Artificial intelligence-based mental health intelligent assessment method and system

Abstract

The artificial intelligence-based mental health intelligent evaluation method and system provided in the application recall similar candidate will behavior description information from multiple reference will behavior description information, on the one hand, improve the recall efficiency, on the other hand, since part of the will behavior description information carries a preference knowledge fragment, the knowledge fragment of part of the will behavior description information can be more accurately captured, and the recall accuracy is improved. Then, according to the second sharing coefficient of the candidate will behavior description information corresponding to the user psychological state data to be processed, the user mental health intelligent evaluation result is screened out from the candidate will behavior description information, on the one hand, the recognition range can be reduced, on the other hand, the user mental health intelligent evaluation result more similar to the user psychological state data to be processed in the whole can be screened out from the candidate will behavior description information.

Description

Technical Field

[0001] This application relates to the field of data assessment technology, and more specifically, to an intelligent assessment method and system for mental health based on artificial intelligence. Background Technology

[0002] Mental health refers to a good or normal state in all aspects and processes of the mind. The ideal state of mental health is to maintain a sound personality, normal intelligence, correct cognition, appropriate emotions, rational will, positive attitude, appropriate behavior, and good adaptation.

[0003] Mental health is influenced by both genetics and environment, especially the parenting style of the family of origin during early childhood, which has a significant impact on the development of mental health. Mental health is characterized by the ability to maintain good communication and cooperation with others in social, productive, and daily life activities, and to handle various situations that arise in life effectively.

[0004] With the continuous development and progress of science and technology, artificial intelligence technology is involved in an increasingly wide range of fields. When artificial intelligence technology is combined with mental health assessment, it can improve the efficiency of the assessment. However, in actual operation, the accuracy of the assessment is low. Therefore, there is an urgent need for a technical solution to improve the above-mentioned technical problems. Summary of the Invention

[0005] To address the technical problems existing in related technologies, this application provides an intelligent assessment method and system for mental health based on artificial intelligence.

[0006] Firstly, it provides an artificial intelligence-based intelligent assessment method for mental health, including: Obtain user psychological state data that needs to be processed; The user psychological state data that needs to be processed is analyzed and processed by important knowledge fragments to obtain multiple partial descriptions of will behavior carrying preference knowledge fragments; For each partial description of will behavior, obtain multiple reference descriptions of will behavior corresponding to the partial description of will behavior, and a first sharing coefficient between the partial description of will behavior and each reference description of will behavior; Based on the first sharing coefficient, one or more candidate will behavior descriptions that match the partial will behavior descriptions among the multiple reference will behavior descriptions are determined, and a second sharing coefficient is determined for the candidate will behavior descriptions corresponding to the user psychological state data that needs to be processed. By combining the second sharing coefficient, the user's mental health intelligent assessment result is determined from the candidate will behavior description information.

[0007] In this application, obtaining multiple reference will behavior descriptions corresponding to the partial will behavior description information includes: Knowledge fragments are extracted from the aforementioned partial information describing willpower behavior to obtain knowledge fragments of willpower behavior description information; Initial knowledge fragment variables that match the knowledge fragment describing the will behavior are selected from a plurality of preset initial knowledge fragment variables and determined as reference knowledge fragment variables; the plurality of initial knowledge fragment variables are extracted from a plurality of preset initial will behavior description information. The reference knowledge fragment variables are subjected to tracing processing of will behavior description information to obtain the initial will behavior description information corresponding to the reference knowledge fragment variables; The initial will behavior description information corresponding to the reference knowledge fragment variable is determined as the reference will behavior description information.

[0008] In this application, the step of obtaining an initial knowledge fragment variable that matches the knowledge fragment describing the will behavior from a plurality of preset initial knowledge fragment variables and determining it as a reference knowledge fragment variable includes: Obtain the candidate sharing coefficients between the knowledge fragments describing the will behavior and the multiple initial knowledge fragment variables; The multiple initial knowledge fragment variables are sorted according to the order set by the candidate sharing coefficients to obtain the sorted initial knowledge fragment variables; The first X initial knowledge fragment variables in the sorted initial knowledge fragment variables are determined as the initial knowledge fragment variables that match the information knowledge fragment describing the will behavior, where X is a positive integer; The initial knowledge fragment variable that matches the knowledge fragment describing the will behavior is determined as the reference knowledge fragment variable.

[0009] In this application, the step of obtaining candidate sharing coefficients between the knowledge fragments describing the will behavior and the plurality of initial knowledge fragment variables includes: For each initial knowledge fragment variable, the initial knowledge fragment variable is processed by the image of the control capability to obtain a partial three-dimensional coordinate system corresponding to the initial knowledge fragment variable. The partial three-dimensional coordinate system includes multiple partial variables with the same dimension, and the dimension of the partial variables is smaller than the dimension of the initial knowledge fragment variable. The partial three-dimensional coordinate system is classified by a preset classification network to obtain the classification criterion of the partial three-dimensional coordinate system; By compressing the partial three-dimensional coordinate system using the classification criterion, the category compression corresponding to the partial three-dimensional coordinate system is obtained; By compressing the aforementioned types and the knowledge fragments describing the will-behavior information, candidate sharing coefficients are determined between the knowledge fragments describing the will-behavior information and the initial knowledge fragment variables.

[0010] In this application, obtaining the first sharing coefficient between the partial will behavior description information and each reference will behavior description information includes: For the reference will behavior description information, identify the number of reference knowledge fragment variables corresponding to the reference will behavior description information; If the number of variables is one, then the reference knowledge fragment variable corresponding to the reference will behavior description information and the will behavior description information knowledge fragment corresponding to the partial will behavior description information are used to calculate the sharing coefficient to obtain the first sharing coefficient between the partial will behavior description information and the reference will behavior description information. If there are multiple variables, then the multiple reference knowledge fragment variables corresponding to the reference will behavior description information are respectively used to calculate the sharing coefficient with the will behavior description information knowledge fragment corresponding to the partial will behavior description information to obtain multiple initial sharing coefficients. Then, the multiple initial sharing coefficients are fused to obtain the first sharing coefficient between the partial will behavior description information and the reference will behavior description information.

[0011] In this application, determining one or more candidate will-behavior descriptions that match the partial will-behavior descriptions among the plurality of reference will-behavior descriptions, in conjunction with the first sharing coefficient, includes: The multiple reference will behavior description information are sorted according to the order set by the first sharing coefficient to obtain sorted reference will behavior description information; The first a reference will behavior descriptions in the sorted reference will behavior descriptions are determined as the candidate will behavior descriptions, where a is a positive integer.

[0012] In this application, determining the second sharing coefficient of the candidate will behavior description information corresponding to the user psychological state data to be processed, in conjunction with the first sharing coefficient, includes: For each candidate will-behavior description, the first shared coefficients corresponding to each part of the will-behavior description are fused to obtain the second shared coefficients corresponding to the candidate will-behavior description and the user psychological state data that needs to be processed.

[0013] In this application, the step of fusing the first shared coefficients of the candidate will-behavior description information corresponding to each part of the will-behavior description information to obtain the second shared coefficients of the candidate will-behavior description information corresponding to the user psychological state data to be processed includes: Obtain the fusion weight of each part of the candidate will behavior description information and the number of multiple parts of the will behavior description information; The first shared coefficient of each part of the will behavior description information corresponding to the candidate will behavior description information is weighted by the fusion weight to obtain the fusion shared coefficient. The second sharing coefficient is obtained by comparing the fusion sharing coefficient with the number of partial will behavior description information.

[0014] In this application, the user's psychological state data to be processed undergoes important knowledge fragment analysis to obtain multiple partial descriptions of will and behavior carrying preference knowledge fragments, including: Identify the element description fragments in the user psychological state data that needs to be processed, wherein the element description fragments are the time periods in which knowledge fragments in the user psychological state data that need to be processed change. The user's psychological state data that needs to be processed is analyzed and processed by the element description fragments to obtain multiple partial will behavior description information carrying preference knowledge fragments.

[0015] In this application, identifying the element description fragments in the user's psychological state data that needs to be processed includes: Knowledge fragments are extracted from the user psychological state data that needs to be processed to obtain a knowledge fragment ranking, where each knowledge fragment in the knowledge fragment ranking corresponds to a time period in the user psychological state data that needs to be processed. The sorting of the knowledge fragments is divided into a first sub-knowledge fragment sorting and a second sub-knowledge fragment sorting. Matrix calculations are performed on the sorting of the first sub-knowledge fragment, the sorting of the second sub-knowledge fragment, and the sorting of the knowledge fragment respectively to obtain the first matrix corresponding to the sorting of the first sub-knowledge fragment, the second matrix corresponding to the sorting of the second sub-knowledge fragment, and the third matrix corresponding to the sorting of the knowledge fragment. Based on the first matrix, the second matrix, and the third matrix, the maximum weight value is determined; if the maximum weight value meets a preset condition, the time period corresponding to the last knowledge segment in the first sub-knowledge segment sorting is determined as the element description segment.

[0016] In this application, determining the user's mental health intelligent assessment result from the candidate willpower behavior description information by combining the second sharing coefficient includes: The candidate will-behavior description information with the largest second shared coefficient among all candidate will-behavior description information matched with the multiple reference will-behavior description information is determined as the user's mental health intelligent assessment result.

[0017] Secondly, an artificial intelligence-based intelligent mental health assessment system is provided, comprising a processor and a memory that communicate with each other, wherein the processor is used to read a computer program from the memory and execute it to implement the above-mentioned method.

[0018] The AI-based intelligent assessment method and system for mental health provided in this application embodiment can, after acquiring user mental state data to be processed, split the user mental state data to be processed into multiple partial will behavior description information carrying preference knowledge fragments. Then, for each partial will behavior description information, multiple reference will behavior description information corresponding to that partial will behavior description information are obtained, as well as a first sharing coefficient between the partial will behavior description information and each reference will behavior description information. Then, based on the first sharing coefficient, one or more candidate will behavior description information that matches the partial will behavior description information among the multiple reference will behavior description information are determined, and a second sharing coefficient is determined for the candidate will behavior description information corresponding to the user mental state data to be processed. Finally, based on the second sharing coefficient, the user mental health intelligent assessment result is determined from the candidate will behavior description information. In the embodiments of this application, for each part of the will-behavioral description information of the user's psychological state data to be processed, some similar reference will-behavioral description information can be initially recalled based on this part of the will-behavioral description information. Then, based on the first sharing coefficient between this part of the will-behavioral description information and each reference will-behavioral description information, similar candidate will-behavioral description information is recalled from multiple reference will-behavioral description information. On the one hand, this improves recall efficiency; on the other hand, since some will-behavioral description information carries preference knowledge fragments, the knowledge fragments of some will-behavioral description information can be captured more accurately, thus improving recall accuracy. Then, based on the second sharing coefficient of the candidate will-behavioral description information corresponding to the user's psychological state data to be processed, the user's mental health intelligent assessment result is screened from the candidate will-behavioral description information. On the one hand, this narrows the identification range; on the other hand, it allows the selection of user mental health intelligent assessment results that are more similar to the user's psychological state data to be processed overall from the candidate will-behavioral description information. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a flowchart of an artificial intelligence-based intelligent assessment method for mental health provided in an embodiment of this application. Detailed Implementation

[0021] To better understand the above technical solutions, the technical solutions of this application will be described in detail below with reference to the accompanying drawings and specific embodiments. It should be understood that the embodiments of this application and the specific knowledge fragments in the embodiments are detailed descriptions of the technical solutions of this application, rather than limitations on the technical solutions of this application. In the absence of conflict, the embodiments of this application and the technical knowledge fragments in the embodiments can be combined with each other.

[0022] Please see Figure 1 This paper presents an intelligent assessment method for mental health based on artificial intelligence, which may include the technical solutions described in steps 101-105.

[0023] 101. Obtain user psychological state data that needs to be processed.

[0024] For example, the basic meaning of mental health refers to a good or normal state in all aspects and processes of the mind. The ideal state of mental health is maintaining a complete personality, normal intelligence, correct cognition, appropriate emotions, rational will, positive attitude, appropriate behavior, and good adaptation. In contrast to mental health are sub-health and mental illness. Mental health has different meanings from different perspectives, and its measurement standards also vary.

[0025] 102. Perform important knowledge fragment analysis on the user psychological state data that needs to be processed to obtain multiple partial descriptions of will behavior carrying preference knowledge fragments.

[0026] Among these, the analysis and processing of important knowledge fragments refers to the decomposition and extraction of different knowledge fragments from the information describing will behavior. Knowledge fragments can be understood as features.

[0027] Among them, some descriptive information of will behavior carries a preference knowledge fragment indicating that the descriptive information fragment of the partial will behavior is relatively independent in terms of knowledge fragment, that is, each partial descriptive information of will behavior contains specific knowledge fragment information or content, and does not need to rely on the entire user psychological state data or other partial descriptive information of will behavior to understand its meaning.

[0028] In one possible implementation, when the user's psychological state data to be processed does not carry knowledge fragments, the user's psychological state data to be processed can be analyzed and processed by important knowledge fragments of will behavior description information to obtain multiple partial will behavior description information carrying preference knowledge fragments. That is, different parts of the split partial will behavior description information can carry different will behavior description information knowledge fragments.

[0029] In one possible implementation embodiment, in step 102, the specific implementation of the step "analyzing and processing the user's psychological state data to be processed to obtain multiple partial descriptions of will and behavior carrying preference knowledge fragments" may include: A1. Identify the element description fragments in the user's psychological state data that need to be processed. The element description fragments are the time periods in which the knowledge fragments in the user's psychological state data that need to be processed change.

[0030] In one possible implementation, a pre-trained element description fragment prediction thread can be used to identify element description fragments in information describing volitional behavior. This thread can output corresponding element description fragments based on user psychological state data processed according to input requirements. For example, when training the element description fragment prediction thread, a large number of examples of volitional behavior description information can be prepared first. Then, element description fragments in these examples are labeled. The labeled examples are then input into an initial thread for training. The goal is to enable the initial thread to accurately identify element description fragments in the volitional behavior description information, thus obtaining the element description fragment prediction thread. Optionally, the initial thread can be a suitable thread for processing volitional behavior description information signals, such as a recurrent neural network (RNN), a convolutional neural network (CNN), or a deep neural network (DNN). These threads can be used to learn patterns and knowledge fragments in the volitional behavior description information. Optionally, supervised learning methods can be used during the training of the initial thread, adjusting thread parameters using training data.

[0031] In an alternative embodiment, the specific implementation of step A1, "identifying the element description fragments in the user's mental state data that need to be processed," may include: A11a. Extract knowledge fragments from the user psychological state data that needs to be processed to obtain a knowledge fragment ranking. Each knowledge fragment in the knowledge fragment ranking corresponds to a time period in the user psychological state data that needs to be processed.

[0032] A12a. Sort the knowledge fragments into sorting the first sub-knowledge fragment and sorting the second sub-knowledge fragment.

[0033] A13a. Perform matrix calculations on the sorting of the first sub-knowledge fragment, the sorting of the second sub-knowledge fragment, and the sorting of the knowledge fragments respectively to obtain the first matrix corresponding to the sorting of the first sub-knowledge fragment, the second matrix corresponding to the sorting of the second sub-knowledge fragment, and the third matrix corresponding to the sorting of the knowledge fragments.

[0034] A14a. Determine the maximum weight value based on the first matrix, the second matrix, and the third matrix. If the maximum weight value meets the preset conditions, then determine the time period corresponding to the last knowledge segment in the first sub-knowledge segment sorting as the element description segment.

[0035] In an alternative embodiment, the specific implementation of step A1, "identifying the element description fragments in the user's mental state data that need to be processed," may include: A11b. Extract psychological quality knowledge fragments from the user psychological state data that needs to be processed to obtain multiple psychological quality knowledge fragments. Each psychological quality knowledge fragment corresponds to a time period in the user psychological state data that needs to be processed, and the multiple psychological quality knowledge fragments are arranged in order of time period from first to last.

[0036] The number and dimensions of psychological quality knowledge fragments can be customized according to actual needs, and are not limited here.

[0037] A12b. Determine one or more target psychological quality knowledge segments from multiple psychological quality knowledge segments. The target psychological quality knowledge segments are used to divide the multiple psychological quality knowledge segments into multiple sets of psychological quality knowledge segments, and the psychological quality knowledge segments in each set of psychological quality knowledge segments satisfy a Gaussian distribution.

[0038] In one possible implementation, taking a sorting of multiple psychological quality knowledge fragments as an example, a specific implementation method for determining one or more target psychological quality knowledge fragments from multiple psychological quality knowledge fragments may include: selecting a psychological quality knowledge fragment from the sorting of psychological quality knowledge fragments and determining it as a reference psychological quality knowledge fragment.

[0039] Then, the psychological quality knowledge fragments that are listed in the reference psychological quality knowledge fragments and those that are listed in the ranking of psychological quality knowledge fragments are identified as the set of psychological quality knowledge fragments to be tested, and it is determined whether the set of psychological quality knowledge fragments to be tested satisfies a Gaussian distribution.

[0040] If the conditions are met, the reference psychological quality knowledge fragment can be identified as the target psychological quality knowledge fragment. The reference psychological quality knowledge fragment and the psychological quality knowledge fragments preceding it in the ranking of psychological quality knowledge fragments are then removed from the ranking of psychological quality knowledge fragments to obtain a new ranking of psychological quality knowledge fragments. Based on the new ranking of psychological quality knowledge fragments, the above steps of "selecting a psychological quality knowledge fragment from the ranking of psychological quality knowledge fragments to identify as the reference psychological quality knowledge fragment" are repeated until there is no set of psychological quality knowledge fragments to be tested that satisfy a Gaussian distribution in the new ranking of psychological quality knowledge fragments.

[0041] If the condition is not met, the adjacent psychological quality knowledge fragments that follow the reference psychological quality knowledge fragments in the psychological quality knowledge fragment sorting are determined as new reference psychological quality knowledge fragments. Based on the new reference psychological quality knowledge fragments, the above steps of "determining the reference psychological quality knowledge fragments and the psychological quality knowledge fragments that precede the reference psychological quality knowledge fragments in the psychological quality knowledge fragment sorting as the set of psychological quality knowledge fragments to be tested, and determining whether the set of psychological quality knowledge fragments to be tested satisfies a Gaussian distribution" are repeated until the set of psychological quality knowledge fragments to be tested satisfies a Gaussian distribution.

[0042] In this embodiment, psychological quality knowledge segments are selected one by one from multiple psychological quality knowledge segments to determine reference psychological quality knowledge segments. A ranking of psychological quality knowledge segments to be tested is constructed based on the reference psychological quality knowledge segments. Then, it is identified whether the ranking of the psychological quality knowledge segments to be tested satisfies a Gaussian distribution to determine whether the reference psychological quality knowledge segment is the target psychological quality knowledge segment. In this way, the target psychological quality knowledge segment can be accurately determined from multiple psychological quality knowledge segments, avoiding the omission of the target psychological quality knowledge segment.

[0043] In an alternative embodiment, in step A12b, the specific implementation of the step "determining one or more target psychological quality knowledge fragments from multiple psychological quality knowledge fragments" may include: A121b. Select a candidate psychological quality knowledge fragment from multiple psychological quality knowledge fragments.

[0044] In one possible implementation, a psychological quality knowledge segment can be randomly selected from multiple psychological quality knowledge segments to determine a candidate psychological quality knowledge segment. When selecting a candidate psychological quality knowledge segment, the first and last psychological quality knowledge segments in the multiple segments can be ignored. For example, if multiple psychological quality knowledge segments form a sequence, when selecting a candidate psychological quality knowledge segment, the segment closest to the middle of the sequence can be preferentially selected as a candidate psychological quality knowledge segment.

[0045] A122b. Based on the candidate psychological quality knowledge fragments, classify multiple psychological quality knowledge fragments into a first set of psychological quality knowledge fragments and a second set of psychological quality knowledge fragments; the first set of psychological quality knowledge fragments includes candidate psychological quality knowledge fragments and psychological quality knowledge fragments whose time period precedes that of the candidate psychological quality knowledge fragments; the second set of psychological quality knowledge fragments includes psychological quality knowledge fragments whose time period follows that of the candidate psychological quality knowledge fragments.

[0046] A123b. Based on the first set of psychological quality knowledge fragments, the second set of psychological quality knowledge fragments, and multiple psychological quality knowledge fragments, determine the evaluation value corresponding to the candidate psychological quality knowledge fragment; the evaluation value represents the probability that both the first set of psychological quality knowledge fragments and the second set of psychological quality knowledge fragments satisfy a Gaussian distribution.

[0047] Specifically, in step A123b, the specific implementation method for determining the evaluation value corresponding to the candidate psychological quality knowledge segment based on the first set of psychological quality knowledge segments, the second set of psychological quality knowledge segments, and multiple psychological quality knowledge segments may include: A1231b. Based on the first set of psychological quality knowledge fragments, the second set of psychological quality knowledge fragments, and multiple psychological quality knowledge fragments, determine the maximum weight value corresponding to the candidate psychological quality knowledge fragment. The maximum weight value represents the initial probability that both the first set of psychological quality knowledge fragments and the second set of psychological quality knowledge fragments satisfy a Gaussian distribution.

[0048] In one possible implementation, the first set of psychological quality knowledge fragments, the second set of psychological quality knowledge fragments, and multiple psychological quality knowledge fragments can be calculated using matrix calculation formulas to obtain a first matrix corresponding to the first set of psychological quality knowledge fragments, a second matrix corresponding to the second set of psychological quality knowledge fragments, and a third matrix corresponding to multiple psychological quality knowledge fragments.

[0049] Then, the determinants of the first matrix, the second matrix, and the third matrix are calculated respectively to obtain the first determinant of the first matrix, the second determinant of the second matrix, and the third determinant of the third matrix.

[0050] Then, the product of the logarithm of the first determinant and the number of psychological quality knowledge segments in the first set of psychological quality knowledge segments is determined as the first calculated value; the product of the logarithm of the second determinant and the number of psychological quality knowledge segments in the second set of psychological quality knowledge segments is determined as the second calculated value; and the product of the logarithm of the third determinant and the number of psychological quality knowledge segments in multiple sets of psychological quality knowledge segments is determined as the third calculated value.

[0051] Finally, by subtracting the first calculated value from the third calculated value and then subtracting the second calculated value, we can obtain the maximum weight value corresponding to the candidate psychological quality knowledge fragment.

[0052] A1232b: Obtain the number of knowledge fragments and the dimension of the knowledge fragments of multiple psychological quality knowledge fragments, and determine the outliers based on the number of knowledge fragments and the dimension of the knowledge fragments. The outliers are used to adjust the initial probability.

[0053] A1233b. Determine the evaluation value based on the maximum weight value and outliers.

[0054] In one possible implementation, the weight corresponding to the outlier can be obtained, and then the product of the weight and the outlier can be calculated to obtain a fourth calculated value. The evaluation value can be obtained by subtracting the fourth calculated value from the maximum weight value.

[0055] A124b. If the evaluation value is the maximum value in the range of values ​​corresponding to the evaluation value, then the candidate psychological quality knowledge fragment is determined as the target psychological quality knowledge fragment.

[0056] Among them, when the evaluation value is the maximum value in the range of values ​​corresponding to the evaluation value, it indicates that the first set of psychological quality knowledge fragments and the second set of psychological quality knowledge fragments are most likely to satisfy the Gaussian distribution, that is, the time period corresponding to the candidate psychological quality knowledge fragments that divide the first set of psychological quality knowledge fragments and the second set of psychological quality knowledge fragments is most likely to be the element description fragment.

[0057] In this embodiment, by searching for element description fragments, the target psychological quality knowledge fragment corresponding to the element description fragment can be quickly found among multiple psychological quality knowledge fragments, thereby improving the efficiency of determining the target psychological quality knowledge fragment.

[0058] A13b. Determine the time period corresponding to the knowledge fragment of the target psychological qualities as the element description fragment.

[0059] A2. Based on the element description fragments, perform important knowledge fragment analysis on the user psychological state data that needs to be processed to obtain multiple partial will behavior description information carrying preference knowledge fragments.

[0060] 103. For each partial description of will behavior, obtain multiple reference descriptions of will behavior corresponding to the partial description of will behavior, and the first sharing coefficient between the partial description of will behavior and each reference description of will behavior.

[0061] The reference will-behavior description information can be benchmark will-behavior description information used for comparison with the user's psychological state data that needs to be processed. For example, compared to the user's psychological state data that needs to be processed, the benchmark will-behavior description information can be known, complete, and free of noise. Optionally, the reference will-behavior description information can be pre-stored in the aforementioned will-behavior description information database.

[0062] Each partial description of will behavior can be pre-mapped with multiple reference descriptions of will behavior. The reference descriptions of will behavior corresponding to the partial descriptions of will behavior can be descriptions of will behavior that carry similar knowledge fragments to the partial descriptions of will behavior.

[0063] In one possible implementation, in step 103, the specific implementation of obtaining multiple reference will behavior description information corresponding to partial will behavior description information may include: S1, extracting knowledge fragments from partial will behavior description information to obtain will behavior description information knowledge fragments.

[0064] In one possible implementation embodiment, a pre-defined knowledge fragment extraction unit for describing will behavior can be used to extract knowledge fragments from a portion of the will behavior description information. Optionally, the extracted knowledge fragments of will behavior description information can be matched with the aforementioned reference knowledge fragment variables in the knowledge fragment dimension to facilitate the calculation of the sharing coefficient between the two.

[0065] S2. Select the initial knowledge fragment variable that matches the knowledge fragment describing the will behavior from multiple preset initial knowledge fragment variables and determine it as the reference knowledge fragment variable; the multiple initial knowledge fragment variables are extracted from multiple preset initial will behavior description information.

[0066] In one possible implementation, step S2 may include: S21, obtaining candidate sharing coefficients between the knowledge fragment describing will behavior and multiple initial knowledge fragment variables.

[0067] In one possible implementation, the sharing coefficients of the information fragment describing the will behavior can be directly compared with the sharing coefficients of each of the multiple initial knowledge fragment variables to obtain candidate sharing coefficients.

[0068] In an alternative embodiment, reference knowledge fragment variables can be obtained through a variable recall method. Therefore, a specific implementation of obtaining the candidate sharing coefficients between the will behavior description information knowledge fragments and multiple initial knowledge fragment variables in S21 may include: S211. For each initial knowledge fragment variable, the initial knowledge fragment variable is processed by the image of the control capability to obtain a partial three-dimensional coordinate system corresponding to the initial knowledge fragment variable. The partial three-dimensional coordinate system includes multiple partial variables with the same dimension, and the dimension of the partial variables is smaller than the dimension of the initial knowledge fragment variable.

[0069] S212. Classify a portion of the three-dimensional coordinate system using a pre-defined classification network to obtain the classification datum for the portion of the three-dimensional coordinate system.

[0070] For example, partitioning Xethods include the a-XEANS algorithm, the a-XEDOIDS algorithm, and the CLARANS algorithm.

[0071] S213. Based on the classification benchmark, some three-dimensional coordinate systems are compressed to obtain the category compression corresponding to some three-dimensional coordinate systems.

[0072] S214. Based on category compression and knowledge fragments describing will behavior, determine the candidate sharing coefficients between knowledge fragments describing will behavior and initial knowledge fragment variables.

[0073] Optionally, the classification benchmark and the knowledge fragment describing will behavior in a portion of the three-dimensional coordinate system can be selected first through category compression to calculate the sharing coefficient. Then, all variables in the portion of the three-dimensional coordinate system are compared to obtain variables similar to the knowledge fragment describing will behavior. The sharing coefficient can be calculated by the inner product of variables.

[0074] Optionally, the knowledge fragment describing will behavior can also be segmented into partial variables describing will behavior, and then the sum of the differences between each partial variable describing will behavior and the corresponding image variables of control ability in the three-dimensional coordinate system can be calculated to obtain candidate sharing coefficients.

[0075] S22. Sort the multiple initial knowledge fragment variables according to the order set by the candidate sharing coefficients to obtain the sorted initial knowledge fragment variables.

[0076] S23. The first X initial knowledge fragment variables in the sorted initial knowledge fragment variables are determined as the initial knowledge fragment variables that match the information knowledge fragment describing will behavior, where X is a positive integer.

[0077] S24. Determine the initial knowledge fragment variable for matching the knowledge fragment of the information describing will behavior as the reference knowledge fragment variable.

[0078] S3. Perform tracing processing on the will behavior description information of the reference knowledge fragment variables to obtain the initial will behavior description information corresponding to the reference knowledge fragment variables.

[0079] The system allows for the storage of multiple pre-defined initial knowledge fragment variables in a pre-defined variable library. This library contains these variables and a variable mapping table, which records the initial will behavior description information corresponding to each initial knowledge fragment variable. The initial knowledge fragment variables are extracted from their corresponding initial will behavior description information. Therefore, after determining the initial will behavior description information, the corresponding initial knowledge fragment variable can be found in the pre-defined variable library based on the variable mapping table. One initial will behavior description information can correspond to one or more initial knowledge fragment variables.

[0080] S4. Determine the initial will behavior description information corresponding to the reference knowledge fragment variable as the reference will behavior description information.

[0081] In one possible implementation, in step 103, the specific implementation of obtaining the first shared coefficient between the partial will behavior description information and each reference will behavior description information may include: for the reference will behavior description information, identifying the number of reference knowledge fragment variables corresponding to the reference will behavior description information.

[0082] If there is only one variable, the sharing coefficient is calculated between the reference knowledge fragment variable corresponding to the reference will behavior description information and the will behavior description information knowledge fragment corresponding to the partial will behavior description information to obtain the first sharing coefficient between the partial will behavior description information and the reference will behavior description information.

[0083] If there are multiple variables, the multiple reference knowledge fragment variables corresponding to the reference will behavior description information are respectively used to calculate the sharing coefficient with the will behavior description information knowledge fragments corresponding to the partial will behavior description information to obtain multiple initial sharing coefficients. The multiple initial sharing coefficients are then fused to obtain the first sharing coefficient between the partial will behavior description information and the reference will behavior description information.

[0084] For example, the specific method for obtaining the first sharing coefficient can be as follows: B1. Retrieve the reference knowledge fragment variable corresponding to each reference will behavior description information from the preset variable library.

[0085] The information describing the reference will behavior can correspond to one or more reference knowledge fragment variables.

[0086] B2. Extract knowledge fragments from some information describing will behavior to obtain knowledge fragments describing will behavior.

[0087] B3. For each reference will behavior description information, calculate the sharing coefficient between the reference knowledge fragment variable and the will behavior description information knowledge fragment corresponding to the reference will behavior description information to obtain the first sharing coefficient between the partial will behavior description information and the reference will behavior description information.

[0088] When there are multiple reference knowledge fragment variables corresponding to the reference will behavior description information, in step B3, the sharing coefficient is calculated for the reference knowledge fragment variables corresponding to the reference will behavior description information and the will behavior description information knowledge fragment to obtain the first sharing coefficient between the partial will behavior description information and the reference will behavior description information. The specific implementation of this method may include: calculating the sharing coefficient for the will behavior description information knowledge fragment and the reference will behavior description information corresponding to multiple reference knowledge fragment variables respectively to obtain multiple initial sharing coefficients.

[0089] By fusing multiple initial shared coefficients, a first shared coefficient is obtained between partial will behavior description information and reference will behavior description information.

[0090] 104. Based on the first sharing coefficient, determine one or more candidate will behavior descriptions that match a portion of the will behavior descriptions among multiple reference will behavior descriptions, and determine the second sharing coefficient of the user psychological state data that needs to be processed corresponding to the candidate will behavior descriptions.

[0091] In one possible implementation, in step 104, determining one or more candidate will behavior descriptions that match a partial will behavior description based on a first sharing coefficient may include: identifying reference will behavior descriptions with a first sharing coefficient greater than or equal to a sharing coefficient threshold as candidate will behavior descriptions. For example, the first sharing coefficients between partial will behavior description x and reference will behavior descriptions Z1, Z2, and Z3 are 0.5, 0.2, and 0.3, respectively. If the sharing coefficient threshold is 0.3, then reference will behavior descriptions Z1 and Z3 can be identified as candidate will behavior descriptions corresponding to partial will behavior description x.

[0092] In an alternative embodiment, in step 104, the specific implementation of determining one or more candidate will behavior descriptions that match a portion of the will behavior descriptions among the multiple reference will behavior descriptions according to the first sharing coefficient may include: sorting the multiple reference will behavior descriptions according to the order set by the first sharing coefficient to obtain sorted reference will behavior descriptions.

[0093] The first 'a' reference will behavior descriptions in the sorted reference will behavior descriptions are determined as candidate will behavior descriptions, where 'a' is a positive integer.

[0094] In one possible implementation, the reference will behavior description information can also be directly identified as the candidate will behavior description information.

[0095] In one possible implementation, in step 104, the specific implementation of determining the second shared coefficient of the user psychological state data that needs to be processed corresponding to the candidate will behavior description information based on the first shared coefficient may include: for each candidate will behavior description information, fusing the first shared coefficient of each part of the will behavior description information corresponding to the candidate will behavior description information to obtain the second shared coefficient of the user psychological state data that needs to be processed corresponding to the candidate will behavior description information.

[0096] In one possible implementation, the specific implementation of the step "fusing the first shared coefficients of each part of the candidate will behavior description information to obtain the second shared coefficients of the user psychological state data to be processed corresponding to the candidate will behavior description information" may include: obtaining the fusion weights of each part of the candidate will behavior description information and the number of multiple parts of the will behavior description information; weighting the first shared coefficients of each part of the candidate will behavior description information based on the fusion weights to obtain the fusion shared coefficients; and calculating the comparison between the fusion shared coefficients and the number of multiple parts of the will behavior description information to obtain the second shared coefficients.

[0097] 105. Based on the second sharing coefficient, determine the user's mental health intelligent assessment results from the candidate will behavior description information.

[0098] This application employs the concept of artificial intelligence (AI), which is described as follows: Artificial Intelligence (AI) is the theory, methods, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to achieve optimal results. In other words, AI is a comprehensive technology within computer science that attempts to understand the essence of intelligence and produce a new type of intelligent machine capable of reacting in a manner similar to human intelligence. AI also studies the design principles and implementation methods of various intelligent machines, enabling them to possess perception, reasoning, and decision-making capabilities.

[0099] Artificial intelligence (AI) is a comprehensive discipline encompassing a wide range of fields, including both hardware and software technologies. Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies primarily encompass computer vision, speech processing, natural language processing, and machine learning / deep learning.

[0100] In the embodiments of this application, the main artificial intelligence software technologies involved include the aforementioned speech processing technologies and machine learning. For example, it may involve speech recognition technology (AutoXaticSpeech Recognition, ASR) in speech technology, including speech signal preprocessing, speech signal frequency analyzing, speech signal feature extraction, speech signal feature matching / recognition, and speech training.

[0101] Artificial intelligence (AI) is a technology that uses digital computers to simulate human perception of the environment, acquisition of knowledge, and use of that knowledge. This technology enables machines to possess functions similar to human perception, reasoning, and decision-making. Fundamental AI technologies generally include sensors, dedicated AI chips, cloud computing, distributed storage, big data processing, operating / interactive systems, and mechatronics. AI software technologies mainly encompass computer vision, speech processing, natural language processing, machine learning / deep learning, autonomous driving, and intelligent transportation.

[0102] In one possible implementation, in step 105, the specific implementation of determining the user's mental health intelligent assessment result from the candidate will behavior description information according to the second sharing coefficient may include: determining the candidate will behavior description information with the largest second sharing coefficient among all candidate will behavior description information matched by multiple reference will behavior description information as the user's mental health intelligent assessment result.

[0103] In one possible implementation embodiment, after determining the user's mental health intelligent assessment result, the user's mental health intelligent assessment result can also be output. Optionally, in addition to outputting the user's mental health intelligent assessment result, information describing the user's will and behavior corresponding to the user's mental health intelligent assessment result can also be output.

[0104] As can be seen, in this embodiment, after acquiring the user's psychological state data to be processed, the data is divided into multiple partial will-behavioral descriptions carrying preference knowledge fragments. For each partial will-behavioral description, multiple reference will-behavioral descriptions are obtained, along with a first sharing coefficient between each partial will-behavioral description and each reference will-behavioral description. Then, based on the first sharing coefficient, one or more candidate will-behavioral descriptions matching the partial will-behavioral descriptions are determined, along with a second sharing coefficient corresponding to the user's psychological state data to be processed for each candidate will-behavioral description. Finally, based on the second sharing coefficient, the user's mental health intelligent assessment result is determined from the candidate will-behavioral descriptions. Because for each partial will-behavioral description of the user's psychological state data to be processed, similar candidate will-behavioral descriptions are recalled from multiple reference will-behavioral descriptions based on the first sharing coefficient between each partial will-behavioral description and each reference will-behavioral description, recall efficiency is improved. Furthermore, since the partial will-behavioral descriptions carry preference knowledge fragments, the knowledge fragments of the partial will-behavioral descriptions can be captured more accurately, thus improving recall accuracy. Then, based on the second sharing coefficient of the user's psychological state data that needs to be processed corresponding to the candidate will-behavioral description information, the user's mental health intelligent assessment results are filtered from the candidate will-behavioral description information. This not only narrows down the identification range but also allows for the selection of user mental health intelligent assessment results that are more similar overall to the user's psychological state data that needs to be processed from the candidate will-behavioral description information. This improves the accuracy and efficiency of will-behavioral description information identification.

[0105] Based on the above, an artificial intelligence-based intelligent mental health assessment device is provided, the device comprising: The data acquisition module is used to acquire user psychological state data that needs to be processed. The data acquisition module is used to perform important knowledge fragment analysis on the user psychological state data that needs to be processed, and obtain multiple partial will behavior description information carrying preference knowledge fragments; The first coefficient acquisition module is used to acquire, for each partial will behavior description information, multiple reference will behavior description information corresponding to the partial will behavior description information, and a first shared coefficient between the partial will behavior description information and each reference will behavior description information. The second coefficient acquisition module is used to combine the first shared coefficient to determine one or more candidate will behavior description information that matches the partial will behavior description information among the multiple reference will behavior description information, and to determine the second shared coefficient of the candidate will behavior description information corresponding to the user psychological state data that needs to be processed. The problem data determination module is used to combine the second shared coefficient to determine the user's mental health intelligent assessment result from the candidate will behavior description information.

[0106] Based on the above, an artificial intelligence-based intelligent mental health assessment system is presented, including a processor and a memory that communicate with each other. The processor is used to read computer programs from the memory and execute them to implement the above-described method.

[0107] Based on the above, a computer-readable storage medium is also provided, on which a computer program stored implements the above method during runtime.

[0108] In summary, based on the above scheme, this application embodiment can, after obtaining the user's psychological state data that needs to be processed, split the user's psychological state data into multiple partial will behavior description information carrying preference knowledge fragments, and then, for each partial will behavior description information, obtain multiple reference will behavior description information corresponding to that partial will behavior description information, as well as a first sharing coefficient between the partial will behavior description information and each reference will behavior description information; then, based on the first sharing coefficient, determine one or more candidate will behavior description information that matches the partial will behavior description information among the multiple reference will behavior description information, and determine a second sharing coefficient for the user's psychological state data that needs to be processed corresponding to the candidate will behavior description information; finally, based on the second sharing coefficient, determine the user's mental health intelligent assessment result from the candidate will behavior description information.

[0109] In the embodiments of this application, for each part of the will-behavioral description information of the user's psychological state data to be processed, some similar reference will-behavioral description information can be initially recalled based on this part of the will-behavioral description information. Then, based on the first sharing coefficient between this part of the will-behavioral description information and each reference will-behavioral description information, similar candidate will-behavioral description information is recalled from multiple reference will-behavioral description information. On the one hand, this improves recall efficiency; on the other hand, since some will-behavioral description information carries preference knowledge fragments, the knowledge fragments of some will-behavioral description information can be captured more accurately, thus improving recall accuracy. Then, based on the second sharing coefficient of the candidate will-behavioral description information corresponding to the user's psychological state data to be processed, the user's mental health intelligent assessment result is screened from the candidate will-behavioral description information. On the one hand, this narrows the identification range; on the other hand, it allows the selection of user mental health intelligent assessment results that are more similar to the user's psychological state data to be processed overall from the candidate will-behavioral description information.

[0110] It should be understood that the systems and modules described above can be implemented in various ways. For example, in some embodiments, the systems and modules can be implemented by hardware, software, or a combination of both. The hardware portion can be implemented using dedicated logic; the software portion can be stored in memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated-design hardware. Those skilled in the art will understand that the methods and systems described above can be implemented using computer-executable instructions and / or included in processor control code, for example, on a media such as a disk, CD, or DVD-ROX, a programmable memory such as read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The systems and modules of this application can be implemented not only by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field-programmable gate arrays, programmable logic devices, etc., but also by software executed by various types of processors, or by a combination of the aforementioned hardware circuits and software (e.g., firmware).

[0111] It should be noted that different embodiments may produce different beneficial effects. In different embodiments, the beneficial effects may be any one or a combination of the above, or any other possible beneficial effects.

Claims

1. An intelligent assessment method for mental health based on artificial intelligence, characterized in that, include: Obtain user psychological state data that needs to be processed; The user psychological state data that needs to be processed is analyzed and processed by important knowledge fragments to obtain multiple partial descriptions of will behavior carrying preference knowledge fragments; For each partial will-behavior description information, multiple reference will-behavior description information corresponding to the partial will-behavior description information are obtained, as well as a first sharing coefficient between the partial will-behavior description information and each reference will-behavior description information; wherein, the first sharing coefficient is used to characterize a first similarity, each reference will-behavior description information is understood as a data example in the database, and the similarity between the partial will-behavior description information and the data example in the database; Combining the first sharing coefficient, one or more candidate will-behavior description information that matches the partial will-behavior description information among the multiple reference will-behavior description information are determined, and a second sharing coefficient is determined for the candidate will-behavior description information corresponding to the user psychological state data that needs to be processed; wherein, the second sharing coefficient is used to characterize the second similarity and to determine the ecological problems in the user psychological state data that needs to be processed; By combining the second sharing coefficient, the user's mental health intelligent assessment result is determined from the candidate will behavior description information.

2. The AI-based intelligent assessment method for mental health as described in claim 1, characterized in that, The step of combining the first sharing coefficient to determine one or more candidate will behavior descriptions that match the partial will behavior descriptions among the plurality of reference will behavior descriptions includes: The multiple reference will behavior description information are sorted according to the order set by the first sharing coefficient to obtain sorted reference will behavior description information; The first a reference will behavior descriptions in the sorted reference will behavior descriptions are determined as the candidate will behavior descriptions, where a is a positive integer.

3. The artificial intelligence-based intelligent assessment method for mental health as described in claim 1, characterized in that, Based on the first sharing coefficient, the second sharing coefficient corresponding to the candidate will behavior description information and the user psychological state data to be processed is determined, including: For each candidate will-behavior description, the first shared coefficients corresponding to each part of the will-behavior description are fused to obtain the second shared coefficients corresponding to the candidate will-behavior description and the user psychological state data that needs to be processed.

4. The artificial intelligence-based intelligent assessment method for mental health as described in claim 1, characterized in that, The user psychological state data that needs to be processed is analyzed for important knowledge fragments to obtain partial descriptions of will and behavior carrying multiple knowledge fragments of preferences, including: Identify the element description fragments in the user's psychological state data that need to be processed. The element description fragments are the time periods in which the knowledge fragments in the user's psychological state data that need to be processed change. Based on the element description fragments, the user psychological state data that needs to be processed is analyzed and processed by important knowledge fragments to obtain multiple partial will behavior description information carrying preference knowledge fragments. This includes identifying descriptive fragments of user psychological state data that need to be processed, including: The psychological quality knowledge fragments are extracted from the user psychological state data that needs to be processed, resulting in multiple psychological quality knowledge fragments. Each of the multiple psychological quality knowledge fragments corresponds to a time period in the user psychological state data that needs to be processed, and the multiple psychological quality knowledge fragments are arranged in the order of the time periods from first to last. One or more target psychological quality knowledge segments are determined from multiple psychological quality knowledge segments. The target psychological quality knowledge segments are used to divide the multiple psychological quality knowledge segments into multiple sets of psychological quality knowledge segments, and the psychological quality knowledge segments in each set of psychological quality knowledge segments satisfy a Gaussian distribution.

5. The artificial intelligence-based intelligent assessment method for mental health as described in claim 1, characterized in that, The acquisition of multiple reference will behavior descriptions corresponding to the partial will behavior description information includes: Knowledge fragments are extracted from the aforementioned partial information describing willpower behavior to obtain knowledge fragments describing willpower behavior. Initial knowledge fragment variables that match the knowledge fragment describing the will behavior are selected from a plurality of preset initial knowledge fragment variables and determined as reference knowledge fragment variables; the plurality of initial knowledge fragment variables are extracted from a plurality of preset initial will behavior description information. The reference knowledge fragment variables are subjected to tracing processing of will behavior description information to obtain the initial will behavior description information corresponding to the reference knowledge fragment variables; The initial will behavior description information corresponding to the reference knowledge fragment variable is determined as the reference will behavior description information.

6. The artificial intelligence-based intelligent assessment method for mental health as described in claim 5, characterized in that, The step of obtaining initial knowledge fragment variables that match the knowledge fragment describing the will behavior from a plurality of preset initial knowledge fragment variables and determining them as reference knowledge fragment variables includes: Obtain the candidate sharing coefficients between the knowledge fragments describing the will behavior and the multiple initial knowledge fragment variables; The multiple initial knowledge fragment variables are sorted according to the order set by the candidate sharing coefficients to obtain the sorted initial knowledge fragment variables; The first X initial knowledge fragment variables in the sorted initial knowledge fragment variables are determined as the initial knowledge fragment variables that match the information knowledge fragment describing the will behavior, where X is a positive integer; The initial knowledge fragment variable that matches the knowledge fragment describing the will behavior is determined as the reference knowledge fragment variable.

7. The artificial intelligence-based intelligent assessment method for mental health as described in claim 6, characterized in that, The process of obtaining candidate sharing coefficients between the knowledge fragments describing the will behavior and the plurality of initial knowledge fragment variables includes: For each initial knowledge fragment variable, the initial knowledge fragment variable is processed by the image of the control capability to obtain a partial three-dimensional coordinate system corresponding to the initial knowledge fragment variable. The partial three-dimensional coordinate system includes multiple partial variables with the same dimension, and the dimension of the partial variables is smaller than the dimension of the initial knowledge fragment variable. The partial three-dimensional coordinate system is classified by a preset classification network to obtain the classification criterion of the partial three-dimensional coordinate system; By compressing the partial three-dimensional coordinate system using the classification criterion, the category compression corresponding to the partial three-dimensional coordinate system is obtained; By compressing the aforementioned categories and the knowledge fragments describing willpower and behavior, candidate sharing coefficients are determined between the knowledge fragments describing willpower and behavior and the initial knowledge fragment variables; wherein, the candidate sharing coefficients are used to characterize the similarity of user psychological state information in the initial data.

8. The artificial intelligence-based intelligent assessment method for mental health as described in claim 1, characterized in that, The step of obtaining the first sharing coefficient between the partial will behavior description information and each reference will behavior description information includes: For the reference will behavior description information, identify the number of reference knowledge fragment variables corresponding to the reference will behavior description information; If the number of variables is one, then the reference knowledge fragment variable corresponding to the reference will behavior description information and the will behavior description information knowledge fragment corresponding to the partial will behavior description information are used to calculate the sharing coefficient to obtain the first sharing coefficient between the partial will behavior description information and the reference will behavior description information. If there are multiple variables, then the multiple reference knowledge fragment variables corresponding to the reference will behavior description information are respectively used to calculate the sharing coefficient with the will behavior description information knowledge fragment corresponding to the partial will behavior description information to obtain multiple initial sharing coefficients. Then, the multiple initial sharing coefficients are fused to obtain the first sharing coefficient between the partial will behavior description information and the reference will behavior description information.

9. The artificial intelligence-based intelligent assessment method for mental health as described in claim 3, characterized in that, The step of fusing the first shared coefficients of the candidate will behavior description information corresponding to each part of the will behavior description information to obtain the second shared coefficients of the candidate will behavior description information corresponding to the user psychological state data to be processed includes: Obtain the fusion weight of each part of the candidate will behavior description information and the number of multiple parts of the will behavior description information; The first shared coefficient of each part of the candidate will behavior description information is weighted by the fusion weight to obtain the fusion shared coefficient. The second sharing coefficient is obtained by comparing the fusion sharing coefficient with the number of partial will behavior description information.

10. An intelligent mental health assessment system based on artificial intelligence, characterized in that, The method includes a processor and a memory that communicate with each other, the processor being configured to read a computer program from the memory and execute it to implement the method of any one of claims 1-9.

Images