Psychological assessment tutoring method and system based on artificial intelligence and large model

Abstract

The invention discloses a psychological assessment tutoring method and system based on artificial intelligence and a large model, and belongs to the technical field of psychological assessment tutoring, and the method comprises the following steps: S1, obtaining and processing real-time data of a user; s2, feature selection and extraction; s3, constructing a psychological assessment large model, and predicting the psychological state and the development trend of the user; and S4, providing a psychological tutoring scheme for the user, intervening the psychology of the user in time, and optimizing and adjusting the psychological tutoring scheme according to the feedback condition of the user. According to the method and the device, the problems that the psychological state of the user cannot be accurately evaluated and a more effective psychological tutoring scheme cannot be formulated in the prior art, so that the user cannot obtain timely psychological tutoring intervention and the psychological tutoring effect of the user is poor are solved. The psychological state of the user can be accurately evaluated, so that a more effective psychological tutoring scheme can be formulated, the user can obtain timely psychological tutoring intervention, and the psychological tutoring effect of the user can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of psychological assessment and counseling, and specifically to a psychological assessment and counseling method and system based on artificial intelligence and large models. Background Technique

[0002] A Chinese patent with the publication number CN106126916A discloses a psychological test system, including: a tested terminal and a control host. An encryption KEY is connected to the USB interface of the tested terminal, a fingerprint scanner is connected to the USB interface of the tested terminal, the output end of the evaluation module is connected to the input end of the evaluation result module, the output end of the evaluation result module is connected to the input end of the data processing module, the output end of the social software module is connected to the input end of the conversion module through a docking program module, the output end of the conversion module is connected to the input end of the data processing module, and the data processing module is provided with an archive generation module, a crisis warning module, and a solution output module. The output end of the solution output module is connected to the storage module, which has the advantages of high automation, high test efficiency, and convenient use; however, this patent has the following defects: The existing technology cannot accurately evaluate the psychological state of users, and cannot formulate more effective psychological counseling plans, resulting in users not being able to receive timely psychological counseling interventions, leading to poor psychological counseling effects for users. Summary of the Invention

[0003] The purpose of the present invention is to provide a psychological assessment and counseling method and system based on artificial intelligence and large models, which can accurately evaluate the psychological state of users, thereby being able to formulate more effective psychological counseling plans, enabling users to receive timely psychological counseling interventions, and improving the psychological counseling effect of users, solving the problems raised in the above background technique.

[0004] To achieve the above purpose, the present invention provides the following technical solutions: A psychological assessment and counseling method based on artificial intelligence and large models includes the following steps: S1. Acquisition and processing of user real-time data: Collect personal information and behavior habits of users through various channels, determine user real-time data based on artificial intelligence, and clean and transform the user real-time data based on artificial intelligence to determine standardized user real-time data based on artificial intelligence; S2. Feature selection and extraction: Use machine learning technology to denoise and reduce the dimension of the standardized user real-time data based on artificial intelligence, select the feature data that contributes the most to the model performance, and extract features from the feature data to reduce the dimension of the feature data and create new user feature data; S3. Construct a large psychological assessment model: Use deep learning algorithms to train a large psychological assessment model, test and optimize the large psychological assessment model, determine the optimal large psychological assessment model, and analyze the user feature data based on the optimal large psychological assessment model to predict the user's psychological state and its development trend; S4. Provide a psychological counseling plan: According to the user's psychological assessment results, provide a psychological counseling plan for the user, intervene in the user's psychology in a timely manner, and track the user's psychological counseling intervention situation in real time. Optimize and adjust the psychological counseling plan according to the user's feedback to meet the user's changing psychological counseling needs.

[0005] Preferably, in S1, collect the user's personal information and behavior habits through multiple channels, including: Real-time collect the user's name, age, gender, occupation, marital status, health status, contact information, educational background and family background through multiple channels based on social media, online questionnaires and applications to obtain the user's personal information; Real-time collect the user's browsing behavior, purchase behavior, social behavior, mobile behavior, entertainment behavior and reading behavior through multiple channels based on social media, online questionnaires and applications to obtain the user's behavior habits; Among them, based on the real-time collected user personal information and user behavior habits, determine the user real-time data based on artificial intelligence, and keep the user real-time data based on artificial intelligence secure and confidential in accordance with privacy protection regulations.

[0006] Preferably, in S1, process the user real-time data based on artificial intelligence, including: Import the user real-time data based on artificial intelligence into a data cleaning tool; Clean the user real-time data based on artificial intelligence based on the data cleaning tool; Among them, view the type and distribution of the user real-time data based on artificial intelligence, check the user real-time data based on artificial intelligence, judge whether there are duplicate values, missing values and outliers in the user real-time data based on artificial intelligence, and process the duplicate values, missing values and outliers in the user real-time data based on artificial intelligence; For duplicate values, delete the duplicate values in the user real-time data based on artificial intelligence to maintain the uniqueness of the user real-time data based on artificial intelligence; For missing values, delete the missing values in the user real-time data based on artificial intelligence, or use filling or interpolation methods to process the missing values in the user real-time data based on artificial intelligence; For outliers, delete the outliers in the user real-time data based on artificial intelligence, or use replacement or smoothing methods to process the outliers in the user real-time data based on artificial intelligence; Perform type and format conversion on the cleaned real-time user data based on artificial intelligence, reduce the dimensional difference between the real-time user data based on artificial intelligence, and determine the standardized real-time user data based on artificial intelligence.

[0007] Preferably, in S2, use machine learning technology to denoise and reduce the dimension of the real-time user data based on artificial intelligence, and extract user feature data, including: Based on the feature selection method of the model, use the feature importance scoring mechanism of the decision tree to select feature data from the standardized real-time user data based on artificial intelligence. By calculating the contribution of each feature data to the model performance, select the feature data with the greatest contribution to the model performance, and at the same time remove redundant and irrelevant noise feature data; Based on the machine learning algorithm, perform feature extraction on the selected feature data. Through linear transformation, project the feature data into a lower-dimensional space, while retaining most of the variance of the feature data, reduce the dimension of the feature data, and create new user feature data, including user mood changes, behavior patterns, and social interactions.

[0008] Preferably, in S3, use the deep learning algorithm to train the psychological evaluation large model, including: Collect user historical data, process the user historical data, and divide the processed user historical data. Determine the training set and the test set according to the ratio of 7:3; Use the deep learning algorithm to train the deep learning model with the training set, so that the deep learning model can learn complex psychological patterns from the user historical data, and enable it to accurately predict the user's psychological state and its development trend, and determine the psychological evaluation large model; Based on the test set, perform performance testing on the psychological evaluation large model, and use the evaluation indicators of accuracy and recall to evaluate the psychological evaluation large model, judge whether the psychological evaluation large model can achieve the expected effect, and use cross-validation and adjust hyperparameters to optimize the performance of the psychological evaluation large model, and determine the optimal psychological evaluation large model.

[0009] Preferably, in S3, predict the user's psychological state and its development trend, including: Obtain the optimal psychological evaluation large model, and deploy the optimal psychological evaluation large model in the actual psychological evaluation and counseling environment; Input the user feature data into the optimal psychological evaluation large model, analyze the user feature data based on the optimal psychological evaluation large model, predict the user's psychological state and its development trend, including the user's emotional state, stress level, and anxiety level, and determine the user psychological evaluation result.

[0010] Preferably, in S4, according to the user psychological assessment result, a psychological counseling plan is provided for the user, including: According to the user psychological assessment result, a psychological counseling plan is provided for the user, and the user's psychology is intervened in a timely manner, including online counseling, face-to-face counseling, and psychological training course counseling; For online counseling, according to the user psychological assessment result, a psychological counselor is recommended for the user. The user makes an appointment with the psychological counselor through an online platform. After the psychological counselor establishes an online connection with the user, the psychological counselor assesses the user's psychological state and jointly formulates a psychological counseling plan with the user, including the goals, content, methods, and time arrangements of psychological counseling. According to the psychological counseling plan, the psychological counselor conducts regular psychological counseling for the user through the online platform. During the psychological counseling process, the psychological counselor provides psychological support, education, and intervention for the user; For face-to-face counseling, according to the user psychological assessment result, a psychological counselor is recommended for the user. The user makes an appointment with the psychological counselor through a reservation platform. After the psychological counselor meets with the user, the psychological counselor has a face-to-face conversation with the user. During the conversation, the psychological counselor directly observes the user's non-verbal behaviors, including facial expressions and body languages, so as to understand the user's psychological condition more comprehensively, and then conducts psychological counseling for the user; For psychological training course counseling, according to the user psychological assessment result, psychological training courses are recommended for the user, including theoretical explanations, practical exercises, and group discussion courses. According to the course content and user characteristics, teaching methods and tools are selected to enable the user to participate in case analysis, role-playing, and psychological tests, focusing on the user's interactive behaviors, and encouraging the user to ask questions and share experiences; Among them, the psychological counseling intervention situation of the user is tracked in real time, the user's psychological state and psychological counseling effect are continuously monitored, and the psychological counseling plan is optimized and adjusted according to the user feedback situation to meet the user's changing psychological counseling needs.

[0011] According to another aspect of the present invention, a psychological assessment and counseling system based on artificial intelligence and large models is provided, which is used to implement the psychological assessment and counseling method based on artificial intelligence and large models as described above, including: A data acquisition and processing module, which is used to acquire real-time user data based on artificial intelligence and process the real-time user data based on artificial intelligence; A feature selection and extraction module, which is used to perform feature selection and extraction on the real-time user data based on artificial intelligence to determine user feature data; A large model construction and prediction module, which is used to construct a psychological assessment large model and predict the user's psychological state and its development trend based on the psychological assessment large model to determine the user psychological assessment result; A psychological counseling control module, which is used to provide a psychological counseling plan for a user based on the user's psychological assessment results, and timely intervene and control the user's psychology.

[0012] Preferably, the psychological counseling control module includes: A video data acquisition unit, which is used to acquire video data of a video acquisition device configured at a preset position, and screen the video data to obtain relevant video data of the user to be controlled; An audio data acquisition unit, which is used to acquire audio data collected by an audio acquisition device of a control terminal worn by the user to be controlled, and screen the audio data to obtain data to be analyzed; A scenario analysis unit, which is used to determine the current scenario based on the relevant video data and the data to be analyzed; A vital sign data acquisition unit, which is used to acquire vital sign data monitored by the control terminal; An intervention unit, which is used to comprehensively analyze the current scenario and the vital sign data, determine an intervention strategy and execute it.

[0013] Preferably, the scenario analysis unit determines the current scenario based on the relevant video data and the data to be analyzed, and performs the following operations: Extract features from the relevant video data to obtain multiple first feature parameters; Extract features from the data to be analyzed to obtain multiple second feature parameters; Arrange the multiple first feature parameters and second feature parameters in the order specified by a preset scenario analysis template to form an analysis data set; Match the analysis data set with the scenario sets corresponding to each scenario in a pre-configured scenario analysis library; Extract the scenario corresponding to the scenario set that matches the analysis data set as the current scenario; Among them, extracting features from the relevant video data to obtain multiple first feature parameters includes: Identify objects and people in the relevant video data to obtain first feature parameters representing each object and person; Analyze the positional relationships between each object and each person in the relevant video data to obtain first feature parameters representing the relative positions of each person and object; Determine the user to be controlled in the relevant video data and calculate the probability of interaction between the user to be controlled and other people according to the following formula. The calculation formula is as follows: ; Among them, is the probability of interaction between the user to be controlled and another person within a certain period of time, is the time value, is the current position vector of the user to be controlled and another person, is a preset adjustment factor used to adjust the error in speed recognition, is the current speed vector of the user to be controlled and another person, which is obtained by differentiating the position vectors between frames, is a probability function, , the variable in the probability function is non - negative, and the value of the probability function decreases monotonically as the variable increases; Generate a first characteristic parameter representing the probability of interaction between the user to be controlled and other people.

[0014] Preferably, the psychological counseling and control module further includes: An online real - time pairing and intervention unit for pairing online intervention personnel with the users to be controlled one by one, and conducting one - to - one docking interventions on the users to be controlled based on the paired online intervention personnel; Among them, the online real - time pairing and intervention unit pairs online intervention personnel with the users to be controlled one by one and performs the following operations: Analyze the intervention records of online intervention personnel to construct multiple first portrait sets; the first portrait sets are constructed based on the situations of the users to be controlled corresponding to the intervention records; Analyze the intervention records of the users to be controlled to construct multiple second portrait sets; the second portrait sets are constructed based on the situations of each online intervention personnel in the intervention records; Analyze the intervention records of online intervention personnel and the feedback of the corresponding intervention records to determine the ability value; Calculate the matching degree between the online intervention personnel and the users to be controlled based on multiple first portrait sets, multiple second portrait sets and the ability value. The calculation formula is as follows: ; In the formula, represents the matching degree between the online intervention personnel and the users to be controlled; is a preset weight coefficient; is the th data in the portrait set of the user to be controlled; represents the th data in the th first portrait set constructed; is the total number of data in the first portrait set; is the total number of the first portrait sets constructed; is the influence coefficient corresponding to the th first portrait set constructed; is the th data in the portrait set of the online intervention personnel; is the the th data in the second image set; is the total number of data in the second image set; is the total number of the constructed second image sets; is the influence coefficient corresponding to the th second image set; is the preset influence coefficient corresponding to the image and the ability respectively; is the initial ability value obtained by analyzing the ability of the online intervention personnel through the preset ability analysis library; is the ability value; is the total number of intervention records; is the correction factor determined according to whether the intervention record is positive. When the score feedback in the intervention record is higher than the preset feedback threshold, it is determined to be positive, and the correction factor is positive at this time, otherwise it is negative; represents the adjustment range, represents the supplementary value of the adjustment range; are the upper limit value and the lower limit value of the ability value respectively; is the actual usage time of the th intervention record; is the preset standard intervention time corresponding to the event corresponding to the th intervention record; is the event scoring function, when it is zero, decreases monotonically with the increase of , the minimum is 0, and the maximum can be configured to any one of 2 to 10; Pair the online intervention personnel with the largest matching degree with the controlled users one by one.

[0015] Compared with the prior art, the beneficial effects of the present invention are: The present invention collects the personal information and behavior habits of users through various channels, determines the real-time user data based on artificial intelligence, cleans and transforms the real-time user data based on artificial intelligence, determines the standardized real-time user data based on artificial intelligence, uses machine learning technology to denoise and reduce the dimension of the standardized real-time user data based on artificial intelligence, extracts user feature data, trains a psychological assessment large model using a deep learning algorithm, analyzes the user feature data based on the optimal psychological assessment large model, predicts the user's psychological state and its development trend, provides a psychological counseling plan for the user according to the user's psychological assessment result, intervenes in the user's psychology in a timely manner, and tracks the situation of the user's psychological counseling intervention in real time, optimizes and adjusts the psychological counseling plan according to the user's feedback, so as to meet the ever-changing psychological counseling needs of the user, can accurately evaluate the user's psychological state, thus can formulate a more effective psychological counseling plan, enable the user to receive timely psychological counseling intervention, and can improve the effect of the user's psychological counseling. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 is a flowchart of the psychological assessment and counseling method based on artificial intelligence and large model of the present invention; Figure 2 is a module diagram of the psychological assessment and counseling system based on artificial intelligence and large model of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0018] In order to solve the problems that the existing technology cannot accurately evaluate the user's psychological state, cannot formulate a more effective psychological counseling plan, so that the user cannot receive timely psychological counseling intervention, resulting in poor psychological counseling effect for the user, please refer to Figure 1 - Figure 2 This embodiment provides the following technical solutions: A psychological assessment and counseling method based on artificial intelligence and large model includes the following steps: S1. Acquisition and processing of user real-time data: Collect the personal information and behavior habits of users through various channels, determine the real-time user data based on artificial intelligence, and clean and transform the real-time user data based on artificial intelligence to determine the standardized real-time user data based on artificial intelligence; In this embodiment, collecting the personal information and behavior habits of users through various channels includes: Real-time collection of users' personal information such as name, age, gender, occupation, marital status, health status, contact information, educational background, and family background through various channels such as social media, online questionnaires, and applications to obtain users' personal information; Real-time collection of users' browsing behavior, purchase behavior, social behavior, mobile behavior, entertainment behavior, and reading behavior through various channels such as social media, online questionnaires, and applications to obtain users' behavior habits; Among them, based on the real-time collected users' personal information and behavior habits, determine the users' real-time data based on artificial intelligence, and ensure the security and confidentiality of the users' real-time data based on artificial intelligence in accordance with privacy protection regulations.

[0019] It should be noted that browsing behavior includes web browsing records, search keywords, click behavior, and residence time, etc.; purchase behavior includes purchase history, purchase frequency, purchase preference, and consumption level, etc.; social behavior includes interactions on social media, people followed, content published, likes, and comments, etc.; mobile behavior includes geographical location information, movement trajectory, and travel habits, etc.; entertainment behavior includes watching videos, listening to music, and playing games, etc.; reading behavior includes reading articles, books, and news, etc.

[0020] In this embodiment, the processing of the users' real-time data based on artificial intelligence includes: Import the users' real-time data based on artificial intelligence into a data cleaning tool; Clean the users' real-time data based on artificial intelligence based on the data cleaning tool; Among them, view the type and distribution of the users' real-time data based on artificial intelligence, check the users' real-time data based on artificial intelligence, judge whether there are duplicate values, missing values, and outliers in the users' real-time data based on artificial intelligence, and process the duplicate values, missing values, and outliers existing in the users' real-time data based on artificial intelligence; For duplicate values, delete the duplicate values in the users' real-time data based on artificial intelligence to maintain the uniqueness of the users' real-time data based on artificial intelligence; For missing values, delete the missing values in the users' real-time data based on artificial intelligence, or use filling or interpolation methods to process the missing values in the users' real-time data based on artificial intelligence; For outliers, delete the outliers in the users' real-time data based on artificial intelligence, or use replacement or smoothing methods to process the outliers in the users' real-time data based on artificial intelligence; Perform type and format conversion on the cleaned users' real-time data based on artificial intelligence to reduce the dimensional difference between the users' real-time data based on artificial intelligence and determine the standardized users' real-time data based on artificial intelligence.

[0021] S2. Feature Selection and Extraction: Apply machine learning techniques to denoise and reduce the dimension of the standardized real-time user data based on artificial intelligence, select the feature data that contributes the most to the model performance, and perform feature extraction on the feature data to reduce the dimension of the feature data and create new user feature data; In this embodiment, applying machine learning techniques to denoise and reduce the dimension of the real-time user data based on artificial intelligence, and extracting user feature data, including: Based on the model-based feature selection method, use the feature importance scoring mechanism of the decision tree to select feature data from the standardized real-time user data based on artificial intelligence. By calculating the contribution of each feature data to the model performance, select the feature data that contributes the most to the model performance, and at the same time remove redundant and irrelevant noise feature data; which helps to improve the generalization ability of the model, reduce the risk of overfitting, and accelerate the training speed of the model; Based on machine learning algorithms, perform feature extraction on the selected feature data. Project the feature data into a lower-dimensional space through linear transformation, while retaining most of the variance of the feature data, reduce the dimension of the feature data, and create new user feature data, including user mood changes, behavior patterns, and social interactions.

[0022] S3. Construct a Psychological Assessment Big Model: Use deep learning algorithms to train the psychological assessment big model, and test and optimize the psychological assessment big model to determine the optimal psychological assessment big model. Analyze the user feature data based on the optimal psychological assessment big model to predict the user's psychological state and its development trend; In this embodiment, using deep learning algorithms to train the psychological assessment big model, including: Collect user historical data, process the user historical data, and divide the processed user historical data. Determine the training set and the test set according to the ratio of 7:3; Using deep learning algorithms, train the deep learning model with the training set, so that the deep learning model can learn complex psychological patterns from the user historical data, and enable it to accurately predict the user's psychological state and its development trend, and determine the psychological assessment big model; Based on the test set, perform performance testing on the psychological assessment big model, and use evaluation metrics such as accuracy and recall to evaluate the psychological assessment big model, determine whether the psychological assessment big model can achieve the expected effect, and use cross-validation and adjust hyperparameters to optimize the performance of the psychological assessment big model to determine the optimal psychological assessment big model.

[0023] In this embodiment, predicting the user's psychological state and its development trend, including: Obtain the optimal psychological assessment big model and deploy the optimal psychological assessment big model in the actual psychological assessment and counseling environment; Input the user characteristic data into the optimal psychological assessment large model, analyze the user characteristic data based on the optimal psychological assessment large model, predict the user's psychological state and its development trend, including the user's emotional state, stress level and anxiety degree, and determine the user's psychological assessment result.

[0024] S4. Provide a psychological counseling plan: According to the user's psychological assessment result, provide a psychological counseling plan for the user, intervene in the user's psychology in a timely manner, and track the user's psychological counseling intervention situation in real time. Optimize and adjust the psychological counseling plan according to the user's feedback to meet the user's changing psychological counseling needs.

[0025] In this embodiment, according to the user's psychological assessment result, a psychological counseling plan is provided for the user, including: According to the user's psychological assessment result, provide a psychological counseling plan for the user and intervene in the user's psychology in a timely manner, including online counseling, face-to-face counseling and psychological training course counseling; For online counseling, according to the user's psychological assessment result, recommend a psychological counselor for the user. The user makes an appointment with the psychological counselor through an online platform. After the psychological counselor establishes an online connection with the user, the psychological counselor assesses the user's psychological state and jointly formulates a psychological counseling plan with the user, including the goals, content, methods and time arrangements of psychological counseling. According to the psychological counseling plan, the psychological counselor conducts regular psychological counseling for the user through the online platform. During the psychological counseling process, the psychological counselor provides psychological support, education and intervention for the user; For face-to-face counseling, according to the user's psychological assessment result, recommend a psychological counselor for the user. The user makes an appointment with the psychological counselor through an appointment platform. After the psychological counselor meets with the user, the psychological counselor has a face-to-face conversation with the user. During the conversation, the psychological counselor directly observes the user's non-verbal behaviors, including facial expressions and body languages, so as to understand the user's psychological condition more comprehensively and then conduct psychological counseling for the user; For psychological training course counseling, according to the user's psychological assessment result, recommend psychological training courses for the user, including theoretical explanations, practical exercises and group discussion courses. Select teaching methods and tools according to the course content and user characteristics, enable the user to participate in case analysis, role-playing and psychological tests, focus on the user's interactive behaviors, and encourage the user to ask questions and share experiences; Among them, the user's psychological counseling intervention situation is tracked in real time, the user's psychological state and psychological counseling effect are continuously monitored, and the psychological counseling plan is optimized and adjusted according to the user's feedback to meet the user's changing psychological counseling needs.

[0026] To better demonstrate the implementation process of the psychological assessment and counseling method based on artificial intelligence and large models, this embodiment now provides a psychological assessment and counseling system based on artificial intelligence and large models, which is used to implement the psychological assessment and counseling method based on artificial intelligence and large models as described above, including: A data acquisition and processing module, which is used to acquire real-time user data based on artificial intelligence and process the real-time user data based on artificial intelligence; A feature selection and extraction module, which is used to perform feature selection and extraction on the real-time user data based on artificial intelligence to determine user feature data; A large model construction and prediction module, which is used to construct a psychological assessment large model and predict the user's psychological state and its development trend based on the psychological assessment large model to determine the user's psychological assessment result; A psychological counseling and control module, which is used to provide a psychological counseling plan for the user based on the user's psychological assessment result and timely intervene and control the user's psychology.

[0027] In summary, by processing and extracting features from the user's real-time data, the user's psychological state can be evaluated, and thus a more effective psychological counseling plan can be formulated. In addition, the psychological assessment large model constructed using deep learning algorithms has powerful prediction capabilities, which helps to discover potential psychological problems of users and give timely intervention.

[0028] In order to achieve real-time intervention for users, in one embodiment, the psychological counseling and control module includes: A video data acquisition unit, which is used to acquire video data of a video acquisition device configured at a preset position, and screen the video data to obtain relevant video data of the user to be controlled; The screening mainly extracts the video data with the user to be controlled. An audio data acquisition unit, which is used to acquire audio data collected by an audio acquisition device of a control terminal worn by the user to be controlled, and screen the audio data to obtain data to be analyzed; The screening of the audio data is performed through a preset key trigger word library, and the audio data is screened and intercepted according to the time interception rule corresponding to the key trigger word to obtain the data to be analyzed; A scenario analysis unit, which is used to determine the current scenario based on the relevant video data and the data to be analyzed; A vital sign data acquisition unit, which is used to acquire vital sign data monitored by the control terminal; The vital sign data includes: body temperature, pulse, heart rate, An intervention unit, which is used to comprehensively analyze the current scenario and the vital sign data to determine and execute an intervention strategy. Through the characteristic parameters of the current scenario and the vital sign data, query the pre-configured intervention strategy library to obtain the intervention strategy; The specific intervention strategies include: playing music, electric current stimulation, playing prompt voice, one or more combinations of them; Among them, the scenario analysis unit determines the current scenario based on relevant video data and data to be analyzed, and performs the following operations: Extract features from the relevant video data to obtain multiple first feature parameters; Extract features from the data to be analyzed to obtain multiple second feature parameters; the second feature parameters include: feature parameters representing trigger words, parameters representing the time length of the analysis data, etc.; Arrange the multiple first feature parameters and second feature parameters in the order specified by a preset scenario analysis template to form an analysis data set; Match the analysis data set with the scenario sets corresponding to each scenario in a pre-configured scenario analysis library; Extract the scenario corresponding to the scenario set that matches the analysis data set as the current scenario; Among them, extracting features from the relevant video data to obtain multiple first feature parameters includes: Identify objects and people in the relevant video data to obtain first feature parameters representing each object and person; Analyze the positional relationships between each object and each person in the relevant video data to obtain first feature parameters representing the relative positions of each person and object; Identify the users to be controlled in the relevant video data and calculate the probability of interaction between the users to be controlled and other people according to the following formula. The calculation formula is as follows: ; Among them, is the probability of interaction between the user to be controlled and another person within a certain time, is the time value, is the current position vector of the user to be controlled and another person, is a preset adjustment factor for adjusting the error in speed recognition, is the current speed vector of the user to be controlled and another person, which is obtained by taking the difference of the position vectors between frames, is the probability function, , the variable in the probability function is non-negative, and the value of the probability function decreases monotonically as the variable increases; Generate first feature parameters representing the probability of interaction between the user to be controlled and other people.

[0029] The system corresponding to the present invention also provides online one-on-one real-person remote intervention. Due to the different capabilities of each online intervention personnel and the situations of the users to be controlled, in order to achieve accurate one-on-one personnel intervention online and ensure the intervention quality; in one embodiment, the psychological counseling control module further includes: An online real-time pairing and intervention unit is used to pair online intervention personnel with users to be controlled one by one, and perform one-to-one docking interventions on the users to be controlled based on the paired online intervention personnel. Among them, the online real-time pairing and intervention unit pairs online intervention personnel with users to be controlled one by one and performs the following operations: Analyze the intervention records of online intervention personnel to construct multiple first portrait sets; the first portrait sets are constructed based on the situations of the users to be controlled corresponding to the intervention records; the data in the first portrait sets include: the quantitative data corresponding to the age, gender, scene number, psychological assessment parameters, etc. of the controlled personnel in the intervention records; the influence coefficient corresponding to the first portrait set is the ratio between the number of intervention records corresponding to the first portrait set and the total number. Analyze the intervention records of the users to be controlled to construct multiple second portrait sets; the second portrait sets are constructed based on the situations of each online intervention personnel in the intervention records; the data in the second portrait sets include: the quantitative data corresponding to the age, gender, scene number, ability assessment parameters, etc. of each online intervention personnel in the intervention records; among them, the scene number is the unique identification code determined by querying the preset scene number library according to the scene data in the intervention records or the intervention records; the influence coefficient of the second portrait set is the coefficient value obtained after normalizing the evaluation value obtained by querying the preset evaluation form based on the number of intervention records corresponding to the second portrait set and the evaluation (evaluation by professionals or the responsible personnel of the controlled personnel). Analyze the intervention records of online intervention personnel and the feedback of the corresponding intervention records to determine the ability value. Based on multiple first portrait sets, multiple second portrait sets and the ability value, calculate the matching degree between the online intervention personnel and the controlled users. The calculation formula is as follows: ; In the formula, represents the matching degree between the online intervention personnel and the controlled users; is the preset weight coefficient; is the rd data in the portrait set of the user to be controlled; represents the th first portrait set constructed; rd data in it; is the total number of data in the first portrait set; is the total number of the constructed first portrait sets; is the influence coefficient corresponding to the th constructed first portrait set; is the rd data in the portrait set of the online intervention personnel; represents the The th data in the second image set; is the total number of data in the second image set; is the total number of the constructed second image sets; is the influence coefficient corresponding to the th constructed second image set; is the initial ability value obtained by analyzing the abilities of online intervention personnel through a preset ability analysis library; is the ability value; is the total number of intervention records; is a correction factor determined according to whether the intervention record is positive. When the score feedback in the intervention record is higher than the preset feedback threshold, it is determined to be positive, and the correction factor is positive at this time, otherwise it is negative; represents the adjustment range, represents the supplementary value of the adjustment range; are respectively the upper limit value and the lower limit value of the ability value; is the actual usage time of the th intervention record; is the preset standard intervention time corresponding to the event corresponding to the th intervention record; When it is zero, along with increasing, it monotonically decreases, with a minimum of 0 and a maximum that can be configured to any value from 2 to 10; One-to-one pairing is performed between the online intervention personnel with the highest matching degree and the controlled users.

[0030] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0031] Although the embodiments of the present invention have been shown and described, for those of ordinary skill in the art, it can be understood that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A psychological assessment and counseling method based on artificial intelligence and large models, characterized by: The steps include: S1. Acquisition and processing of real-time user data: Collecting users’ personal information and behavioral habits through various channels, determining real-time user data based on artificial intelligence, and cleaning and converting real-time user data based on artificial intelligence to determine standardized real-time user data based on artificial intelligence; Among them, checking the type and distribution of the real-time user data based on artificial intelligence, checking the real-time user data based on artificial intelligence, judging whether there are duplicate values, missing values ​​and abnormal values ​​in the real-time user data based on artificial intelligence, and processing the duplicate values, missing values ​​and abnormal values ​​in the real-time user data based on artificial intelligence; S2. Feature selection and extraction: Use machine learning technology to denoise and reduce the dimension of standardized AI-based user real-time data, select the feature data that contributes most to model performance, extract features from the feature data, reduce the dimension of the feature data, and create new user feature data; S3. Build a psychological assessment model: Use deep learning algorithms to train the psychological assessment model, test and optimize the psychological assessment model, determine the optimal psychological assessment model, analyze user feature data based on the optimal psychological assessment model, and predict the user's psychological state and development trend; S4. Provide psychological counseling plans: Provide users with psychological counseling plans based on the results of user psychological assessments, intervene in users' psychology in a timely manner, and track users' psychological counseling interventions in real time. Optimize and adjust psychological counseling plans based on user feedback to meet users' changing psychological counseling needs.

2. The psychological assessment and counseling method based on artificial intelligence and large models as claimed in claim 1, characterized in that: In S1, the user's personal information and behavior habits are collected through various channels, including: Collect the user's name, age, gender, occupation, marital status, health status, contact information, educational background and family background in real time through social media, online questionnaires and applications to obtain the user's personal information; Collect users' browsing behavior, purchasing behavior, social behavior, mobile behavior, entertainment behavior and reading behavior in real time through social media, online questionnaires and applications to obtain user behavior habits; Among them, based on the user personal information and user behavior habits collected in real time, the real-time user data based on artificial intelligence is determined, and based on privacy protection regulations, the real-time user data based on artificial intelligence is kept safe and confidential.

3. The psychological assessment and counseling method based on artificial intelligence and large models as claimed in claim 2, characterized in that: In S1, the real-time user data based on artificial intelligence is processed, including: Import AI-based user real-time data into data cleaning tools; Clean real-time user data based on artificial intelligence using data cleaning tools; For duplicate values, delete the duplicate values ​​in the AI-based user real-time data to maintain the uniqueness of the AI-based user real-time data; For missing values, delete the missing values ​​in the AI-based user real-time data, or use filling or interpolation methods to handle the missing values ​​in the AI-based user real-time data; For outliers, delete outliers in the AI-based user real-time data, or use replacement or smoothing methods to process outliers in the AI-based user real-time data; The cleaned AI-based user real-time data is converted in type and format to reduce the dimensional differences between the AI-based user real-time data and determine standardized AI-based user real-time data.

4. The psychological assessment and counseling method based on artificial intelligence and large models as claimed in claim 3, characterized in that: In S2, machine learning technology is used to perform denoising and dimensionality reduction processing on the real-time user data based on artificial intelligence, and to extract user feature data, including: Model-based feature selection method, which uses the feature importance scoring mechanism of decision tree to select feature data from standardized AI-based user real-time data, and selects the feature data that contributes the most to model performance by calculating the contribution of each feature data to model performance, while removing redundant and irrelevant noise feature data; Feature extraction is performed on the selected feature data based on machine learning algorithms, and the feature data is projected into a lower dimensional space through linear transformation, while retaining most of the variance of the feature data, reducing the dimension of the feature data, and creating new user feature data, including user emotional changes, behavioral patterns, and social interactions.

5. The psychological assessment and counseling method based on artificial intelligence and large models as claimed in claim 4, characterized in that: In S3, a deep learning algorithm is used to train a large psychological assessment model, including: Collect user historical data, process the user historical data, and divide the processed user historical data into training sets and test sets according to a ratio of 7:3; Using deep learning algorithms and training sets to train deep learning models, the deep learning models can learn complex psychological patterns from historical user data, predict the user's psychological state and development trend, and determine the psychological assessment model; Based on the test set, the performance of the psychological assessment model is tested, and the accuracy and recall evaluation indicators are used to evaluate the psychological assessment model to determine whether the psychological assessment model can achieve the expected effect. Cross-validation and adjustment of hyperparameters are used to optimize the performance of the psychological assessment model and determine the optimal psychological assessment model.

6. The psychological assessment and counseling method based on artificial intelligence and large models as claimed in claim 5, characterized in that: In S3, predicting the user's psychological state and development trend includes: Obtain the optimal psychological assessment model and deploy it in the actual psychological assessment and counseling environment; Input the user characteristic data into the optimal psychological assessment model, analyze the user characteristic data based on the optimal psychological assessment model, predict the user's psychological state and development trend, including the user's emotional state, stress level and anxiety level, and determine the user's psychological assessment results.

7. The psychological assessment and counseling method based on artificial intelligence and large models as claimed in claim 6, characterized in that: In S4, according to the psychological assessment result of the user, a psychological counseling program is provided to the user, including: Provide users with psychological counseling solutions based on the results of user psychological assessments to intervene in their psychology, including online counseling, face-to-face counseling, and psychological training course counseling; Among them, the user's psychological counseling intervention situation is tracked in real time, the user's psychological state and psychological counseling effect are continuously monitored, and the psychological counseling plan is optimized and adjusted according to user feedback to meet the user's changing psychological counseling needs.

8. A psychological assessment and counseling system based on artificial intelligence and a large model, used to implement the psychological assessment and counseling method based on artificial intelligence and a large model as claimed in claim 7, characterized in that: include: A data acquisition and processing module is used to acquire and process the real-time user data based on artificial intelligence; Feature selection and extraction module, used to select and extract features of real-time user data based on artificial intelligence to determine user feature data; A large model building prediction module is used to build a psychological assessment large model, and predict the user's psychological state and its development trend based on the psychological assessment large model to determine the user's psychological assessment results; The psychological counseling control module is used to provide users with psychological counseling solutions based on the results of user psychological assessments, and to intervene and control users' psychology in a timely manner; The psychological counseling management and control module includes: A video data acquisition unit, used to acquire video data from a video acquisition device configured at a preset position, and to filter the video data to obtain relevant video data of the controlled user; An audio data collection unit is used to collect audio data collected by an audio collection device of a control terminal worn by a controlled user, and to filter the audio data to obtain data to be analyzed; A scenario analysis unit, for determining a current scenario based on relevant video data and data to be analyzed; A life characteristic data collection unit, used to collect life characteristic data monitored by the control terminal; The intervention unit is used to comprehensively analyze the current situation and vital signs data, determine the intervention strategy and implement it.

9. The psychological assessment and counseling system based on artificial intelligence and large models as claimed in claim 8, characterized in that: The scenario analysis unit determines the current scenario based on the relevant video data and the data to be analyzed, and performs the following operations: Extracting features from relevant video data to obtain a plurality of first feature parameters; Perform feature extraction on the data to be analyzed to obtain a plurality of second feature parameters; Arrange the plurality of first characteristic parameters and the second characteristic parameters in an order specified by a preset scenario analysis template to form an analysis data set; Matching the analysis data set with the scenario sets corresponding to each scenario in the pre-configured scenario analysis library; Extracting the scenario corresponding to the scenario set matching the analysis data set as the current scenario; The process of extracting features from relevant video data to obtain a plurality of first feature parameters includes: Identify objects and people in relevant video data to obtain first characteristic parameters representing each object and person; Analyze the positional relationship between various objects and various people in the relevant video data to obtain a first characteristic parameter representing the relative position of each person and object; Determine the controlled user in the relevant video data and calculate the probability of interaction between the controlled user and other characters according to the following formula: ； in, is the probability that the controlled user interacts with another character within a certain period of time. is the time value, is the current position vector of the controlled user and the other person, It is a preset adjustment factor used to adjust the error of speed recognition. is the current velocity vector of the controlled user and the other character, is obtained by differentiating the position vector between frames, is the probability function, , the variable in the probability function is non-negative, and the value of the probability function decreases monotonically as the variable increases; A first characteristic parameter representing a probability of interaction between the controlled user and other characters is generated.

10. The psychological assessment and counseling system based on artificial intelligence and large models as claimed in claim 9, characterized in that: The psychological counseling management and control module also includes: An online real-time pairing intervention unit is used to pair online intervention personnel with users to be controlled one by one, and the paired online intervention personnel perform one-to-one docking intervention on the users to be controlled; Among them, the online real-time pairing intervention unit pairs the online intervention personnel with the users to be controlled one by one and performs the following operations: Analyze the intervention records of online intervention personnel and construct multiple first portrait sets; the first portrait sets are constructed based on the conditions of the users to be controlled corresponding to the intervention records; Analyze the intervention records of the users to be controlled and construct multiple second portrait sets; the second portrait sets are constructed based on the situations of each online intervention person in the intervention records; Analyze the intervention records of online intervention personnel and the corresponding feedback of intervention records to determine the capability value; Based on multiple first portrait sets, multiple second portrait sets and capability values, the matching degree between the online intervention personnel and the controlled users is calculated. The calculation formula is as follows: ； In the formula, It indicates the matching degree between online intervention personnel and the controlled users; is the preset weight coefficient; This is the portrait of the controlled user. individual data; Indicates the constructed The first portrait collection individual data; is the total number of data in the first portrait set; is the total number of first portrait sets constructed; For the corresponding construction The influence coefficient of the first image set; This is the first in a series of portraits of online interventionists. individual data; For the construction of The second portrait collection individual data; is the total number of data in the second portrait set; is the total number of second portrait sets constructed; For the construction of The influence coefficient corresponding to the second image set; The influence coefficients of the preset corresponding portraits and capabilities respectively; To obtain the initial capability value after conducting capability analysis on online intervention personnel through a preset capability analysis library; is the ability value; total number of interventions recorded; The correction factor is determined to indicate whether the intervention record is positive. When the feedback score in the intervention record is higher than the preset feedback threshold, it is determined to be positive. In this case, the correction factor is positive, otherwise it is negative. Indicates the adjustment range. A supplementary value indicating the adjustment range; They are the upper and lower limits of the ability value respectively; For the The actual time of use of each intervention record; For the The preset standard intervention time corresponding to the event corresponding to each intervention record; is the event scoring function, When it is zero, along with The increase of decreases monotonically, the minimum is 0, and the maximum can be configured to any value between 2 and 10; The online intervention personnel with the highest matching degree are paired one-to-one with the controlled users.

Images