mainly includes over-caution, introversion, lack of interest, rigidity and isolation; abnormal sense of security mainly includes fear, dissatisfaction, irritability, tension, guilt and vigilance; cognitive abnormalities mainly include overconfidence, strong control and self-doubt; social abnormalities mainly include hallucinations, indifference and impulsiveness; emotional abnormalities mainly include poor speech and irritability.

[0021] 6. Preferably, the weighting of the quantified characteristic words of common college students' mental health problems in step 4 is weighted according to factors such as social harm, disease incidence, degree of impact on the patient's quality of life, and difficulty of treatment. We classify common college students' mental health problems according to the severity of the disease: severe: schizophrenia, bipolar disorder, moderate: personality disorder (borderline, antisocial, avoidant, etc.), paranoia, mania, post-traumatic stress disorder, depression, mild: attention deficit hyperactivity disorder, obsessive-compulsive disorder, phobia. The weight ratio of psychological problems determined according to different classifications is 26:8:5.

[0022] 7. Preferably, in step 4, the characteristic words quantified for common psychological problems are scored. First, the results of the primary clustering are scored according to the psychological problems of different severity. The scores of each category are calculated according to the number of occurrences of the features. Then, the percentage system is scored according to the proportion to obtain the scores of abnormal behavior, abnormal sense of security, abnormal cognition, abnormal sociality and abnormal emotions, which are 12:34:9:19:26 respectively. Then, in the secondary clustering, the total score of the small cluster is determined according to the score of the primary cluster. According to the silhouette coefficient of each small cluster center, the proportion of different small cluster parts is scored, and the scores of each group of small cluster texts are unified to obtain the scores of each secondary cluster. Among them, the scores of behavioral abnormality include over-caution, introversion, lack of interest, rigidity, and isolation, with scores of 1, 3, 0, 1, and 7 respectively; the scores of security abnormality include fear, dissatisfaction, irritability, tension, guilt, and vigilance, with scores of 5, 10, 2, 2, 8, and 7 respectively; the scores of cognitive abnormality include overconfidence, strong control desire, and self-doubt, with scores of 2, 2, and 5 respectively; the scores of social abnormality include hallucination, indifference, and paranoia, with scores of 4, 11, and 4 respectively; the scores of emotional abnormality include poor speech and irritability, with scores of 9 and 17. Among them, lack of interest will cause the phenomenon of not completing the questionnaire, and there is no score, which will be combined with subsequent analysis.

[0023] 8. Preferably, the quantified feature words of the profiling data in step 5 are weighted, and the profiling data based on the teaching evaluation questionnaire is weighted in importance using the score of each teaching evaluation question.

[0024] 9. Preferably, in step 6, the mental health assessment value is corrected and weighted, and the course score of the teaching evaluation questionnaire is used for weighting, so the correction coefficient k is introduced:

[0025]

[0026] To calculate the mental health assessment value:

[0027] Z=k*(Z1+Z2+Z3+Z4+Z5)

[0028] Z is the mental health assessment value, Z1 is the behavioral abnormality score, Z2 is the sense of security abnormality score, Z3 is the cognitive abnormality score, Z4 is the social abnormality score, and Z5 is the emotional abnormality score.

[0029] (III) Beneficial effects

[0030] The present invention improves the existing mental health monitoring algorithm and proposes a method for dynamically monitoring the mental health status of college students and providing early warnings using profiling data, which can more comprehensively and accurately capture the changing trends of college students' mental health status. Traditional static assessment methods usually have a long cycle and are difficult to reflect students' psychological fluctuations in a timely manner. The present invention integrates multi-dimensional data in the profiling questionnaire, such as the use of emotional vocabulary, score volatility, and participation rate, to establish a more sophisticated real-time monitoring model, which can effectively identify potential mental health problems and improve the efficiency and accuracy of psychological monitoring.

[0031] This dynamic monitoring method provides a new tool for mental health management. By quickly responding to students' psychological changes, it can timely discover and identify student groups that may be at risk of mental health, and then implement targeted intervention measures to avoid further deterioration of psychological problems. At the same time, the dynamic analysis method that integrates multi-dimensional data can not only improve the accuracy of psychological monitoring, but also provide continuous and detailed data support for long-term mental health research.

[0032] The invention has a wide range of applications and can effectively promote the intelligent and precise management of mental health in colleges and universities. It not only provides important guarantees for the mental health management of schools, but also reduces the incidence of academic interruptions and campus crisis events caused by psychological problems through more effective monitoring methods. In addition, the method is highly consistent with all aspects of education management, can provide more comprehensive support for students' psychological development, promote the improvement of teaching quality in colleges and universities, optimize student management and development strategies, and meet the sustainable development goals of the modern education management system. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0034] Figure 1 It is the process of using profiling data to identify common psychological problems among college students;

[0035] Figure 2 is the elbow rule diagram;

[0036] Figure 3 is the silhouette coefficient plot;

[0037] Figure 4 It is the result of a clustering;

[0038] Figure 5 It is the result of partial secondary clustering;

[0039] Figure 6 It is the weighted result of the profiling data;

[0040] Figure 7 It is the result of a clustering feature assignment;

[0041] Figure 8 It is the result of secondary clustering feature assignment;

[0042] Fig. 9 It is an abnormal radar chart of the data indicator to be tested;

[0043] Fig.10 It is a radar chart of the psychological problem tendencies of the data to be tested. DETAILED DESCRIPTION

[0044] In order to enable those skilled in the art to better understand the technical solution of the present invention, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0045] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0046] See also Figure 1-10 , a method for dynamically monitoring the mental health status of college students and providing early warnings using profile data, characterized in that it includes the following steps:

[0047] 1. A method for dynamically monitoring the mental health status of college students and providing early warnings using profile data, comprising the following steps:

[0048] Step 1: Extract features from profile data and common college students’ mental health problems;

[0049] Step 2: Use Word2vec to convert the extracted features into word vectors through the vector space model;

[0050] Step 3: In the framework of k-means clustering method, the metric distance is modified to cosine distance to measure the similarity of feature words and determine the cluster center point, and the K value of clustering is determined according to the silhouette coefficient graph and elbow rule graph;

[0051] Step 4: Sort and weight the quantified characteristic words of common psychological problems according to factors such as social harm, disease incidence, degree of impact on patients' quality of life, and difficulty of treatment;

[0052] Step 5: Weighted values ​​are assigned to the quantified feature words of the profile data according to the importance of the factors reflected by the profile data;

[0053] Step 6: Modify and weight the mental health assessment value, set the specific quantitative standards in the threshold theory, input the data to be tested, obtain the mental health assessment results in five dimensions: behavior, sense of security, cognition, sociality and emotion, and output the tendencies of different psychological problems.

[0054] 2. Preferably, the questionnaire questions for the profile data in step 1 specifically include: (1) The teaching objectives of this course are clear and the arrangement is clear and reasonable. (2) The teaching content of this course keeps up with the forefront of the subject and social development, which can stimulate interest in learning. (3) The teacher pays attention to the integration of theoretical knowledge and positive values ​​in the teaching process, which has cultivated and improved my moral character. (4) Auxiliary teaching is very helpful for my study. (5) The teacher's words and deeds are appropriate in the teaching process, actively spread excellent culture and positive energy, the classroom order is good, and the students are treated fairly and caringly. (6) The lecturer's lectures are clear and organized, the examples are appropriate, and the key points are highlighted. (7) The lecturer inspired my learning ideas and stimulated my learning enthusiasm through methods such as cases, scenario simulations, and problem discussions. (8) Homework can be corrected by the lecturer in a timely manner, and questions in learning can be answered and guided by the lecturer in a timely manner. (9) The classroom activities are rich in form, and there are opportunities to interact and communicate with teachers and classmates. (10) Through the study of this course, I have mastered the basic theoretical knowledge of the course. (11) Through the study of this course, I have improved my ability to learn independently and to analyze and solve problems by applying knowledge comprehensively. (12) Through the study of this course, I have improved my scientific literacy and humanistic literacy. The options for each question are divided into five levels according to the degree of satisfaction.

[0055] 3. Preferably, the Word2vec model in step 2 is trained mainly through two architectures: CBOW and Skip-Gram. Word2Vec converts words in the text into vector representations with semantic information to complete word quantization.

[0056] 4. Preferably, in step 3, the metric distance is modified to be the cosine distance under the framework of the k-means clustering method to measure the similarity of the feature words and determine the cluster center point.

[0057] The formula for cosine similarity is:

[0058]

[0059] Among them, A and B are vectors representing text, A·B is the dot product of vectors, and ‖A‖×‖B‖ is a vector.

[0060] 5. Preferably, the K value of k-means clustering in step 3 is selected based on the silhouette coefficient diagram and the elbow rule diagram. The K value of the large cluster is selected as 5, and the five clusters are respectively determined as abnormal behavior, abnormal sense of security, abnormal cognition, abnormal social interaction and abnormal emotions. There are still many synonyms obtained, so a secondary cluster analysis is further performed. The results of the secondary cluster analysis are as follows: abnormal behavior mainly includes over-caution, introversion, lack of interest, rigidity and isolation; abnormal sense of security mainly includes fear, dissatisfaction, irritability, tension, guilt and vigilance; cognitive abnormalities mainly include overconfidence, strong control and self-doubt; social abnormalities mainly include hallucinations, indifference and impulsiveness; emotional abnormalities mainly include poor speech and irritability.

[0061] 6. Preferably, the weighting of the quantified characteristic words of common college students' mental health problems in step 4 is weighted according to factors such as social harm, disease incidence, degree of impact on the patient's quality of life, and difficulty of treatment. We classify common college students' mental health problems according to the severity of the disease: severe: schizophrenia, bipolar disorder, moderate: personality disorder (borderline, antisocial, avoidant, etc.), paranoia, mania, post-traumatic stress disorder, depression, mild: attention deficit hyperactivity disorder, obsessive-compulsive disorder, phobia. The weight ratio of psychological problems determined according to different classifications is 26:8:5.

[0062] 7. Preferably, in step 4, the characteristic words quantified for common psychological problems are scored. First, the results of the primary clustering are scored according to the psychological problems of different severity. The scores of each category are calculated according to the number of occurrences of the features. Then, the percentage system is scored according to the proportion to obtain the scores of abnormal behavior, abnormal sense of security, abnormal cognition, abnormal sociality and abnormal emotions, which are 12:34:9:19:26 respectively. Then, in the secondary clustering, the total score of the small cluster is determined according to the score of the primary cluster. According to the silhouette coefficient of each small cluster center, the proportion of different small cluster parts is scored, and the scores of each group of small cluster texts are unified to obtain the scores of each secondary cluster. Among them, the scores of behavioral abnormality include over-caution, introversion, lack of interest, rigidity, and isolation, with scores of 1, 3, 0, 1, and 7 respectively; the scores of security abnormality include fear, dissatisfaction, irritability, tension, guilt, and vigilance, with scores of 5, 10, 2, 2, 8, and 7 respectively; the scores of cognitive abnormality include overconfidence, strong control desire, and self-doubt, with scores of 2, 2, and 5 respectively; the scores of social abnormality include hallucination, indifference, and paranoia, with scores of 4, 11, and 4 respectively; the scores of emotional abnormality include poor speech and irritability, with scores of 9 and 17. Among them, lack of interest will cause the phenomenon of not completing the questionnaire, and there is no score, which will be combined with subsequent analysis.

[0063] 8. Preferably, the quantified feature words of the profiling data in step 5 are weighted, and the profiling data based on the teaching evaluation questionnaire is weighted in importance using the score of each teaching evaluation question.

[0064] 9. Preferably, in step 6, the mental health assessment value is corrected and weighted, and the course score of the teaching evaluation questionnaire is used for weighting, so the correction coefficient k is introduced:

[0065]

[0066] To calculate the mental health assessment value:

[0067] Z=k*(Z1+Z2+Z3+Z4+Z5)

[0068] Z is the mental health assessment value, Z1 is the behavioral abnormality score, Z2 is the sense of security abnormality score, Z3 is the cognitive abnormality score, Z4 is the social abnormality score, and Z5 is the emotional abnormality score.

[0069] Implementation examples:

[0070] We used a data profiling questionnaire to distribute and collect data from University C in City A, and selected eleven common college students' mental health problems, including anxiety, depression, phobia, mania, paranoia, obsessive-compulsive disorder, attention deficit hyperactivity disorder, schizophrenia, post-traumatic stress disorder, personality disorder, and bipolar disorder as data for analysis. The process is as follows: Figure 1 , the specific implementation steps are as follows:

[0071] Step 1: Extract features from profile data and common college students' mental health problems. The main features include: 'demanding', 'uninterested', 'hasty decision', 'detail-oriented', 'isolated', 'disrespectful', 'introverted', 'avoidant', 'talkative', 'stereotyped', 'overly cautious', 'fearful', 'uneasy', 'panic', 'terrified', 'frustrated', 'fearful', 'angry', 'desperate', 'nervous', 'fearful', 'excited', 'guilty', 'anxious', 'guilty', 'irritable', 'self-blame', 'confusion', 'worry', 'vigilance', 'timidity', 'alertness', 'helplessness', 'dissatisfaction', 'depression', 'perfectionism', 'self-denial', 'control', 'overconfidence', 'self-doubt', 'lack of confidence', 'lack of empathy', 'indifference', 'stubbornness', 'coldness', 'arrogance', 'indifference', 'indifference', 'loneliness', 'delusion', 'avoidance', 'reticence', 'sensitivity', 'hallucination', 'shame', 'impulsivity', 'paranoidism';

[0072] Step 2: Use Word2vec to convert the extracted features into word vectors through the vector space model;

[0073] Step 3: Under the framework of k-means clustering method, modify the metric distance to cosine distance to measure the similarity of feature words and determine the cluster center point. Determine the K value of clustering according to the silhouette coefficient graph and elbow rule graph, such as Figure 2 , Figure 3 As shown, determine the K value to be 5, such as Figure 4 As shown in the figure, the five types of clustering are used as abnormal indicators, including: abnormal behavior, abnormal sense of security, abnormal cognition, abnormal social interaction, and abnormal emotions. Because the results of the large clustering still contain a large number of synonyms, a secondary clustering is performed, and the results are as follows Figure 5 As shown;

[0074] Step 4: According to factors such as social harm, disease incidence, degree of impact on patients' quality of life, and difficulty of treatment, the diseases are divided into severe, moderate and mild wards. Severe: schizophrenia, bipolar disorder, moderate: personality disorder (borderline, antisocial, avoidant, etc.), paranoia, mania, post-traumatic stress disorder, depression, mild: attention deficit hyperactivity disorder, obsessive-compulsive disorder, phobia. The weight ratio of different categories of psychological problems is 26:8:5. The weights are combined to assign features and obtain the scores of common college students' mental health problems. First, the results of the first clustering are assigned points according to the characteristics of psychological problems of different severity, and the scores of each category are calculated according to the number of occurrences of the features, and then the percentage is assigned according to the proportion. Then, in the second clustering, the total score of the small cluster is determined according to the score of the first cluster. According to the silhouette coefficient of each small cluster center, the proportion of different small cluster parts is assigned, and the scores of each group of small cluster texts are unified to obtain the scores of each secondary cluster. Among them, the scores of behavioral abnormalities include over-caution, introversion, lack of interest, stereotyped, and isolation, with scores of 1, 3, 0, 1, and 7 respectively; the scores of sense of security abnormalities include fear, dissatisfaction, irritability, tension, guilt, and vigilance, with scores of 5, 10, 2, 2, 8, and 7 respectively; the scores of cognitive abnormalities include overconfidence, strong control desire, and self-doubt, with scores of 2, 2, and 5 respectively; the scores of social abnormalities include hallucinations, indifference, and paranoia, with scores of 4, 11, and 4 respectively; the scores of emotional abnormalities include poor speech and irritability, with scores of 9 and 17, such as Figure 7 , Figure 8 As shown;

[0075] Step 5: Use 6, 8, or 10 points to weight the profiling questionnaire according to the importance of the factors reflected in the profiling data, such as Figure 6 As shown;-

[0076] Step 6: Modify and weight the mental health assessment value, and use the course scores of the teaching evaluation questionnaire for weighting, so the correction coefficient k is introduced:

[0077]

[0078] To calculate the mental health assessment value:

[0079] Z=k*(Z1+Z2+Z3+Z4+Z5)

[0080] Z is the mental health assessment value, Z1 is the behavioral abnormality score, Z2 is the sense of security abnormality score, Z3 is the cognitive abnormality score, Z4 is the social abnormality score, and Z5 is the emotional abnormality score.

[0081] According to the psychological health assessment value, combined with the threshold test, the data to be tested is input to obtain the psychological assessment results. The options for the data to be tested are C, B, D, B, E, B, C, B, D, C, D, C. According to the threshold test, the psychological assessment results of the sample students can be obtained, such as Fig. 9 , Fig.10 As shown. The sample's behavioral abnormality score was 31, the sense of security abnormality score was 55, the cognitive abnormality score was 37, the social abnormality score was 53, and the emotional abnormality score was 61. The corresponding psychological problem tendencies were: depression tendency 77%, anxiety tendency 61%, personality disorder tendency 36%, bipolar disorder tendency 18%, and attention deficit hyperactivity disorder tendency 43%. The psychological assessment results of the sample indicate that there is a high tendency of psychological problems, which should be paid attention to and corresponding measures should be taken to intervene and assist in restoring students' mental health.

[0082] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It is obvious to those skilled in the art that the present invention is not limited to the details of the above exemplary examples, and the present invention can be implemented in other specific forms without departing from the spirit or basic features of the present invention. Therefore, no matter from which point of view, the embodiments should be regarded as exemplary and non-restrictive. The scope of the present invention is defined by the attached claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present invention. Any figure mark in the claims should not be regarded as limiting the claims involved.

[0083] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A method for dynamically monitoring the mental health status of college students and providing early warning using profile data, comprising the following steps: Step 1: Extract features from profile data and common college students’ mental health problems; Step 2: Use Word2vec to convert the extracted features into word vectors through the vector space model; Step 3: In the framework of k-means clustering method, the metric distance is modified to cosine distance to measure the similarity of feature words and determine the cluster center point, and the K value of clustering is determined according to the silhouette coefficient graph and elbow rule graph; Step 4: Sort and weight the quantified characteristic words of common psychological problems according to factors such as social harm, disease incidence, degree of impact on patients' quality of life, and difficulty of treatment; Step 5: Weighted values ​​are assigned to the quantified feature words of the profile data according to the importance of the factors reflected by the profile data; Step 6: Modify and weight the mental health assessment value, set the specific quantitative standards in the threshold theory, input the data to be tested, obtain the mental health assessment results in five dimensions: behavior, sense of security, cognition, sociality and emotion, and output the tendencies of different psychological problems.

2. According to claim 1, a method for dynamically monitoring the mental health status of college students and providing early warning using profile data is characterized in that: The questionnaire questions for profiling data in step 1 specifically include: (1) The teaching objectives of this course are clear and the arrangement is clear and reasonable. (2) The teaching content of this course keeps up with the forefront of the discipline and social development, which can stimulate interest in learning. (3) The teacher focuses on the integration of theoretical knowledge and positive values ​​during the teaching process, which has cultivated and improved my moral character. (4) Assisted teaching is very helpful for my learning. (5) The teacher's words and deeds during the teaching process are appropriate, actively spread excellent culture and positive energy, the classroom order is good, and the students are treated fairly and caringly. (6) The lecturer's lectures are clear and organized, with appropriate examples and highlighted key points. (7) The lecturer inspired my learning ideas and stimulated my learning enthusiasm through cases, scenario simulations, problem discussions and other methods. (8) Homework can be corrected by the lecturer in a timely manner, and questions in learning can be answered and guided by the lecturer in a timely manner. (9) Classroom activities are rich in form, and there are opportunities to interact with teachers and classmates. (10) Through the study of this course, I have mastered the basic theoretical knowledge of this course. (11) Through the study of this course, I have improved my ability to learn independently and to analyze and solve problems by applying knowledge comprehensively. (12) Through the study of this course, I have improved my scientific literacy and humanistic literacy. The options for each question are divided into five levels according to the degree of satisfaction.

3. According to claim 1, a method for dynamically monitoring the mental health status of college students and providing early warning using profile data is characterized in that: The Word2vec model in step 2 is mainly trained using two architectures: CBOW and Skip-Gram. Word2Vec converts words in the text into vector representations with semantic information to complete word quantization.

4. According to claim 1, a method for dynamically monitoring the mental health status of college students and providing early warning using profile data is characterized in that: In step 3, the metric distance is modified to cosine distance under the framework of k-means clustering method to measure the similarity of feature words and determine the cluster center point. The formula for cosine similarity is: Among them, A and B are vectors representing text, A·B is the dot product of vectors, and ||A||×||B|| is a vector.

5. According to claim 1, a method for dynamically monitoring the mental health status of college students and providing early warning using profile data is characterized in that: The K value of k-means clustering in step 3 is determined by combining the silhouette coefficient graph and the elbow rule graph. The K value of the large cluster is selected as 5, and the five clusters are determined as abnormal behavior, abnormal sense of security, abnormal cognition, abnormal social interaction, and abnormal emotions. There are still many synonyms, so a secondary cluster analysis is performed. The results of the secondary cluster analysis are as follows: abnormal behavior mainly includes over-caution, introversion, lack of interest, rigidity, and isolation; abnormal sense of security mainly includes fear, dissatisfaction, irritability, tension, guilt, and vigilance; abnormal cognition mainly includes overconfidence, strong control, and self-doubt; abnormal social interaction mainly includes hallucinations, indifference, and impulsivity; abnormal emotions mainly include poor speech and irritability.

6. The method of dynamically monitoring the mental health status of college students and providing early warning using profile data according to claim 1 is characterized in that: The weighting of the quantified characteristic words of common college students' mental health problems in step 4 is weighted according to factors such as social harm, disease incidence, degree of impact on the patient's quality of life, and difficulty of treatment. We classify common college students' mental health problems according to the severity of the disease: severe: schizophrenia, bipolar disorder, moderate: personality disorder (borderline, antisocial, avoidant, etc.), paranoia, mania, post-traumatic stress disorder, depression, mild: attention deficit hyperactivity disorder, obsessive-compulsive disorder, phobia. The weight ratio of psychological problems determined according to different classifications is 26:8:

5.

7. The method of dynamically monitoring the mental health status of college students and providing early warning using profile data according to claim 1 is characterized in that: In step 4, the characteristic words of common psychological problems are scored. First, the results of the first clustering are scored according to the psychological problems of different severity. The scores of each category are calculated according to the number of occurrences of the features. Then, the scores are assigned in percentage according to the proportions, and the scores of abnormal behavior, abnormal sense of security, abnormal cognition, abnormal sociality and abnormal emotions are obtained, which are 12:34:9:19:26 respectively. Then, in the second clustering, the total score of the small cluster is determined according to the score of the first cluster. According to the silhouette coefficient of each small cluster center, the proportion of different small cluster parts is assigned, and the scores of each group of small cluster words are unified to obtain the scores of each secondary cluster. Among them, the scores of behavioral abnormality include over-caution, introversion, lack of interest, rigidity, and isolation, with scores of 1, 3, 0, 1, and 7 respectively; the scores of security abnormality include fear, dissatisfaction, irritability, tension, guilt, and vigilance, with scores of 5, 10, 2, 2, 8, and 7 respectively; the scores of cognitive abnormality include overconfidence, strong control desire, and self-doubt, with scores of 2, 2, and 5 respectively; the scores of social abnormality include hallucination, indifference, and paranoia, with scores of 4, 11, and 4 respectively; the scores of emotional abnormality include poor speech and irritability, with scores of 9 and 17. Among them, lack of interest will cause the phenomenon of not completing the questionnaire, and there is no score, which will be combined with subsequent analysis.

8. The method of dynamically monitoring the mental health status of college students and providing early warning using profile data according to claim 1 is characterized in that: In step 5, the quantified feature words of the profiling data are weighted, and the profiling data based on the teaching evaluation questionnaire are weighted in importance using the score of each teaching evaluation question.

9. The method of dynamically monitoring the mental health status of college students and providing early warning using profile data according to claim 1 is characterized in that: In step 6, the mental health assessment value is corrected and weighted, and the course score of the teaching evaluation questionnaire is used for weighting, so the correction coefficient k is introduced: To calculate the mental health assessment value: Z=k*(Z1+Z2+Z3+Z4+Z5) Z is the mental health assessment value, Z1 is the behavioral abnormality score, Z2 is the sense of security abnormality score, Z3 is the cognitive abnormality score, Z4 is the social abnormality score, and Z5 is the emotional abnormality score.

Images