Method and apparatus for monitoring mental health status

Abstract

This application discloses a method and apparatus for monitoring mental health status. The method includes: obtaining user input association features of a target user, including textual semantic features of the input text; obtaining an initial psychological risk assessment value for the target user based on the user input association features; obtaining a psychological baseline value corresponding to the target user, the psychological baseline value representing the target user's baseline psychological state within a predetermined time period; and determining the target risk level corresponding to the target user based on the magnitude of change of the initial psychological risk assessment value relative to the psychological baseline value. This method integrates user input association features of textual semantic features with the magnitude of change between the user's specific psychological baseline value and the initial assessment value to determine the risk level. It achieves a multi-dimensional and comprehensive assessment while also taking into account individual psychological specificity and dynamic changes. It can accurately identify relatively abnormal risks and positive changes, improving the targeting and accuracy of mental health monitoring.

Description

Technical Field

[0001] This application relates to the field of mental health monitoring technology, specifically to a method for monitoring mental health status. This application also relates to a mental health monitoring device, a terminal device, and a computer-readable storage medium. Background Technology

[0002] Current mental health monitoring technologies use fixed thresholds to assess risk, classifying risk levels based solely on a uniform score range. However, different users exhibit fundamental differences in psychological resilience and emotional tone, making it impossible to identify relatively abnormal situations with fixed standards. For example, a normally optimistic person might suddenly see their score jump from 20 to 50, failing to reflect the individual specificity of a user's psychological state. The lack of flexible risk adjustment logic and the failure to optimize risk assessment based on individual changes, keyword triggers, and other special circumstances result in monitoring remaining merely a superficial score comparison with insufficient accuracy. Furthermore, these technologies cannot specifically capture users' potential psychological risks and are ill-suited to the diverse daily needs of mental health monitoring, requiring improvement in both practicality and adaptability. Summary of the Invention

[0003] This invention provides a method, device, terminal equipment, and computer-readable storage medium for monitoring mental health status, in order to solve the problems in the prior art.

[0004] To address or improve upon the aforementioned technical problems to some extent, according to one aspect of the present invention, a method for monitoring mental health status is provided, comprising: Obtain the user input association features of the target user, wherein the user input association features include the textual semantic features of the input text; Based on the user input association features, the initial psychological risk assessment value of the target user is obtained; Obtain the psychological baseline value corresponding to the target user, wherein the psychological baseline value represents the benchmark value of the target user's psychological state within a predetermined time period; Based on the magnitude of change of the initial psychological risk assessment value relative to the psychological baseline value, the target risk level corresponding to the target user is determined.

[0005] In one implementation, the input-related features further include input behavior features; Based on the user input association features, an initial psychological risk assessment value for the target user is obtained, including: A first psychological risk value is obtained based on the text semantic features, and the first psychological risk value is adjusted based on the input behavioral features to obtain the initial psychological risk assessment value; or, A first psychological risk value is obtained based on the text semantic features, a second psychological risk value is obtained based on the input behavior features, and an initial psychological risk assessment value is obtained based on the first psychological risk value, the second psychological risk value, and their respective weights.

[0006] In one implementation, after obtaining the initial psychological risk assessment value of the target user, the method further includes: obtaining the first risk level corresponding to the initial psychological risk assessment value; Correspondingly, based on the magnitude of change of the initial psychological risk assessment value relative to the psychological baseline value, the target risk level corresponding to the target user is determined, including: If the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, the first risk level is adjusted to a second risk level, and the second risk level is determined as the target risk level corresponding to the target user; or, If the difference between the initial psychological risk assessment value and the psychological baseline value does not exceed a predetermined threshold, then the first risk level is determined as the target risk level corresponding to the target user.

[0007] In one implementation, if the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, the first risk level is adjusted to a second risk level, including: If the initial psychological risk assessment value is lower than the psychological baseline value, and the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, then the first risk level will be downgraded to the second risk level; or, If the initial psychological risk assessment value is greater than the psychological baseline value, and the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, then the first risk level will be upgraded to the second risk level.

[0008] In one implementation, the method further includes: providing a risk warning to the target user based on the target risk level.

[0009] In one implementation, the input behavior features are obtained in the following manner: Feature extraction is performed on the log data of the input method application to obtain the input behavior features, wherein the input method application is the application used by the target user when performing input operations; or, Data is collected from the input interface of the input method application to obtain input operation behavior data, and feature extraction is performed based on the input operation behavior data to obtain the input behavior features.

[0010] In one implementation, the method further includes: analyzing the input text of the target user based on a large language model to obtain the semantic features of the text.

[0011] According to another aspect of the present invention, a mental health monitoring device is provided, comprising: The feature acquisition unit is used to acquire user input association features of the target user, wherein the user input association features include textual semantic features of the input text; The evaluation value acquisition unit is used to obtain the initial psychological risk assessment value of the target user based on the user input association features; A psychological baseline value acquisition unit is used to obtain the psychological baseline value corresponding to the target user, wherein the psychological baseline value represents the benchmark value of the target user's psychological state within a predetermined time period. The risk level determination unit is used to determine the target risk level corresponding to the target user based on the change range of the initial psychological risk assessment value relative to the psychological baseline value.

[0012] According to another aspect of the present invention, a terminal device is provided, including a processor and a memory; wherein the memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the above-described method.

[0013] According to another aspect of the present invention, a computer-readable storage medium is provided having one or more computer instructions stored thereon, which are executed by a processor to implement the above-described method.

[0014] Compared with the prior art, the present invention has the following advantages: The mental health monitoring method provided by this invention includes: obtaining user input association features of a target user, including textual semantic features of the input text; obtaining an initial psychological risk assessment value for the target user based on the user input association features; obtaining a psychological baseline value corresponding to the target user, the psychological baseline value representing the target user's psychological state benchmark value within a predetermined time period; and determining the target risk level corresponding to the target user based on the change range of the initial psychological risk assessment value relative to the psychological baseline value. This method integrates user input association features of textual semantic features with the change range of the user's specific psychological baseline value and the initial assessment value to determine the risk level. It achieves a multi-dimensional and comprehensive assessment while also considering individual psychological specificity and dynamic changes. It can accurately identify relatively abnormal risks and positive changes, improving the targeting and accuracy of mental health monitoring. Furthermore, it relies on strong adaptability to input scenarios and a robust privacy protection mechanism to meet diverse daily monitoring needs. Attached Figure Description

[0015] Figure 1 This is a flowchart of the mental health monitoring method provided in the embodiments of this application; Figure 2This is a unit block diagram of the mental health monitoring device provided in the embodiments of this application; Figure 3 This is a schematic diagram of the logical structure of the terminal device provided in the embodiments of this application. Detailed Implementation

[0016] Many specific details are set forth in the following description to provide a full understanding of this application. However, this application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this application; therefore, this application is not limited to the specific embodiments disclosed below.

[0017] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this specification. The singular forms “a,” “described,” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0018] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this specification, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0019] In this application, the term "multiple" may refer to two or more, and "at least one" may refer to one, two or more.

[0020] The term "and / or" in this application is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this application generally indicates that the preceding and following related objects have an "or" relationship.

[0021] For the purpose of mental health monitoring, this application provides a method for monitoring mental health status, a corresponding mental health monitoring device, a terminal device, and a computer-readable storage medium. The following embodiments provide a detailed description of the above-mentioned method, device, terminal device, and computer-readable storage medium.

[0022] The mental health monitoring method provided in this application embodiment can be applied to a computing device application used for mental health monitoring, which can run on a network platform server. Figure 1 The flowchart of the mental health monitoring method provided in the embodiments of this application is as follows. Figure 1 The method provided in this embodiment will be described in detail. The embodiments described below are used to explain the principle of the method and are not intended to limit actual use.

[0023] like Figure 1 As shown, the mental health monitoring method provided in this embodiment includes the following steps: S101, Obtain the user input association features of the target user.

[0024] This step is used to obtain the user input association features of the target user, which include the textual semantic features of the input text. Textual semantic features are a core dimension for analyzing a user's mental health risk; they refer to the semantically relevant features extracted from the user's input text (or speech-to-text content) that reflect their psychological state. In one implementation, the target user's input text can be analyzed using a Large Language Model (LLM) to obtain semantic features. The core of this approach is to uncover the underlying psychological state through in-depth analysis of the user's real-world language corpus using a LLM. This includes identifying emotional tendencies (positive / negative / neutral), semantic orientations (such as frequent mentions of keywords like "stress," "helplessness," and "insomnia"), and linguistic styles (such as whether the expression is disorganized or whether negative attributions are present). LLM can overcome the limitations of keyword matching and achieve deep semantic understanding, for example, distinguishing between a simple complaint like "I'm a little tired today" and a high-risk expression like "I'm tired every day and have no motivation to live."

[0025] In this embodiment, the text semantic features may specifically include the following categories of indicators, each with a different weight. Different weights represent different psychological tendencies corresponding to the text: Negativity of Emotion: This metric uses a large language model to determine the positive or negative emotional tendency contained in the text, focusing on the intensity of negative emotions expressed by the user. For example, expressions such as "despair," "don't want to live anymore," and "too frustrated to the point of collapse" are judged as having a high degree of negativity; while expressions such as "happy" and "full of hope" correspond to a low degree of negativity. Cognitive Distortion: This measure captures irrational cognitive patterns reflected in user text, namely one-sided, extreme, or distorted judgments about things. For example, statements such as "I fail at everything I do, I'm destined to be a failure in this life" (absolute thinking) or "Everyone looks down on me" (subjective assumption) will be included in the cognitive distortion index, reflecting the degree of bias in the user's cognition. Self-focus: This measures the proportion of self-centered expressions in the text, focusing on whether users are overly concerned with their own state, feelings, or experiences. For example, frequent use of self-centered expressions such as "I feel," "I have such a hard time," and "I am always ignored" will increase the score of this indicator, reflecting the user's level of self-focus. Excessive self-focus may be related to psychological stress and depressive tendencies. Feelings of worthlessness: Extracting semantic tendencies of self-denial and lack of meaning in the text. For example, expressions such as "My life has no meaning" and "Nobody needs me" in the text will be directly associated with this indicator and are one of the key semantic features for judging high risk of mental health in users.

[0026] In one implementation, input-related features also include input behavior features, which are an important dimension for assisting in judging a user's mental health status. That is, based on text semantic analysis focusing on information such as emotion and cognition within the text content, these features are combined with the user's behavioral characteristics during the input process as foundational data for subsequent mental health monitoring. This broadens the analytical dimensions, avoids the one-sidedness of single data points, and comprehensively captures the differences in mental states corresponding to input behaviors. Input behavior features are correlated with a user's emotions and mental state (e.g., hesitation when inputting during anxiety, decreased input motivation when feeling down). Combining these features with text semantic features to calculate the initial psychological risk assessment value makes the judgment of mental health risk more comprehensive and accurate.

[0027] In one implementation, input behavior features are obtained by: extracting features from the log data of an input method application (the application used by the target user when performing input operations); or, collecting data from the input interface of the input method application to obtain input operation behavior data, and then extracting features from this data. In other words, the source of input behavior features is either input method log data or input interface operation data. Specific extractable indicators include: input speed (frequency of normal input / hesitation / pause), number of deletions / modifications (reflecting thought coherence), input time distribution (e.g., high-frequency input late at night may be associated with sleep problems), and frequency of special symbol usage (e.g., frequent use of negative punctuation). These features are external manifestations of the user's subconscious behavior, are objective, and are not affected by the user's subjective statements.

[0028] In this embodiment, the following relevant features can be extracted based on the user's behavior during the input process: Motivation Index: Reflects the user's enthusiasm and initiative when inputting. It can be calculated through behaviors such as input frequency, input speed, and input continuity. For example, high-frequency and continuous input may correspond to a higher motivation index, while long-term low-frequency and intermittent input may indicate insufficient motivation.

[0029] Hesitation and Anxiety Index: This reflects the user's hesitation or anxiety during the input process. It can be extracted through behaviors such as the frequency of modification, the number of deletions, and the duration of pauses (such as the time interval between inputs). For example, frequently modifying text or pausing for a long time before continuing to input may correspond to a higher hesitation and anxiety index.

[0030] Rhythmic Copywriting Index: Focusing on the regularity and stability of input behavior, it is calculated through behavioral characteristics such as the distribution of input time (e.g., whether input is done at fixed times) and the fluctuation of the length of input content (e.g., whether the amount of text entered each time varies too much) to reflect the rhythm of user input behavior and thus relate to the stability of psychological state.

[0031] S102, based on the user input association features, obtain the initial psychological risk assessment value of the target user.

[0032] After obtaining the user input association characteristics of the target user in the above steps, this step is used to obtain the initial psychological risk assessment value of the target user based on the user input association characteristics.

[0033] When user input features include textual semantic features and input behavior features, the initial psychological risk assessment value of the target user can be obtained through the following two methods: Method 1 involves assessing the text's semantic features to obtain a first psychological risk value, which is then adjusted based on input behavior characteristics to arrive at an initial psychological risk assessment value. Specifically, Method 1 employs a logic of basic assessment combined with dynamic adjustment. First, it utilizes a large language model to conduct in-depth analysis of the text's semantic features, accurately assessing dimensions such as emotional negativity and cognitive distortion to obtain a first psychological risk value, which forms the core basis of the initial assessment. Then, it makes targeted adjustments based on input behavior characteristics. For example, if the input behavior exhibits abnormal behaviors such as high hesitation and anxiety indices or low motivation indices, the first psychological risk value is appropriately increased; if the input behavior is regular, consistent, and without obvious abnormalities, the first psychological risk value is maintained or slightly decreased, ultimately forming an initial psychological risk assessment value that closely reflects the user's actual state.

[0034] Method Two involves assessing the first psychological risk value based on text semantic features and the second psychological risk value based on input behavior features. An initial psychological risk assessment value is then obtained based on the first and second psychological risk values ​​and their respective weights. In other words, Method Two follows a dual-dimensional assessment and weighted fusion approach. First, the two types of features are assessed independently: the first psychological risk value is derived from text semantic features, and the second psychological risk value is derived from input behavior features (such as rhythmic copywriting index, hesitation and anxiety index, etc.). Then, weights are assigned to each type of feature based on their importance in the psychological risk judgment (e.g., the weight of text semantic features can be higher than that of input behavior features). Using a pre-defined weighted calculation rule, the two risk values ​​are merged to obtain an initial psychological risk assessment value that comprehensively reflects both text semantics and input behavior dimensions.

[0035] S103, obtain the psychological baseline value corresponding to the target user.

[0036] This step is used to obtain the psychological baseline value for the target user before or after the steps described above. This baseline value represents the target user's baseline psychological state within a predetermined time period. Each user corresponds to a baseline value within a predetermined time period, used for dynamic auxiliary assessment of the user's mental health. For example, a dynamic baseline for each user is dynamically generated based on a 30-day moving average score to determine the degree of deviation of the user's psychological state from their own baseline. Each user corresponds to a baseline value within a predetermined time period, which represents the baseline level of psychological risk in the user's normal state. For instance, the textual semantics of an introverted user may be biased towards negativity, but this is within the normal state; baseline calibration can avoid misjudgment.

[0037] It should be noted that if the predetermined time period is too long, the baseline value may not be able to reflect changes in the user's psychological state in a timely manner; if the time period is too short, the baseline value may be unstable. Therefore, this embodiment adopts an adaptive sliding window mechanism to adjust the window length according to the stability of user data. When a user exhibits abnormal input behavior or inputs abnormal text content (including corpus data), the baseline value can be automatically updated. When calculating the baseline value, the sliding window method can be used, selecting the average ± standard deviation of the user's initial psychological risk assessment values ​​over the most recent N days as the baseline value. For new users, the group baseline value can be used first, and then switched to the individual baseline value after sufficient data has been accumulated.

[0038] Furthermore, the baseline value needs to be dynamically updated over a predetermined period of time (e.g., once a month) to adapt to long-term changes in the user's psychological state (e.g., the baseline value will decrease accordingly after the user has passed the stress period).

[0039] S104. Based on the change in the initial psychological risk assessment value relative to the psychological baseline value, determine the target risk level corresponding to the target user.

[0040] After obtaining the initial psychological risk assessment value and psychological baseline value for the target user in the above steps, this step is used to determine the target risk level corresponding to the target user based on the change of the initial psychological risk assessment value relative to the psychological baseline value. That is, by maintaining a personal dynamic baseline for each user, the assessment of psychological status is based not only on the absolute score of the initial psychological risk assessment value, but also on the degree of deviation of the initial psychological risk assessment value from the personal baseline. The baseline serves to determine the individual's basic level. After the baseline is determined, the target risk level can be determined by combining the absolute score of the initial psychological risk assessment value and its deviation from the psychological baseline value.

[0041] In one implementation, after obtaining the initial psychological risk assessment value of the target user, the method further includes: obtaining a first risk level corresponding to the initial psychological risk assessment value; for example, the initial psychological risk assessment value... Correspondingly, based on the change in the initial psychological risk assessment value relative to the psychological baseline value, the target risk level for the target user is determined, which can specifically refer to: If the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, the first risk level will be adjusted to the second risk level, and the second risk level will be determined as the target risk level for the target user; or, If the difference between the initial psychological risk assessment value and the psychological baseline value does not exceed the predetermined threshold, then the first risk level is determined as the target risk level corresponding to the target user.

[0042] In one implementation, adjusting the first risk level to the second risk level can specifically mean: if the initial psychological risk assessment value is less than the psychological baseline value and the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, then the first risk level is reduced to the second risk level; or, if the initial psychological risk assessment value is greater than the psychological baseline value and the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, then the first risk level is increased to the second risk level.

[0043] For example, in a scenario where the risk level is rising, the user's baseline value is 0.15, the threshold is 0.1, the current R1 is 0.3, the difference is 0.15 > 0.1, the first risk level is "low risk", and it is adjusted to the second risk level "medium risk".

[0044] For the risk reduction scenario, the user baseline value is 0.5 (medium risk), the threshold is 0.1, the current R1=0.35, the difference 0.15>0.1, the first risk level is "medium risk", and it is adjusted to the second risk level "low risk".

[0045] In one implementation, after determining the target risk level for each user, risk warnings must be issued based on that risk level. Different target risk levels correspond to different warnings and intervention measures to ensure the accuracy of the intervention. For example, for low-risk levels, routine monitoring can be used without proactive intervention; only the initial psychological risk assessment value and baseline value changes need to be continuously tracked. For medium-risk levels, mental health science content can be pushed, users can be guided to use built-in emotional support tools (such as meditation audio), or users can be reminded to pay attention to their own state. For high-risk levels, an emergency warning mechanism needs to be triggered, such as notifying the user's emergency contact, connecting with a professional psychological counseling platform, or pushing relevant information to a mental health service institution under compliant conditions (user privacy must be strictly protected). Differentiated intervention based on target risk levels avoids excessive disturbance to low-risk users while ensuring that high-risk users receive timely attention.

[0046] This method comprehensively collects user input-related information, including semantic features in the text that represent user emotions and thoughts, as well as behavioral characteristics during the input process. The objectivity of behavioral features compensates for the subjectivity of semantic features, while the depth of semantic features compensates for the superficiality of behavioral features. The combination of the two improves the accuracy of the assessment and achieves a multi-dimensional comprehensive evaluation. Simultaneously, this method establishes a unique psychological baseline for each user (i.e., a benchmark of psychological state over a period of time), and then compares the initial risk assessment value with the baseline to determine the risk level. This allows for the accurate detection of relatively abnormal situations, such as a sudden deterioration in mood in a normally optimistic person, as well as capturing positive changes in a user's mood. It takes into account individual psychological differences and state fluctuations, making the mental health monitoring process more targeted and accurate. Furthermore, this method is adaptable to various input scenarios, effectively protects user privacy, and can meet diverse daily mental health monitoring needs.

[0047] Furthermore, this method does not require users to actively fill out questionnaires or use specialized equipment; data collection can be completed using commonly used input methods, resulting in high user acceptance and strong monitoring continuity. The objectivity of behavioral features compensates for the subjectivity of semantic features, while the depth of semantic features compensates for the superficiality of behavioral features; the combination of the two improves the accuracy of the assessment.

[0048] The above embodiments provide a method for monitoring mental health status. Correspondingly, another embodiment of this application also provides a device for monitoring mental health status. Since the device embodiment is basically similar to the method embodiment, it is described in a relatively simple way. For details of the relevant technical features, please refer to the corresponding description of the method embodiment provided above. The following description of the device embodiment is merely illustrative.

[0049] Please refer to Figure 2 Understanding this embodiment, Figure 2This is a block diagram of the mental health monitoring device provided in this embodiment, such as... Figure 2 As shown, the mental health monitoring device provided in this embodiment includes: The feature acquisition unit 201 is used to acquire the user input association features of the target user, including the text semantic features of the input text; The evaluation value acquisition unit 202 is used to obtain the initial psychological risk assessment value of the target user based on the user input association features; The psychological baseline value acquisition unit 203 acquires the psychological baseline value corresponding to the target user. The psychological baseline value represents the benchmark value of the target user's psychological state within a predetermined time period. Risk level determination unit 204 is used to determine the target risk level of the target user based on the magnitude of change of the initial psychological risk assessment value relative to the psychological baseline value.

[0050] In one implementation, the input-related features further include input behavior features; Based on the user input association features, an initial psychological risk assessment value for the target user is obtained, including: A first psychological risk value is obtained based on the text semantic features, and the first psychological risk value is adjusted based on the input behavioral features to obtain the initial psychological risk assessment value; or, A first psychological risk value is obtained based on the text semantic features, a second psychological risk value is obtained based on the input behavior features, and an initial psychological risk assessment value is obtained based on the first psychological risk value, the second psychological risk value, and their respective weights.

[0051] In one implementation, after obtaining the initial psychological risk assessment value of the target user, the method further includes: obtaining the first risk level corresponding to the initial psychological risk assessment value; Correspondingly, based on the magnitude of change of the initial psychological risk assessment value relative to the psychological baseline value, the target risk level corresponding to the target user is determined, including: If the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, the first risk level is adjusted to a second risk level, and the second risk level is determined as the target risk level corresponding to the target user; or, If the difference between the initial psychological risk assessment value and the psychological baseline value does not exceed a predetermined threshold, then the first risk level is determined as the target risk level corresponding to the target user.

[0052] In one implementation, if the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, the first risk level is adjusted to a second risk level, including: If the initial psychological risk assessment value is lower than the psychological baseline value, and the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, then the first risk level will be downgraded to the second risk level; or, If the initial psychological risk assessment value is greater than the psychological baseline value, and the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, then the first risk level will be upgraded to the second risk level.

[0053] In one implementation, the method further includes: providing a risk warning to the target user based on the target risk level.

[0054] In one implementation, the input behavior features are obtained in the following manner: Feature extraction is performed on the log data of the input method application to obtain the input behavior features, wherein the input method application is the application used by the target user when performing input operations; or, Data is collected from the input interface of the input method application to obtain input operation behavior data, and feature extraction is performed based on the input operation behavior data to obtain the input behavior features.

[0055] In one implementation, the method further includes: analyzing the input text of the target user based on a large language model to obtain the semantic features of the text.

[0056] This invention also provides a terminal device that can be programmed with the aforementioned mental health monitoring device to execute the mental health monitoring method provided in this invention. Optionally, one possible hardware structure of the terminal device may be as follows: Figure 3 As shown, it includes: at least one processor 301, at least one communication interface 302, at least one memory 303 and at least one communication bus 304; Optionally, communication interface 02 can be an interface of a communication module, such as the interface of a GSM module; Processor 01 may be a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits configured to implement embodiments of the present invention.

[0057] Memory 303 may include high-speed RAM memory, and may also include non-volatile memory, such as at least one disk storage device.

[0058] The memory 303 stores a program, and the processor 01 calls the program stored in the memory 03 to execute the mental health monitoring method provided in this embodiment of the invention.

[0059] In the above embodiments, a method for monitoring mental health status, a device for monitoring mental health status, and a terminal device are provided. Furthermore, another embodiment of this application provides a computer-readable storage medium for implementing the above method. The computer-readable storage medium embodiments provided in this application are described relatively simply; relevant parts can be found in the corresponding descriptions of the above method embodiments. The embodiments described below are merely illustrative. The computer-readable storage medium provided in this embodiment stores computer instructions, which, when executed by a processor, perform the following steps: Obtain the user input association features of the target user, wherein the user input association features include the textual semantic features of the input text; Based on the user input association features, the initial psychological risk assessment value of the target user is obtained; Obtain the psychological baseline value corresponding to the target user, wherein the psychological baseline value represents the benchmark value of the target user's psychological state within a predetermined time period; Based on the magnitude of change of the initial psychological risk assessment value relative to the psychological baseline value, the target risk level corresponding to the target user is determined.

[0060] In one implementation, the input-related features further include input behavior features; Based on the user input association features, an initial psychological risk assessment value for the target user is obtained, including: A first psychological risk value is obtained based on the text semantic features, and the first psychological risk value is adjusted based on the input behavioral features to obtain the initial psychological risk assessment value; or, A first psychological risk value is obtained based on the text semantic features, a second psychological risk value is obtained based on the input behavior features, and an initial psychological risk assessment value is obtained based on the first psychological risk value, the second psychological risk value, and their respective weights.

[0061] In one implementation, after obtaining the initial psychological risk assessment value of the target user, the method further includes: obtaining the first risk level corresponding to the initial psychological risk assessment value; Correspondingly, based on the magnitude of change of the initial psychological risk assessment value relative to the psychological baseline value, the target risk level corresponding to the target user is determined, including: If the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, the first risk level is adjusted to a second risk level, and the second risk level is determined as the target risk level corresponding to the target user; or, If the difference between the initial psychological risk assessment value and the psychological baseline value does not exceed a predetermined threshold, then the first risk level is determined as the target risk level corresponding to the target user.

[0062] In one implementation, if the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, the first risk level is adjusted to a second risk level, including: If the initial psychological risk assessment value is lower than the psychological baseline value, and the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, then the first risk level will be downgraded to the second risk level; or, If the initial psychological risk assessment value is greater than the psychological baseline value, and the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, then the first risk level will be upgraded to the second risk level.

[0063] In one implementation, the method further includes: providing a risk warning to the target user based on the target risk level.

[0064] In one implementation, the input behavior features are obtained in the following manner: Feature extraction is performed on the log data of the input method application to obtain the input behavior features, wherein the input method application is the application used by the target user when performing input operations; or, Data is collected from the input interface of the input method application to obtain input operation behavior data, and feature extraction is performed based on the input operation behavior data to obtain the input behavior features.

[0065] In one implementation, the method further includes: analyzing the input text of the target user based on a large language model to obtain the semantic features of the text.

[0066] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0067] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0068] 1. Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include non-transitory computer-readable media, such as modulated data signals and carrier waves.

[0069] 2. Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can 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, etc.) containing computer-usable program code.

[0070] Although this application discloses preferred embodiments as described above, it is not intended to limit this application. Any person skilled in the art can make possible changes and modifications without departing from the spirit and scope of this application. Therefore, the scope of protection of this application should be determined by the scope defined in the claims of this application.

Claims

1. A method for monitoring mental health status, characterized in that, include: Obtain the user input association features of the target user, wherein the user input association features include the textual semantic features of the input text; Based on the user input association features, the initial psychological risk assessment value of the target user is obtained; Obtain the psychological baseline value corresponding to the target user, wherein the psychological baseline value represents the benchmark value of the target user's psychological state within a predetermined time period; Based on the magnitude of change of the initial psychological risk assessment value relative to the psychological baseline value, the target risk level corresponding to the target user is determined.

2. The method according to claim 1, characterized in that, The input-related features also include input behavior features; Based on the user input association features, an initial psychological risk assessment value for the target user is obtained, including: A first psychological risk value is obtained based on the text semantic features, and the first psychological risk value is adjusted based on the input behavior features to obtain the initial psychological risk assessment value. or, A first psychological risk value is obtained based on the text semantic features, a second psychological risk value is obtained based on the input behavior features, and an initial psychological risk assessment value is obtained based on the first psychological risk value, the second psychological risk value, and their respective weights.

3. The method according to claim 1, characterized in that, After obtaining the initial psychological risk assessment value of the target user, the method further includes: obtaining the first risk level corresponding to the initial psychological risk assessment value; Correspondingly, based on the magnitude of change of the initial psychological risk assessment value relative to the psychological baseline value, the target risk level corresponding to the target user is determined, including: If the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, the first risk level is adjusted to a second risk level, and the second risk level is determined as the target risk level corresponding to the target user; or, If the difference between the initial psychological risk assessment value and the psychological baseline value does not exceed a predetermined threshold, then the first risk level is determined as the target risk level corresponding to the target user.

4. The method according to claim 3, characterized in that, If the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, the first risk level is adjusted to a second risk level, including: If the initial psychological risk assessment value is lower than the psychological baseline value, and the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, then the first risk level will be downgraded to the second risk level; or, If the initial psychological risk assessment value is greater than the psychological baseline value, and the difference between the initial psychological risk assessment value and the psychological baseline value exceeds a predetermined threshold, then the first risk level will be upgraded to the second risk level.

5. The method according to claim 1, characterized in that, Also includes: Risk warnings are issued to the target users based on the target risk level.

6. The method according to claim 2, characterized in that, The input behavior features are obtained in the following way: Feature extraction is performed on the log data of the input method application to obtain the input behavior features, wherein the input method application is the application used by the target user when performing input operations; or, Data is collected from the input interface of the input method application to obtain input operation behavior data, and feature extraction is performed based on the input operation behavior data to obtain the input behavior features.

7. The method according to claim 1 or 2, characterized in that, Also includes: The target user's input text is analyzed using a large language model to obtain the semantic features of the text.

8. A mental health monitoring device, characterized in that, include: The feature acquisition unit is used to acquire user input association features of the target user, wherein the user input association features include textual semantic features of the input text; The evaluation value acquisition unit is used to obtain the initial psychological risk assessment value of the target user based on the user input association features; A psychological baseline value acquisition unit is used to obtain the psychological baseline value corresponding to the target user, wherein the psychological baseline value represents the benchmark value of the target user's psychological state within a predetermined time period. The risk level determination unit is used to determine the target risk level corresponding to the target user based on the change range of the initial psychological risk assessment value relative to the psychological baseline value.

9. A terminal device, characterized in that, Includes processor and memory; among which, The memory is used to store one or more computer instructions, wherein the one or more computer instructions are executed by the processor to implement the method as described in any one of claims 1-7.

10. A computer-readable storage medium storing one or more computer instructions thereon, characterized in that, The instruction is executed by the processor to implement the method as described in any one of claims 1-7.

Images