Student mental health assessment system based on big data

Abstract

The invention relates to the technical field of psychological health assessment, in particular to a student psychological health assessment system based on big data, which comprises a health data acquisition module used for acquiring physiological health data and activity data of a target in a monitoring period, and a voice acquisition module used for acquiring a target voice file in the monitoring period, the voice data processing module is used for processing a target voice file, the voice data analysis module is used for constructing emotion weights of all words in a target text word group, the semantic analysis module is used for analyzing a target semantic state, and the health state analysis module is used for analyzing a physical sign state and a vitality state of a target. The health state analysis module is used for analyzing the health state of the target, and the psychological health assessment module is used for assessing the psychological state of the target. According to the invention, the efficiency and accuracy of target mental health assessment are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of mental health assessment, and in particular to a student mental health assessment system based on big data. Background Art

[0002] With the development of society, students are facing increasing pressure, and mental health issues have gradually received widespread attention. Traditional mental health assessment methods mainly rely on questionnaires and personal interviews, which are time-consuming and labor-intensive and difficult to fully cover all students. Therefore, there is an urgent need for a method and technical means to efficiently and accurately assess the mental state of a large number of students.

[0003] Chinese patent publication number CN116453656A discloses a mental health assessment and early warning system and method, the method comprising: using assessment questions to conduct a mental health assessment test on a target user to obtain multiple test results, and during the test process, collecting the target user's physiological parameters for stress analysis to obtain multiple stress analysis results, and judging whether they are greater than a preset stress threshold, if the proportion of unqualified stress analysis results is greater than a preset proportion threshold, then a warning is issued, or if not, an assessment mode is determined according to qualified stress analysis results, and it is input into a corresponding mental health assessment model to obtain a mental health assessment result, and when the mental health assessment result is unqualified, a warning is issued; it can be seen that the invention just analyzes the user's physiological sign parameters to obtain the user's stress level, does not combine big data and the user's voice to judge the user's emotions, and only uses sign parameters to analyze the user's mental state, which has the problems of low efficiency and inaccuracy. Summary of the invention

[0004] The purpose of the present invention is to provide a student mental health assessment system based on big data to solve at least one of the problems existing in the prior art.

[0005] To achieve the above object, the present invention adopts the following technical solutions: A student mental health assessment system based on big data, characterized by comprising: A health data collection module is used to collect physiological health data and activity data of the target within the monitoring period; Voice collection module, used to collect target voice files within the monitoring period; An audio data processing module is used to process the target voice file to obtain a target text word group; A speech data analysis module, for constructing the sentiment weight of each word in the target text word group according to the target text word group and the volume of each word in the target text word group; A semantic analysis module is used to analyze the target semantic state according to the sentiment weight construction result of each word in the target text word group; A health status analysis module is used to analyze the target's physical status and vitality status according to the target's physiological health data and the target's activity data during the monitoring period, and to analyze the target's health status according to the analysis results; The mental health assessment module is used to assess the target's mental state based on the target's semantic state analysis results and the target's health state analysis results during the monitoring period, and to alert the user based on the assessment results.

[0006] Furthermore, the audio data processing module inputs the target voice file into the language model to obtain the target communication text, and generates the target text word group according to the target communication text, and the audio data processing module is also used to record the time of each word in the target text word group; The audio data processing module is also used to extract the volume db(i) of each word in the target text word group according to the target voice file and the target text word group, wherein db(i) represents the volume of the i-th word in the target text word group.

[0007] Furthermore, the speech data analysis module includes a sentiment assignment unit, which is used to generate a target sentiment dictionary based on the historical target text, and assign sentiment vectors to the target text word group based on the target sentiment dictionary to obtain the sentiment vector qg(i) of each word in the target text word, wherein qg(i) represents the sentiment vector of the i-th word in the target text word group.

[0008] Furthermore, the sentiment analysis unit is used to calculate the sentiment index α(i) of each word in the target text word group, and compare the sentiment index α(i) of each word in the target text word group with each preset sentiment index, and classify each word in the target text word group according to the comparison result, and the classification result of each word in the target text word group includes negative words, ordinary words and positive words, and the sentiment weight of the negative word is set to We1, the sentiment weight of the ordinary word is set to We2, and the sentiment weight of the positive word is set to We3.

[0009] Furthermore, the semantic analysis module includes a semantic analysis unit, which is used to calculate a target semantic index β and compare the target semantic index β with each preset semantic index to determine a target semantic state, wherein the target semantic state includes a negative state, a calm state and an excited state.

[0010] Furthermore, the semantic analysis module also includes a real-time state analysis unit, which is used to calculate the target score floating ratio μ based on the target's historical test ranking and the target's most recent test ranking, and compare the target score floating ratio μ with each preset floating ratio to determine the real-time state of the target. The real-time state of the target includes depressed, normal and excited, and when the real-time state of the target is depressed, the first preset semantic index is adjusted to B1'; when the real-time state of the target is excited, the second preset semantic index is adjusted to B2'.

[0011] Furthermore, the semantic analysis module also includes a fatigue analysis unit, which is used to calculate the target's fatigue index β according to the target's study time and sleep time during the monitoring period, and when the target's fatigue index β exceeds a preset fatigue index B, optimize the first preset floating ratio to U1'.

[0012] Further, the health status analysis module includes a physical sign status analysis unit, which is used to calculate the target's physiological standard index γ according to the target's physiological health data, and compare the target's physiological standard index γ with the physiological index threshold value Y to determine the target's physical sign status, and the target's physical sign status includes normal and abnormal; The health status analysis module also includes a vitality status analysis unit, which is used to analyze the vitality status of the target according to the activity data of the target during the monitoring period: if a1×t2 / T+a2×t3 / t1<K, the vitality status analysis unit determines that the vitality status of the target during the monitoring period is abnormal; if a1×t2 / T+a2×t3 / t1≥K, the vitality status analysis unit determines that the vitality status of the target during the monitoring period is normal; wherein a1 is the sleep proportion weight, a2 is the exercise learning proportion weight, t3 is the target exercise duration, a1+a2=1 and a1>a2, t2 is the deep sleep duration, and K is the relief proportion threshold.

[0013] Furthermore, the health status analysis module also includes a health status analysis unit, which is used to analyze the target's health status based on the target's vital sign status analysis results and vitality status analysis results during the monitoring period. The target's health status analysis results include normal, abnormal vital sign, abnormal rest and abnormal health.

[0014] Furthermore, the mental health assessment module is used to assess the mental state of the target according to the target semantic state analysis results and the target health state analysis results within the monitoring period, and to alert the user according to the assessment results: If the target semantic state is calm or excited and the health state is normal, the mental health assessment module determines that the target's mental state is healthy and does not alert the user; If the target semantic state is a negative state and the health state is a physical sign abnormality, the mental health assessment module determines that the target's mental state is a third-level abnormality and warns the user; if the target semantic state is a negative state and the health state is a rest abnormality, the mental health assessment module determines that the target's mental state is a second-level abnormality and warns the user; if the target semantic state is a negative state and the health state is a health abnormality, the mental health assessment module determines that the target's mental state is a first-level abnormality; If the target semantic state is an excited state or a calm state, at this time, when the health state is abnormal physical signs or abnormal health, the mental health assessment module sends a disease alarm to the user.

[0015] Compared with the prior art, the beneficial effects of the present invention are: by performing sentiment analysis on the target voice file to determine the target semantic state, and then by performing multi-dimensional analysis on the target's health status to achieve two-way analysis of the target's physical sign state and fatigue state, finally the target's psychological state is analyzed and an alarm is issued to the user, effectively improving the efficiency and accuracy of the target's psychological health assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] 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.

[0017] Figure 1 This is a schematic diagram of the structure of the student mental health assessment system based on big data in this embodiment.

[0018] Figure 2 Schematic diagram of the structure of the voice data analysis module in this embodiment.

[0019] Figure 3 Schematic diagram of the structure of the semantic analysis module of this embodiment.

[0020] Figure 4 Schematic diagram of the structure of the health status analysis module of this embodiment. DETAILED DESCRIPTION

[0021] In order to more clearly illustrate the present invention, the present invention is further described below in conjunction with preferred embodiments and accompanying drawings. Similar components in the accompanying drawings are represented by the same reference numerals. It should be understood by those skilled in the art that the content specifically described below is illustrative rather than restrictive, and should not be used to limit the scope of protection of the present invention.

[0022] It should be noted that, although the terms first, second, third, etc. may be used to describe in the embodiments of the present application, these descriptions should not be limited to these terms. These terms are only used to distinguish the descriptions. For example, without departing from the scope of the embodiments of the present application, the first may also be referred to as the second, and similarly, the second may also be referred to as the first.

[0023] Specifically, the student mental health assessment system based on big data described in this embodiment is applied to the mental health status assessment of students in high schools. The system described in this embodiment is set in the school psychological assessment server.

[0024] See also Figure 1 As shown, it is a schematic diagram of the structure of the student mental health assessment system based on big data in this embodiment, including: The health data collection module is used to collect the physiological health data and activity data of the target within the monitoring period; the physiological health data of the target includes blood pressure and heart rate; the activity data includes study time, sleep time, deep sleep time and target exercise time.

[0025] Specifically, the target described in this embodiment refers to students; the physiological health data and activity data of the target described in this embodiment are obtained through two methods: sensors built into the wearable smart watch and target interactive input.

[0026] Specifically, this embodiment does not impose any specific limitation on the value of the duration of the monitoring period. Technical personnel in this field can set it freely as long as the value requirements of the duration of the monitoring period are met. In this embodiment, the value of the duration of the monitoring period is 1 week; it can be understood that the physiological health data and activity data of the target during the monitoring period collected by this embodiment are all average data during the monitoring period.

[0027] Please continue reading Figure 1 As shown, the system also includes a voice collection module for collecting target voice files within a monitoring period.

[0028] Specifically, the target voice file described in this embodiment is specifically the voice audio of the target communication, and this embodiment collects the target voice file through an audio receiving device; it should be noted that the target information (including but not limited to the target's physiological health data and activity data) and data (including but not limited to those used for the target voice file) involved in this application are all information and data authorized by the target or fully authorized by all parties, and the collection, use and processing of relevant data need to comply with relevant national and regional laws, regulations and standards.

[0029] See also Figure 1As shown, the system includes an audio data processing module, which is connected to the voice acquisition module, and is used to process the target voice file to obtain a target text word group, and use the target text word group as the basic data for the analysis process of the emotion contained in the target semantics; The audio data processing module inputs the target voice file into a Seq2Seq (Sequence-to-Sequence) model to obtain a target communication text, and analyzes the target text word group according to the target communication text. The audio data processing module is also used to record the time of each word in the target text word group; The audio data processing module is also used to input the time of each word in the target voice file and the target text word group into the librosa library of Python to obtain the volume db(i) of each word in the target text word group, wherein db(i) represents the volume of the i-th word in the target text word group, and extracts the target text word group and the volume of each word in the target text word group to extract the data of the emotion contained in the target voice text, thereby providing data support for the subsequent analysis process of the emotion contained in the target voice.

[0030] It can be understood that the "time of each word in the target text word group" described in this embodiment is the time when each word in the target text word group appears in the target voice file, and the target voice file is a continuous voice audio with a time axis.

[0031] Specifically, the process of "analyzing the target text word group according to the target communication text" in this embodiment is to obtain the text word group after performing a word segmentation operation on the target communication text, and the word segmentation operation is implemented by Jieba.

[0032] Please continue reading Figure 1 As shown, the system also includes a speech data analysis module, which is connected to the audio data processing module, and the speech data analysis module is used to construct the emotional weight of each word in the target text word group based on the target text word group and the volume of each word in the target text word group.

[0033] See also Figure 2 As shown, the speech data analysis module includes a sentiment assignment unit, which is used to assign sentiment vectors to the target text word group, so as to accurately assign the sentiment of the target speech file in combination with the target context; The sentiment assignment unit is used to input the historical target text into the LSTM model to obtain the target sentiment dictionary, and to assign sentiment vectors to the target text word group according to the target sentiment dictionary to obtain the sentiment vector qg(i) of each word in the target text word group, wherein qg(i) represents the sentiment vector of the i-th word in the target text word group; the sentiment assignment unit is combined with the existing technology to generate the sentiment dictionary of each target context, perform big data differentiation analysis on each target, and then accurately assign sentiment to the target text word group, thereby achieving accurate control of the sentiment of each target word, and then achieving accurate classification of each word in the target text word group.

[0034] Specifically, the process of "inputting the historical target text into the LSTM model to obtain the target sentiment dictionary" in this embodiment is the prior art and will not be elaborated in this embodiment; at the same time, in this embodiment; it can be understood that the historical target text described in this embodiment is the target's historical communication voice text, which is collected and stored in the cloud server through a wearable smart watch.

[0035] Please continue reading Figure 2 As shown, the speech data analysis module also includes a sentiment analysis unit, which is connected to the sentiment assignment unit, and the sentiment analysis unit is used to construct a sentiment weight of the target text word group according to the sentiment vector assignment result of the target text word and the volume of each word in the target text word group, and the sentiment weight of the target text word group refers to the sentiment information contained in each word; The sentiment analysis unit calculates the sentiment index α(i) of each word in the target text word group, and sets α(i)=qg(i) / QG×db(i), where QG is the statistical communication volume; The sentiment analysis unit compares the sentiment index α(i) of each word in the target text word group with each preset sentiment index, and classifies each word in the target text word group according to the comparison result, and is also used to construct the sentiment weight of each word in the target text word group according to the classification result: if α(i)<A1, the sentiment analysis unit determines that the word is a negative word, and sets the sentiment weight of the word to We1, setting We1=[α(i)-A1] / A1; if A1≤α(i)<A2, the sentiment analysis unit determines that the word is a normal word. word, and set the sentiment weight of the word to We2, setting We2=0; if α(i)≥A2, the sentiment analysis unit determines that the word is a positive word, and sets the sentiment weight of the word to We3, setting We3=[α(i)-A2] / A2; calculate the sentiment index of each word by volume and the sentiment of the word itself, and classify each word in the existing context according to the calculation result, so as to realize the accurate interpretation of each word in the target text phrase in the target voice file, thereby improving the accuracy of the target semantic state analysis; Wherein, A1 is the first preset emotion index, A2 is the second preset emotion index, and A1<A2.

[0036] It can be understood that, in this embodiment, a negative sentiment weight indicates that the sentiment tends to be negative, and a positive sentiment weight indicates that the sentiment tends to be positive.

[0037] Specifically, this embodiment does not make specific limitations on the values ​​of the first preset emotion index A1 and the second preset emotion index A2. Those skilled in the art can set them freely as long as the value requirements of the first preset emotion index A1 and the second preset emotion index A2 are met. In this embodiment, the optimal value of the first preset emotion index A1 is set to -0.6, and the optimal value of the second preset emotion index A2 is set to 0.9; it can be understood that the value of the statistical communication volume QG described in this embodiment is obtained by collecting the communication voice volume of the target through big data and taking the average value of the collected data.

[0038] Please continue reading Figure 1 As shown, the system also includes a semantic analysis module, which is connected to the speech data analysis module. The semantic analysis module is used to construct the emotional weight of each word in the target text word group according to the volume of each word in the target text word group to analyze the target semantic state.

[0039] See also Figure 3 As shown, the semantic analysis module includes a semantic analysis unit, which is used to construct the emotional weight of each word in the target text word group according to the volume of each word in the target text word group to analyze the target semantic state, and use the target semantic state as a summary of the emotion contained in the target voice file; The semantic analysis unit is used to calculate the target semantic index β, and set β=Σα(i); The semantic analysis unit is used to compare the target semantic index β with each preset semantic index, and analyze the target semantic state according to the comparison result: if β<B1, the semantic analysis unit determines that the target semantic state is a negative state; if B1≤β<B2, the semantic analysis unit determines that the target semantic state is a calm state; if β≥B2, the semantic analysis unit determines that the target semantic state is an excited state; wherein B1 is the first preset semantic index, B2 is the second preset semantic index, and B1<B2.

[0040] Specifically, this embodiment does not impose any specific restrictions on the values ​​of the first preset semantic index B1 and the second preset semantic index B2. Those skilled in the art can set them freely as long as the value requirements of the first preset semantic index B1 and the second preset semantic index B2 are met. In this embodiment, the optimal value of the first preset semantic index B1 is -3, and the optimal value of the second preset semantic index B2 is 2.5.

[0041] Please continue reading Figure 3 As shown, the semantic analysis module also includes a real-time state analysis unit, which is connected to the semantic analysis unit. The real-time state analysis unit is used to calculate the target score floating ratio μ according to the target's historical test ranking and the target's most recent test ranking, and set μ=(PMC-MC) / PMC; wherein MC represents the target's most recent test ranking, and PMC represents the target's historical test ranking; The real-time state analysis unit is used to compare the target performance floating ratio μ with each preset floating ratio, and analyze the real-time state of the target according to the comparison result. The real-time state analysis unit is also used to adjust the analysis process of the semantic state of the target according to the analysis result: if μ<U1, the real-time state analysis unit determines that the real-time state of the target is depressed, and adjusts the first preset semantic index to B1', setting B1'=B1×exp[(U1-μ) / U1]; if U1≤μ<U2, the real-time state analysis unit determines that the real-time state of the target is normal; if μ≥U2, the real-time state analysis unit determines that the real-time state of the target is excited, and adjusts the second preset semantic index to B2', setting B2'=B2×[1+(μ-U2) / U2]; wherein U1 is the first preset floating ratio, and U2 is the second preset floating ratio; the real-time state analysis unit analyzes the floating changes of the target's performance, and then determines the impact of the target performance changes on the target's psychology, thereby improving the accurate judgment of the target's semantic state.

[0042] Specifically, this embodiment does not make specific restrictions on the values ​​of the first preset floating ratio U1 and the second preset floating ratio U2. Those skilled in the art can set them freely as long as the value requirements of the first preset floating ratio U1 and the second preset floating ratio U2 are met. In this embodiment, the first preset floating ratio U1 and the second preset floating ratio U2 are assigned values ​​by means of big data survey statistics. The specific process is to set up a questionnaire, the content of which is whether the target is satisfied with the results, and conduct multiple statistics to determine the satisfaction standard of each target with the results. The satisfactory result with the smallest decrease in the target ranking is taken as the original calculation data of the first preset floating ratio U1, and the satisfactory result with the smallest increase in the target ranking is taken as the original data of the second preset floating ratio U2. The process of calculating the first preset floating ratio U1 and the second preset floating ratio U2 according to the original data is the same as the process of calculating the target score floating ratio μ; it can be understood that the frequency of the target test described in this embodiment is 1 week / time.

[0043] Please continue reading Figure 3As shown, the semantic analysis module also includes a fatigue analysis unit, which is connected to the real-time state analysis unit. The fatigue analysis unit is used to calculate the target's fatigue index β according to the target's learning time and sleep time during the monitoring period. The target's fatigue index refers to the target's learning fatigue degree, and β=t1 / T is set, where t1 is the target's learning time and T is the sleep time; The fatigue analysis unit is used to compare the target's fatigue index β with the preset fatigue index B, and optimize the target's real-time state analysis process according to the comparison result: if β<B, the fatigue analysis unit determines that the target is fatigued during the monitoring period, and optimizes the first preset floating ratio to U1', setting U1'=U1×[1+(β-B) / B]; if β≥B, the fatigue analysis unit determines that the target is normal during the monitoring period and does not perform optimization; by analyzing the target's learning fatigue level, and analyzing the impact of the target's learning fatigue level combined with the target's performance fluctuation on the target's semantic state, the accuracy of the target's semantic state analysis is improved.

[0044] Specifically, this embodiment does not impose any specific limitation on the value of the preset fatigue index B, and those skilled in the art can freely set it as long as the value requirement of the preset fatigue index B is met. In this embodiment, the best value of the preset fatigue index B is 2.

[0045] Please continue reading Figure 1 As shown, the system also includes a health status analysis module, which is connected to the health data acquisition module. The health status analysis module is used to analyze the health status of the target based on the target's physiological health data and the target's activity data during a monitoring period.

[0046] See also Figure 4 As shown, the health status analysis module includes a physical sign status analysis unit, which is used to calculate the target's physiological standard index according to the target's physiological health data, and analyze the target's physical sign status according to the calculation result, and the target's physical sign status refers to the target's physical health level; The physical sign state analysis unit calculates the target's physiological standard index γ according to the target's physiological health data, and sets γ=ln{1+|xy-XY| / XY+|xl-XL| / XL}; wherein xy is the target's average blood pressure during the monitoring period, XY is the standard normal blood pressure, xl is the target's average heart rate during the monitoring period, and XL is the standard normal efficiency; The physical sign status analysis unit compares the target's physiological standard index γ with the physiological index threshold value Y, and analyzes the target's physical sign status based on the comparison result: if γ<Y, the physical sign status analysis unit determines that the target's physical sign status is normal; if γ≥Y, the physical sign status analysis unit determines that the target's physical sign status is abnormal; by analyzing the target's physical sign status in combination with the target's physical sign parameters, the target's status is judged at the physical level, thereby achieving accurate physical level judgment of the target's health status.

[0047] Specifically, this embodiment does not impose any specific limitation on the value of the physiological index threshold value Y, and those skilled in the art can freely set it as long as the value requirement of the physiological index threshold value Y is met. In this embodiment, the optimal value of the physiological index threshold value Y is 0.32; it can be understood that the values ​​of the standard normal blood pressure XY and the standard normal efficiency XL in this embodiment are the average values ​​of the historical normal blood pressure and efficiency of the target within the normal medical reference range.

[0048] Please continue reading Figure 4 As shown, the health status analysis module also includes a vitality status analysis unit, which is used to analyze the vitality status of the target according to the activity data of the target during the monitoring period, and the vitality status of the target refers to whether the target's fatigue and rest are balanced: if a1×t2 / T+a2×t3 / t1<K, the vitality status analysis unit determines that the vitality status of the target during the monitoring period is abnormal; if a1×t2 / T+a2×t3 / t1≥K, the vitality status analysis unit determines that the vitality status of the target during the monitoring period is normal; wherein, a1 is the sleep proportion weight, a2 is the exercise learning proportion weight, t3 is the target exercise duration, a1+a2=1 and a1>a2, t2 is the deep sleep duration, and K is the relief proportion threshold; by analyzing the rest-fatigue level of the body that cannot be expressed by physical sign parameters, an accurate judgment of the rest-fatigue level of the target health status can be achieved.

[0049] Specifically, this embodiment does not impose any specific limitation on the value of the relief ratio threshold K, and those skilled in the art can freely set it as long as the value requirement of the relief ratio threshold K is met. In this embodiment, the optimal value of the relief ratio threshold K is 0.9.

[0050] Please continue reading Figure 4As shown, the health status analysis module also includes a health status analysis unit, which is connected to the vitality status analysis unit and the vital sign status analysis unit, and the health status analysis unit is used to analyze the health status of the target according to the vital sign status analysis results and the vitality status analysis results of the target during the monitoring period: if the target's vital sign status is normal and the vitality status is normal, the health status analysis unit determines that the target's health status is normal; if the target's vital sign status is abnormal and the vitality status is normal, the health status analysis unit determines that the target's health status is abnormal vital sign; if the target's vital sign status is normal and the vitality status is abnormal, the health status analysis unit determines that the target's health status is abnormal rest; if the target's vital sign status is abnormal and the vitality status is abnormal, the health status analysis unit determines that the target's health status is abnormal health; by combining the target's vital sign status analysis results and the vitality status analysis results during the monitoring period, a multi-dimensional analysis of the target's vital sign level and rest-fatigue level is achieved to achieve accurate judgment of the target's health status.

[0051] Please continue reading Figure 1 As shown, the system further includes a mental health assessment module, which is connected to the health status analysis module and the semantic analysis module. The mental health assessment module is used to assess the target's mental state according to the target semantic status analysis results and the target's health status analysis results within the monitoring period, and to alert the user according to the assessment results: If the target semantic state is calm or excited and the health state is normal, the mental health assessment module determines that the target's mental state is healthy and does not alert the user; If the target semantic state is a negative state and the health state is a physical sign abnormality, the mental health assessment module determines that the target's mental state is a third-level abnormality and warns the user; if the target semantic state is a negative state and the health state is a rest abnormality, the mental health assessment module determines that the target's mental state is a second-level abnormality and warns the user; if the target semantic state is a negative state and the health state is a health abnormality, the mental health assessment module determines that the target's mental state is a first-level abnormality; If the target semantic state is an excited state or a calm state, at this time, when the health state is abnormal physical signs or abnormal health, the mental health assessment module sends a disease alarm to the user; when the health state is abnormal rest, the mental health assessment module sends a fatigue alarm to the user; the mental health assessment module combines the target voice and the target health status analysis results to perform a preliminary screening and judgment on the target's mental state, and at the same time, it also realizes effective monitoring of the target's health and fatigue, thereby improving the efficiency and accuracy of the target's mental health assessment.

[0052] Specifically, the first-level abnormality, second-level abnormality and third-level abnormality described in this embodiment are psychological abnormalities from severe to mild, and the third-level abnormality is a mild abnormality, which is a normal negative emotion; the second-level abnormality and the first-level abnormality need to be monitored separately to overcome the psychological abnormality problem; the user described in this embodiment is the target administrator.

[0053] So far, the technical solutions of the present invention have been described in conjunction with the preferred embodiments shown in the accompanying drawings. However, it is easy for those skilled in the art to understand that the protection scope of the present invention is obviously not limited to these specific embodiments. Without departing from the principle of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will fall within the protection scope of the present invention.

Claims

1. A student mental health assessment system based on big data, characterized in that: include: A health data collection module is used to collect physiological health data and activity data of the target within the monitoring period; Voice collection module, used to collect target voice files within the monitoring period; An audio data processing module is used to process the target voice file to obtain a target text word group; A speech data analysis module, for constructing the sentiment weight of each word in the target text word group according to the target text word group and the volume of each word in the target text word group; A semantic analysis module is used to analyze the target semantic state according to the sentiment weight construction result of each word in the target text word group; A health status analysis module is used to analyze the target's physical status and vitality status according to the target's physiological health data and the target's activity data during the monitoring period, and to analyze the target's health status according to the analysis results; The mental health assessment module is used to assess the target's mental state based on the target's semantic state analysis results and the target's health state analysis results during the monitoring period, and to alert the user based on the assessment results.

2. The student mental health assessment system based on big data according to claim 1 is characterized in that: The audio data processing module inputs the target voice file into the language model to obtain the target communication text, and generates the target text word group according to the target communication text, and the audio data processing module is also used to record the time of each word in the target text word group; The audio data processing module is also used to extract the volume db(i) of each word in the target text word group according to the target voice file and the target text word group, wherein db(i) represents the volume of the i-th word in the target text word group.

3. The student mental health assessment system based on big data according to claim 2 is characterized in that: The speech data analysis module includes a sentiment assignment unit, which is used to generate a target sentiment dictionary based on the historical target text, and assign sentiment vectors to the target text word group based on the target sentiment dictionary to obtain the sentiment vector qg(i) of each word in the target text word group, wherein qg(i) represents the sentiment vector of the i-th word in the target text word group.

4. The student mental health assessment system based on big data according to claim 3 is characterized in that: The sentiment analysis unit is used to calculate the sentiment index α(i) of each word in the target text word group, and compare the sentiment index α(i) of each word in the target text word group with each preset sentiment index, and classify each word in the target text word group according to the comparison result. The classification results of each word in the target text word group include negative words, ordinary words and positive words, and the sentiment weight of the negative word is set to We1, the sentiment weight of the ordinary word is set to We2, and the sentiment weight of the positive word is set to We3.

5. The student mental health assessment system based on big data according to claim 4 is characterized in that: The semantic analysis module includes a semantic analysis unit, which is used to calculate a target semantic index β and compare the target semantic index β with each preset semantic index to determine a target semantic state, which includes a negative state, a calm state and an excited state.

6. The student mental health assessment system based on big data according to claim 5 is characterized in that: The semantic analysis module also includes a real-time state analysis unit, which is used to calculate the target score floating ratio μ according to the target's historical test ranking and the target's most recent test ranking, and compare the target score floating ratio μ with each preset floating ratio to determine the real-time state of the target. The real-time state of the target includes depressed, normal and excited, and when the real-time state of the target is depressed, the first preset semantic index is adjusted to B1'; when the real-time state of the target is excited, the second preset semantic index is adjusted to B2'.

7. The student mental health assessment system based on big data according to claim 6 is characterized in that: The semantic analysis module also includes a fatigue analysis unit, which is used to calculate the target's fatigue index β according to the target's learning time and sleeping time during the monitoring period, and when the target's fatigue index β exceeds the preset fatigue index B, optimize the first preset floating ratio to U1'.

8. The student mental health assessment system based on big data according to claim 7 is characterized in that: The health status analysis module includes a physical sign status analysis unit, which is used to calculate the target's physiological standard index γ according to the target's physiological health data, and compare the target's physiological standard index γ with the physiological index threshold value Y to determine the target's physical sign status, and the target's physical sign status includes normal and abnormal; The health status analysis module further includes a vitality status analysis unit, which is used to analyze the vitality status of the target according to the activity data of the target during the monitoring period: if a1×t2 / T+a2×t3 / t1<K, the vitality status analysis unit determines that the vitality status of the target during the monitoring period is abnormal; If a1×t2 / T+a2×t3 / t1≥K, the vitality status analysis unit determines that the vitality status of the target during the monitoring period is normal; wherein a1 is the sleep proportion weight, a2 is the exercise learning proportion weight, t3 is the target exercise duration, a1+a2=1 and a1>a2, t2 is the deep sleep duration, and K is the relief proportion threshold.

9. The student mental health assessment system based on big data according to claim 8 is characterized in that: The health status analysis module also includes a health status analysis unit, which is used to analyze the target's health status according to the target's vital sign status analysis results and vitality status analysis results during the monitoring period. The target's health status analysis results include normal, abnormal vital sign, abnormal rest and abnormal health.

10. The student mental health assessment system based on big data according to claim 9 is characterized in that: The mental health assessment module is used to assess the target's mental state according to the target semantic state analysis results and the target health state analysis results during the monitoring period, and to alert the user according to the assessment results: If the target semantic state is calm or excited and the health state is normal, the mental health assessment module determines that the target's mental state is healthy and does not alert the user; If the target semantic state is a negative state and the health state is a physical sign abnormality, the mental health assessment module determines that the target's mental state is a third-level abnormality and warns the user; if the target semantic state is a negative state and the health state is a rest abnormality, the mental health assessment module determines that the target's mental state is a second-level abnormality and warns the user; if the target semantic state is a negative state and the health state is a health abnormality, the mental health assessment module determines that the target's mental state is a first-level abnormality; If the target semantic state is an excited state or a calm state, at this time, when the health state is abnormal physical signs or abnormal health, the mental health assessment module sends a disease alarm to the user.

Images