Psychological assessment system and method based on artificial intelligence

Abstract

The invention discloses a psychological assessment system and method based on artificial intelligence, relates to the technical field of psychological state assessment, and solves the technical problem of low accuracy of final psychological assessment caused by insufficient personalization of an existing psychological test assessment scheme. The data acquisition module is used for acquiring test video data, user physiological data and question and answer data in a test process; the evaluation selection module is used for acquiring psychological test questions and classifying the psychological test questions to generate a plurality of psychological test question groups; and acquiring personnel information of the tested personnel, inputting the personnel information into the preliminary analysis model to obtain a corresponding communication ability score and a plurality of psychological tendency scores, and generating a test scheme according to the communication ability score and the plurality of psychological tendency scores, the psychological assessment module is used for generating a test influence score according to the test video data and the user physiological data, and generating a corresponding psychological tendency assessment report according to the test influence score and the question and answer data; and the accuracy of psychological test evaluation results is improved.

Description

Technical Field

[0001] This application belongs to the technical field of psychological state assessment, involving psychological state assessment technology, and specifically is a psychological assessment system and method based on artificial intelligence. Background Art

[0002] Psychological assessment refers to the general term for comprehensively using methods of conversation, observation, and testing to conduct a comprehensive, systematic, and in-depth analysis of the psychological phenomena of individuals or groups under the joint guidance of the biological, psychological, social, and medical models. Its main role lies in helping counselors and clients discover changes during the counseling process, review the counseling effects, guide the direction and technology selection of psychological counseling, and has important guiding roles in the screening of psychiatric diseases and judging whether referral is needed, etc.

[0003] Existing psychological testing and assessment systems, although to a certain extent providing standardized measurement tools for the field of mental health, this system generally relies on psychiatrists to conduct psychological tests and assessments on different testees according to preset unified test questions and standardized question-and-answer methods; however, the personal experiences, cultural backgrounds, personality traits, and current psychological states of each testee vary greatly; there are significant differences in people's personality traits, emotional response patterns, coping mechanisms, and psychological resilience, etc. Therefore, when using unified psychological test questions and assessment methods, it is difficult to comprehensively and accurately capture the psychological characteristics and potential psychological problems of each testee; this limitation may lead to incomplete or deviated assessment results, thus affecting the formulation of subsequent psychological counseling or treatment plans; therefore, a psychological assessment system and method based on artificial intelligence are needed. Summary of the Invention

[0004] This application aims to solve at least one of the technical problems existing in the prior art; for this purpose, this application proposes a psychological assessment system and method based on artificial intelligence to solve the technical problem that the existing psychological test assessment scheme lacks personalization, resulting in relatively low accuracy of the final psychological assessment.

[0005] To achieve the above object, the first aspect of this application provides a psychological assessment system based on artificial intelligence, including: a data acquisition module, an interaction module, an assessment selection module, a psychological assessment module, a test database, and a database;

[0006] The data acquisition module: obtains test video data, user physiological data, and question-and-answer data during the test through data acquisition devices connected thereto;

[0007] The interaction module: is used to obtain the personal information of the testee and display the psychological test items for the test;

[0008] The evaluation and selection module: obtains psychological test questions through the database connected thereto, classifies the psychological test questions to generate several groups of psychological test questions, and stores them in the database; and, obtains the personal information of the tested person, inputs the personal information into the preliminary analysis model to obtain the corresponding communication ability score and several psychological tendency scores, and generates a test plan according to the communication ability score and several psychological tendency scores;

[0009] The psychological evaluation module: obtains the test video data, user physiological data, and question-and-answer data during the test; generates a test impact score according to the test video data and user physiological data, and generates a corresponding psychological tendency evaluation report according to the test impact score and the question-and-answer data.

[0010] This application classifies psychological test questions, obtains the personal information of the tested person; generates the communication ability score and psychological tendency scores of the tested person according to the personal information; generates a test plan according to the communication ability score and several psychological tendency scores; obtains the test video data, user physiological data, and question-and-answer data during the test; generates a test impact score according to the test video data and user physiological data; generates a corresponding psychological tendency evaluation report according to the test impact score and the question-and-answer data; by pre-analyzing the personal data of the tested person, selects the most suitable test method for the tested person, and generates targeted test questions, increases the personalization of psychological evaluation, and thus improves the accuracy of the psychological test evaluation results.

[0011] Preferably, classifying the psychological test questions to generate several groups of the psychological test questions includes:

[0012] Obtains one or more psychological test items corresponding to the psychological test questions, marks the type of the psychological test question that only contains one psychological test item as a single-choice test question, and marks the type of the psychological test question that contains multiple psychological test items as a multiple-choice test question;

[0013] Divides the single-choice test questions with the same psychological test items into a single-choice psychological test question group; and uses the psychological test item as the test item of the single-choice psychological test question group;

[0014] Divides the multiple-choice test questions with the same psychological test items into a multiple-choice psychological test question group; and uses the psychological test item as the test item of the multiple-choice psychological test question group;

[0015] The group of psychological test questions includes a single-choice psychological test question group and a multiple-choice psychological test question group.

[0016] Preferably, the preliminary analysis model is obtained through training of an artificial intelligence model, and includes:

[0017] Obtain a number of personnel information through a database, as well as the corresponding communication ability scores and a number of psychological tendency scores; the personnel information includes the dialogue text information converted from the dialogue audio provided by the tested personnel, the previous psychological test information of the tested personnel, etc.; the communication ability score is the score inferred by an expert based on the keywords in the personnel information to evaluate the communication ability of the tested personnel; integrate the personnel information, its corresponding communication ability score, and a number of psychological tendency scores into a number of training data and test data;

[0018] Use the training data to train the artificial intelligence model, and use the test data to test the trained artificial intelligence model; finally, obtain a preliminary analysis model with the personnel information as the input and the corresponding communication ability score and a number of psychological tendency scores as the output; among them, the artificial intelligence model includes a recurrent neural network model, etc.

[0019] Preferably, the generation of the test plan according to the communication ability score and a number of psychological tendency scores includes:

[0020] Judge whether the communication ability score is greater than the set communication ability threshold; if so, set the test mode to the dialogue mode, and the dialogue mode is for the psychologist to communicate face-to-face with the tested personnel to complete the psychological test; if not, set the test mode to the answering mode, and the answering mode is for the tested personnel to answer questions through a paper test or an electronic device to complete the psychological test;

[0021] Successively judge whether each psychological tendency score is greater than the set severe tendency score; if so, set the psychological test item corresponding to the psychological tendency score as the key test item; it can be understood that the psychological tendency score corresponds to the psychological test item one by one; if not, when the psychological tendency score is greater than the set medium tendency score, then set the psychological test item corresponding to the psychological tendency score as the auxiliary test item; otherwise, set the psychological test item corresponding to the psychological tendency score as the normal item;

[0022] Obtain the corresponding psychological test items for each key test item; select a set number of single test questions as test questions from the single psychological test question group with the same test item as the psychological test item;

[0023] Obtain the corresponding psychological test items for each auxiliary test item; select a set number of multiple test questions as test questions from the multiple psychological test question group with the same test item as the psychological test item;

[0024] Integrate the test mode and each test question into a test plan.

[0025] Preferably, the generation of the test impact score according to the test video data and the user's physiological data includes:

[0026] Extract the test video data for each recording period, decompose the video data frame by frame into a number of test images; input the test images into an image recognition model to obtain the corresponding image labels;

[0027] Extract a number of test values corresponding to a number of test items for each recording period in the user's physiological data, where the test items include one or more of heart rate, respiratory rate, skin conductance response, blood pressure, etc.;

[0028] Generate a test impact score based on a number of image labels and test values of a number of test items within each recording period.

[0029] This application generates a test impact score through the action performance and physiological responses of the person being evaluated during the evaluation process, and uses the test impact score to correct the score obtained from the Q&A to obtain the final score, further increasing the accuracy of psychological assessment.

[0030] Preferably, the image recognition model is trained through an artificial intelligence model, including:

[0031] Obtain a number of test images and their corresponding image labels through a database. The image labels are labels set for the images by experts according to the actions in the images, including normal and abnormal; integrate the number of test images and their corresponding image labels into a number of sets of training data and test data;

[0032] Use the training data to train the artificial intelligence model, use the test data to test the trained artificial intelligence model, and finally obtain an image recognition model with the input being the test image and the output being its corresponding image label. Among them, the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0033] Preferably, the generating a test impact score based on a number of image labels and test values of a number of test items within each recording period includes:

[0034] Obtain the number of monitored images with abnormal image labels and the number of monitored images with normal image labels within each recording period, and mark them as YS and ZS respectively;

[0035] Fit the test values of a number of test items into a project test curve CSi(t) according to their corresponding acquisition times; i is the number of the test item corresponding to the project test curve;

[0036] Through the formula Calculate the test impact score YP corresponding to the recording period; where ZSi is the normal value corresponding to the test item numbered i; α1 and α2 are proportionality coefficients used to adjust the proportional relationship between the two parameters, and the specific values are set according to expert experience; t ∈ [0, T], T is the length of the recording period; i = 1, 2, …, I; I is the total number of test items;

[0037] Obtain the test impact scores corresponding to each recording period in sequence.

[0038] Preferably, generating the corresponding psychological tendency evaluation report according to the test impact score and the Q&A data includes:

[0039] Obtain the test impact scores of each recording period; determine whether the test impact scores corresponding to each recording period are greater than the set impact threshold; if yes, mark the recording period as an abnormal period; if no, mark the recording period as a normal period;

[0040] Extract the scores of the test questions corresponding to each normal period in the Q&A data; generate the item scores of each psychological test item according to the scores corresponding to each test question and the test impact scores of each normal period.

[0041] Extract the test questions corresponding to each abnormal period in the Q&A data, and mark the test questions as key analysis questions.

[0042] Integrate the item scores of each psychological test item and several key analysis questions into a psychological tendency evaluation report.

[0043] Preferably, generating the item scores of each psychological test item according to the scores corresponding to each test question and the test impact scores of each normal period includes:

[0044] Obtain the test impact scores corresponding to each normal period, perform normalization processing on them and mark them as YPn, n is the recording period and also the number of the test question; since each recording period only contains one test question, so n is both the number of the recording period and the number of the test question;

[0045] Obtain the scores corresponding to the test questions corresponding to each normal period and mark them as CPmn; m is the number of the psychological evaluation item;

[0046] Through the formula Calculate the item score XPn corresponding to the psychological evaluation item numbered m; where SLmn is the set ratio of the psychological evaluation item numbered m in the test question numbered n; n = 1, 2, …, N; N is the total number of recording periods and also the total number of test questions; β is the adjustment coefficient.

[0047] Another aspect of the present application provides an artificial intelligence-based psychological assessment method, including the following steps:

[0048] Step 1: Obtain the personal information of the person to be tested;

[0049] Step 2: Input the personal information into the preliminary analysis model to obtain the corresponding communication ability score and psychological tendency score;

[0050] Step 3: Generate a test plan based on the communication ability score and several psychological tendency scores;

[0051] Step 4: Obtain the test video data, user physiological data, and question-and-answer data during the test;

[0052] Step 5: Generate a test impact score based on the test video data and user physiological data;

[0053] Step 6: Generate a corresponding psychological tendency assessment report based on the test impact score and question-and-answer data.

[0054] Compared with the prior art, the beneficial effects of the present application are as follows:

[0055] 1. In the present application, psychological test questions are classified to obtain the personal information of the person to be tested; a communication ability score and a psychological tendency score of the person to be tested are generated based on the personal information; a test plan is generated based on the communication ability score and several psychological tendency scores; the test video data, user physiological data, and question-and-answer data during the test are obtained; a test impact score is generated based on the test video data and user physiological data; a corresponding psychological tendency assessment report is generated based on the test impact score and question-and-answer data; by pre-analyzing the personal data of the person to be tested, the most suitable test method for the person to be tested is selected, and targeted test questions are generated, increasing the personalization of psychological assessment, and thus improving the accuracy of psychological test assessment results.

[0056] 2. In the present application, a test impact score is generated based on the action performance and physiological reactions of the person being tested during the assessment process, and the score obtained from the question-and-answer is corrected using the test impact score to obtain the final score, further increasing the accuracy of psychological assessment. Description of the Drawings

[0057] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0058] Figure 1Schematic diagram of module connection of the artificial intelligence-based psychological assessment system in this application;

[0059] Figure 2 Schematic diagram of steps of the artificial intelligence-based psychological assessment method in this application. Specific implementation manners

[0060] The technical solutions of this application will be described clearly and completely below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in this application without creative efforts shall fall within the scope of protection of this application.

[0061] Please refer to Figure 1 , a first aspect of this application provides an artificial intelligence-based psychological assessment system, including: a data acquisition module, an interaction module, an assessment selection module, a psychological assessment module, a test database, and a database;

[0062] Data acquisition module: Obtain test video data, user physiological data, and question-and-answer data during the test through data acquisition devices connected thereto; the data acquisition devices include photographic devices and some acquisition devices for testing physiological data; the test video data is the action video of the tested person collected during the psychological test; the user physiological data is one or more of the heartbeat, breathing rate, skin conductance response, and blood pressure of the tested person collected during the psychological test; the question-and-answer data is the test questions, answer content, and corresponding scores collected during the question-and-answer test or paper test of the tested person;

[0063] Interaction module: Used to obtain the personal information of the tested person and display the psychological test items for the test; the personal information includes the dialogue text information converted from the dialogue audio provided by the tested person, the previous psychological test information of the tested person, and other data related to the tested person, which is used to analyze the psychological problem tendency of the tested person; the psychological test items include psychological test questions, etc.;

[0064] Evaluation and Selection Module: Obtain psychological test questions through the database connected to it. The psychological test questions are relevant questions stored in the database for psychological problem testing. It can be understood that the psychological test questions in the database are updated in real time to ensure the accuracy and forward-looking of the psychological test questions. Classify the psychological test questions to generate several groups of psychological test questions and store them in the test database by category. It can be understood that different psychological problems generally require different psychological test questions. Therefore, it is necessary to classify and store the psychological test questions according to different psychological problems for subsequent generation of test plans. Obtain the personal information of the tested person, and input the personal information into the preliminary analysis model to obtain the corresponding communication ability score and several psychological tendency scores. The communication ability score is the score evaluated by experts based on the keywords in the personal information to evaluate the communication ability of the tested person. The psychological tendency score is the possibility that the tested person has the corresponding psychological test item in each psychological test item inferred by experts according to the personal information. Generate a test plan based on the communication ability score and several psychological tendency scores. The test plan includes a test mode and test questions. The test mode includes a dialogue mode and a question-and-answer mode. The test questions are several questions generated specifically according to the personal information of the tested person.

[0065] Psychological Assessment Module: Obtain the test video data, user physiological data, and question-and-answer data during the test. Generate a test impact score based on the test video data and user physiological data. The test impact score is the score for evaluating the answering effect of the corresponding question based on the physiological reaction of the tested person when answering the corresponding question. For example, when the tested person is answering a certain question, if the emotion is too anxious or tense, it will cause the answer to the corresponding question to not accurately express the situation of the tested person, resulting in a poor answering effect, and its corresponding test impact score will be larger. Generate a corresponding psychological tendency assessment report based on the test impact score and the question-and-answer data. The psychological tendency assessment report is the assessment result of the psychological test of the tested person this time, including the item scores of each psychological test item and some key test questions.

[0066] In this embodiment, by classifying the psychological test questions, obtaining the personal information of the tested person, generating the communication ability score and psychological tendency score of the tested person according to the personal information, generating a test plan according to the communication ability score and several psychological tendency scores, obtaining the test video data, user physiological data, and question-and-answer data during the test, generating a test impact score based on the test video data and user physiological data, and generating a corresponding psychological tendency assessment report based on the test impact score and the question-and-answer data, the most suitable test method for the tested person is selected by pre-analyzing the personal data of the tested person, and targeted test questions are generated, increasing the personalization of psychological assessment, and thus improving the accuracy of the psychological test assessment results.

[0067] Classify psychological test questions to generate several groups of psychological test questions, including: obtaining one or more psychological test items corresponding to the psychological test questions, where the psychological test items correspond one-to-one with psychological problems. For example, if the psychological problem is depression, the corresponding psychological test item is depression; mark the type of psychological test questions that contain only one psychological test item as single-item test questions, and single-item test questions are questions used only to test one psychological problem; mark the type of psychological test questions that contain multiple psychological test items as multiple-item test questions; multiple-item test questions are questions from which multiple psychological problem tendencies can be analyzed based on the answers given by the tested person; divide single-item test questions with the same psychological test item into a single-item psychological test question group; and use the psychological test item as the test item for the single-item psychological test question group; divide multiple-item test questions with the same psychological test item into a multiple-item psychological test question group; and use the psychological test item as the test item for the multiple-item psychological test question group; the psychological test question groups include single-item psychological test question groups and multiple-item psychological test question groups; it can be understood that the above functions only run when updating psychological test questions.

[0068] In this embodiment, the psychological test questions are classified according to the psychological test items corresponding to the psychological test questions, which is convenient for quickly and accurately generating test questions subsequently; it is convenient to improve the data processing efficiency.

[0069] The preliminary analysis model is obtained through training with an artificial intelligence model and includes:

[0070] Obtain several pieces of personnel information through the database, as well as the corresponding communication ability scores and several psychological tendency scores; the personnel information includes the dialogue text information converted from the dialogue audio provided by the tested person, the previous psychological test information of the tested person, etc.; the communication ability score is the score inferred by an expert based on the keywords in the personnel information to evaluate the communication ability of the tested person; the higher the communication ability score, the more active the tested person is in language communication, and the better the test effect of the psychological test when communicating face-to-face with the tested person; the lower the communication ability score, the more passive the tested person is in language communication, and the worse the test effect of the psychological test when communicating face-to-face with the tested person; the psychological tendency score is the possibility that the tested person has the corresponding psychological test item in each psychological test item inferred by an expert based on the personnel information. The higher the psychological tendency score of a certain psychological test item, the greater the possibility that the tested person has the psychological test item; integrate the personnel information, as well as the corresponding communication ability scores and several psychological tendency scores into several training data and test data;

[0071] Train an artificial intelligence model using training data and test the trained artificial intelligence model using test data. Specifically, input the personal information in the test data into the trained artificial intelligence model to obtain the corresponding communication ability score and several psychological tendency scores. Determine whether the differences between the output communication ability score and several psychological tendency scores and the corresponding communication ability score and several psychological tendency scores in the test data are all within an acceptable range. If so, it means that the test data passes the test. If not, relevant parameters of the artificial intelligence model need to be adjusted, and this group of test data is used for testing again until a set proportion of the test data passes the test. Finally, obtain a preliminary analysis model with personal information as the input and the corresponding communication ability score and several psychological tendency scores as the output. When a certain psychological tendency score is 0, it means that the tested person has no psychological problem tendency corresponding to the psychological test item corresponding to the psychological tendency score. Among them, the artificial intelligence model includes a recurrent neural network model, etc.

[0072] Generate a test plan based on the communication ability score and several psychological tendency scores, including: determining whether the communication ability score is greater than a set communication ability threshold; if so, set the test mode to the dialogue mode, where the dialogue mode is for a psychologist to communicate face-to-face with the tested person to complete the psychological test; if not, set the test mode to the answering mode, where the answering mode is for the tested person to answer questions through a paper test or an electronic device to complete the psychological test.

[0073] Successively determine whether each psychological tendency score is greater than a set severe tendency score; if so, set the psychological test item corresponding to the psychological tendency score as a key test item; it can be understood that the psychological tendency score corresponds to the psychological test item one by one; if not, when the psychological tendency score is greater than a set medium tendency score, set the psychological test item corresponding to the psychological tendency score as an auxiliary test item; otherwise, set the psychological test item corresponding to the psychological tendency score as a normal item; the key test item is the psychological test item that needs to be key tested, the auxiliary test item is the psychological test item that needs to be examined, and the normal item is the psychological test item that does not need to be examined.

[0074] Obtain the corresponding psychological test items for each key test item; select a set number of single test questions as test questions from the single psychological test question group where the test item is the same as the psychological test item.

[0075] Obtain the psychological test items corresponding to each auxiliary test item; select a set number of test questions from multiple groups of psychological test questions with the same test items as the psychological test items as the test questions; it can be understood that the set number of questions for key test items is greater than the set number of questions for auxiliary test items, and the set number of questions is set by experts; specific questions can also be selected by experts or randomly selected from the corresponding question groups; integrate the test mode and each test question into a test plan.

[0076] In this embodiment, by the characteristics of the tested person, a suitable test mode and suitable test questions are selected, so that the psychological test has strong pertinence and ensures the accuracy of the test results.

[0077] Generate a test impact score based on the test video data and the user's physiological data, including: extract the test video data for each recording period, decompose the video data frame by frame into a number of test images; input the test images into an image recognition model to obtain the corresponding image labels; the recording period is divided according to the questions, and each test question corresponds to a recording period; the labels set for the images by experts according to the actions in the images, including normal and abnormal.

[0078] Extract a number of test values corresponding to a number of test items in the user's physiological data for each recording period, and the test items include one or more of heart rate, respiratory rate, skin conductance response, blood pressure, etc.

[0079] Generate a test impact score based on the number of image labels and the test values of a number of test items within each recording period.

[0080] Since during the test, the tested person may be affected by the questions or the test environment, resulting in the tested person being psychologically tense, anxious, etc., and the results of the corresponding answer questions are not accurate enough. Therefore, in this embodiment, a test impact score is set to adjust the weight of the scores of different test questions to obtain the final item score, making each item score more accurate; at the same time, questions that affect the emotions of the tested person can also be screened out according to the test impact score, facilitating relevant personnel to accurately analyze the psychological problems of the tested person.

[0081] In this embodiment, a test impact score is generated based on the action performance and physiological reactions of the tested person during the evaluation process, and the test impact score is used to correct the score obtained from the question and answer to obtain the final score, further increasing the accuracy of the psychological assessment.

[0082] The image recognition model is obtained by training an artificial intelligence model, including: obtaining a number of test images and their corresponding image labels through a database. The image labels are labels set for the images by experts according to the actions in the images, including normal and abnormal. An abnormal image label indicates that the tested person has made some habitual actions that are prone to occur when nervous or anxious, such as touching the hair, clasping the fingers, and biting the nails during the Q&A process. When the tested person makes these actions, the answers to the questions are difficult to accurately represent the true answers of the tested person due to the influence of the tested person's emotions. Integrating a number of test images and their corresponding image labels into a number of sets of training data and test data;

[0083] Training the artificial intelligence model with the training data and testing the trained artificial intelligence model with the test data. The specific steps of training the model are similar to those of a normal model and will not be described in detail here. Finally, an image recognition model with the input being a test image and the output being its corresponding image label is obtained. Among them, the artificial intelligence model includes a BP neural network model and an RBF neural network model.

[0084] Generating a test impact score according to a number of image labels and test values of a number of test items in each recording period, including: obtaining the number of monitored images with abnormal image labels and the number of monitored images with normal image labels in each recording period, and marking them as YS and ZS respectively;

[0085] Fitting the test values of a number of test items into a project test curve CSi(t) according to their corresponding acquisition times; i is the number of the test item corresponding to the project test curve;

[0086] Through the formula Calculating to obtain the test impact score YP for the corresponding recording period; where ZSi is the normal value corresponding to the test item numbered i; α1 and α2 are proportionality coefficients used to adjust the proportional relationship between the two parameters, and the specific values are set according to expert experience; t ∈ [0, T], T is the length of the recording period; i = 1, 2,..., I; I is the total number of test items. Sequentially obtain the test impact scores corresponding to each recording period.

[0087] In this embodiment, the test impact scores of each test question are calculated through the above formula. When the test data corresponding to each physiological test item deviates more from the normal value in a certain recording period, it indicates that the emotions of the tested person are abnormal at this time, and the greater the impact on the answers of the tested person, the greater the corresponding test impact score is set. When there are more abnormal actions recorded by the tested person in a certain recording period, it indicates that the emotions of the tested person are abnormal at this time, and the greater the impact on the answers of the tested person, the greater the corresponding test impact score is also set.

[0088] Generate a corresponding psychological tendency assessment report based on the test impact score and Q&A data, including: obtaining the test impact scores for each recording period; determining whether the test impact score corresponding to each recording period is greater than a set impact threshold, and the specific value of the impact threshold is set by experts based on experience; if so, mark the recording period as an abnormal period; if not, mark the recording period as a normal period;

[0089] When the test impact score is greater than the set impact threshold, it indicates that the emotional state of the person being tested is too intense at this time, and the corresponding test question answers are too inaccurate and will not be referred to in the final scoring;

[0090] Extract the scores of the test questions corresponding to each normal period in the Q&A data; generate the item scores of each psychological test item based on the scores corresponding to each test question and the test impact scores of each normal period; extract the test questions corresponding to each abnormal period in the Q&A data and mark the test questions as key analysis questions; integrate the item scores of each psychological test item and several key analysis questions into a psychological tendency assessment report.

[0091] Generate the item scores of each psychological test item based on the scores corresponding to each test question and the test impact scores of each normal period, including: obtaining the test impact scores corresponding to each normal period, normalizing them and marking them as YPn, where n is the recording period and also the test question number; since each recording period only contains one test question, n is both the recording period number and the test question number;

[0092] Obtain the scores corresponding to the test questions corresponding to each normal period and mark them as CPmn; m is the number of the psychological assessment item; it can be understood that the score corresponding to the test question is the comprehensive score corresponding to the test question. Through the set proportionality coefficients of each psychological assessment item corresponding to the test question, the scores of the test question corresponding to each psychological assessment item can be calculated, that is, multiplying the set proportionality coefficient by the corresponding comprehensive score can obtain the score of the psychological assessment item corresponding to the set proportionality coefficient; if the set proportionality coefficient of the psychological assessment item of the test question is 0, it means that the test question is not used to test this psychological assessment item at this time; if the set proportionality coefficient of the psychological assessment item of the test question is 1, it means that the test question is only used to test this psychological assessment item, and the value range of the set proportionality coefficient is from 0 to 1, including 0 and 1;

[0093] Through the formula Calculate the item score XPn corresponding to the psychological assessment item numbered m; where SLmn is the set ratio of the psychological assessment item numbered m in the test question numbered n; n = 1, 2, …, N; N is the total number of recording periods and also the total number of test questions; β is an adjustment coefficient; used to adjust the reference influence of the test impact score on the test questions, and the specific value is set according to expert experience.

[0094] In this embodiment, by introducing the test impact scores of each test question, the accuracy of the final score is increased; when the test impact score is larger, the reference of the corresponding test question is smaller; when the number of test questions corresponding to a certain psychological test item is larger, the scores of each test question are larger, and the set ratio coefficient of the psychological test item is larger, the item score of this psychological test item is also larger.

[0095] In this embodiment, the test questions with greater influence on the emotions of the tested person and the item scores corresponding to each psychological test item are integrated into a psychological tendency assessment report; it is convenient for relevant personnel to intuitively analyze the psychological problems of the tested person; and it is possible to conduct targeted analysis of the tested person based on the key analysis questions with greater influence on the emotions of the tested person.

[0096] Please refer to Figure 2 , another aspect of this application provides a psychological assessment method based on artificial intelligence, including the following steps:

[0097] Step 1: Obtain the personal information of the tested person;

[0098] Step 2: Input the personal information into the preliminary analysis model to obtain the corresponding communication ability score and psychological tendency score;

[0099] Step 3: Generate a test plan based on the communication ability score and several psychological tendency scores;

[0100] Step 4: Obtain the test video data, user physiological data, and question-and-answer data during the test;

[0101] Step 5: Generate a test impact score based on the test video data and user physiological data;

[0102] Step 6: Generate a corresponding psychological tendency assessment report based on the test impact score and the question-and-answer data; the specific steps of Steps 4 to 6 are similar to the working principles executed by the corresponding modules.

[0103] Some of the data in the above formula are calculated by removing the dimension and taking their numerical values. The formula is obtained by software simulation of a large amount of collected data to get a formula closest to the real situation; the preset parameters and preset thresholds in the formula are set by those skilled in the art according to the actual situation or obtained through simulation of a large amount of data.

[0104] Working principle of this application:

[0105] In this application, psychological test questions are classified to obtain the personal information of the tested person; the communication ability score and psychological tendency score of the tested person are generated according to the personal information; a test plan is generated according to the communication ability score and several psychological tendency scores; test video data, user physiological data and question-and-answer data during the test are obtained; a test impact score is generated according to the test video data and user physiological data; a corresponding psychological tendency evaluation report is generated according to the test impact score and question-and-answer data; by analyzing the personal data of the tested person in advance, the most suitable test method for the tested person is selected, and targeted test questions are generated to increase the personalization of psychological evaluation, thereby improving the accuracy of the psychological test evaluation results.

[0106] The above embodiments are only used to illustrate the technical method of this application and not to limit it. Although this application has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of this application can be modified or equivalently replaced without departing from the spirit and scope of the technical method of this application.

Claims

1. A psychological assessment system based on artificial intelligence, comprising: Data collection module, assessment and selection module, psychological assessment module and database; characterized in that, The data acquisition module is used to acquire test video data, user physiological data and question-and-answer data during the test process; The evaluation and selection module: obtains psychological test questions, classifies the psychological test questions to generate a plurality of psychological test question groups; and obtains the personnel information of the tested person, inputs the preliminary analysis model according to the personnel information to obtain the corresponding communication ability score and a plurality of psychological tendency scores, and generates a test plan according to the communication ability score and the plurality of psychological tendency scores; The psychological assessment module: obtains test video data, user physiological data and question-and-answer data during the test; generates a test impact score based on the test video data and user physiological data, and generates a corresponding psychological tendency assessment report based on the test impact score and question-and-answer data.

2. The artificial intelligence-based psychological assessment system according to claim 1, characterized in that: Classifying the psychological test questions to generate a plurality of psychological test question groups, including: Obtain one or more psychological test items corresponding to the psychological test question, mark the type of the psychological test question containing only one psychological test item as a single test question, and mark the type of the psychological test question containing multiple psychological test items as a multiple test question; Grouping single test questions with the same psychological test items into a single psychological test question group; and using the psychological test items as test items of the single psychological test question group; Grouping multiple test questions with the same psychological test items into a group of multiple psychological test questions; and using the psychological test items as test items of the group of multiple psychological test questions; The psychological test question group includes a single psychological test question group and a multiple psychological test question group.

3. The artificial intelligence-based psychological assessment system according to claim 1, characterized in that: The preliminary analysis model is obtained through artificial intelligence model training, including: Acquire a number of personnel information, and corresponding communication ability scores and a number of psychological tendency scores through a database; integrate the personnel information, and the corresponding communication ability scores and a number of psychological tendency scores into a number of training data and test data; Use training data to train the artificial intelligence model, and use test data to test the trained artificial intelligence model; ultimately, a preliminary analysis model is obtained with personnel information as input and corresponding communication ability scores and several psychological tendency scores as output.

4. The artificial intelligence-based psychological assessment system according to claim 1, characterized in that: The test scheme is generated based on the communication ability score and several psychological tendency scores, including: Determine whether the communication ability score is greater than the set communication ability threshold; if yes, set the test mode to dialogue mode; if no, set the test mode to answer question mode; It is judged in turn whether each psychological tendency score is greater than the set severe tendency score; if yes, the psychological test item corresponding to the psychological tendency score is set as the key test item; if no, when the psychological tendency score is greater than the set medium tendency score, the psychological test item corresponding to the psychological tendency score is set as the auxiliary test item; otherwise, the psychological test item corresponding to the psychological tendency score is set as the normal item; Obtaining the psychological test items corresponding to each key test item; selecting a set number of single test questions as test questions from a group of single psychological test questions whose test items are the same as the psychological test items; Obtaining psychological test items corresponding to each auxiliary test item; selecting a set number of multiple test questions from a group of multiple psychological test questions whose test items are the same as the psychological test items as test questions; Integrate the test mode and each test question into a test plan.

5. The artificial intelligence-based psychological assessment system according to claim 1, characterized in that: Generating a test impact score according to the test video data and the user's physiological data includes: Extract the test video data of each recording period, decompose the video data frame by frame into several test images; input the test images into the image recognition model to obtain the corresponding image labels; Extracting a number of test values ​​corresponding to a number of test items in each recording period in the user's physiological data; A test impact score is generated according to a number of image labels and test values ​​of a number of test items in each recording period.

6. The artificial intelligence-based psychological assessment system according to claim 5, characterized in that: The image recognition model is obtained by training an artificial intelligence model, including: Acquire a number of test images and their corresponding image labels through a database; integrate the number of test images and their corresponding image labels into a number of groups of training data and test data; The artificial intelligence model is trained using training data, and the trained artificial intelligence model is tested using test data, ultimately obtaining an image recognition model whose input is a test image and whose output is its corresponding image label.

7. The artificial intelligence-based psychological assessment system according to claim 5, characterized in that: The generating of the test impact score according to the image labels and the test values ​​of the test items in each recording period includes: Obtain the number of monitoring images with abnormal image labels and the number of monitoring images with normal image labels in each recording period, marked as YS and ZS respectively; Fit the test values ​​of each test item into a test curve CSi(t) according to the corresponding collection time; i is the number of the test item corresponding to the test curve; By formula The test impact score YP of the corresponding recording period is calculated; wherein ZSi is the normal value corresponding to the test item numbered i; α1 and α2 are proportional coefficients; t∈[0, T], T is the length of the recording period; i=1, 2, …, I; I is the total number of test items; Get the test impact score corresponding to each recording period in turn.

8. The artificial intelligence-based psychological assessment system according to claim 1, characterized in that: The psychological tendency assessment report generated according to the test impact score and question and answer data includes: Obtain the test impact score of each recording period; determine whether the test impact score corresponding to each recording period is greater than a set impact threshold; if yes, mark the recording period as an abnormal period; if no, mark the recording period as a normal period; Extract the scores of the test questions corresponding to each normal period in the question and answer data; generate the project scores of each psychological test item according to the scores corresponding to each test question and the test impact scores of each normal period; Extract the test questions corresponding to each abnormal period in the question-and-answer data, and mark the test questions as key analysis questions; Integrate the item scores of each psychological test item and several key analysis questions into a psychological tendency assessment report.

9. The artificial intelligence-based psychological assessment system according to claim 8, characterized in that: The generating of the item scores of each psychological test item according to the scores corresponding to each test question and the test impact scores of each normal time period includes: Get the test impact score corresponding to each normal time period, normalize it and mark it as YPn, where n is the recording time period and also the number of the test question; Get the scores corresponding to the test questions in each normal period and mark them as CPmn; m is the number of the psychological assessment item; By formula Calculate and obtain the project score XPn corresponding to the psychological assessment project numbered m; wherein SLmn is the set proportion of the psychological assessment project numbered m in the test item numbered n; n = 1, 2, ..., N; N is the total number of recording periods, which is also the total number of test items; β is the adjustment coefficient.

10. A psychological assessment method based on artificial intelligence, applied to a psychological assessment system based on artificial intelligence as claimed in any one of claims 1 to 9; characterized in that: The following steps are involved: Step 1: Obtain the personnel information of the person being tested; Step 2: Input personnel information into the preliminary analysis model to obtain corresponding communication ability scores and psychological tendency scores; Step 3: Generate a test plan based on the communication ability score and several psychological tendency scores; Step 4: Obtain test video data, user physiological data, and question-and-answer data during the test; Step 5: Generate a test impact score based on the test video data and user physiological data; Step 6: Generate a corresponding psychological tendency assessment report based on the test impact score and question and answer data.

Images