WO2022166110A1 - 心理辅导训练方案的确定方法及装置 - Google Patents

心理辅导训练方案的确定方法及装置 Download PDF

Info

Publication number
WO2022166110A1
WO2022166110A1 PCT/CN2021/105287 CN2021105287W WO2022166110A1 WO 2022166110 A1 WO2022166110 A1 WO 2022166110A1 CN 2021105287 W CN2021105287 W CN 2021105287W WO 2022166110 A1 WO2022166110 A1 WO 2022166110A1
Authority
WO
WIPO (PCT)
Prior art keywords
training
user
classification model
cycle
result
Prior art date
Application number
PCT/CN2021/105287
Other languages
English (en)
French (fr)
Inventor
司马华鹏
华冰涛
汤毅平
汪成
Original Assignee
南京硅基智能科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 南京硅基智能科技有限公司 filed Critical 南京硅基智能科技有限公司
Priority to US17/594,499 priority Critical patent/US11682482B2/en
Priority to EP21773443.3A priority patent/EP4064289A4/en
Publication of WO2022166110A1 publication Critical patent/WO2022166110A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H20/00—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance
    • G16H20/70—ICT specially adapted for therapies or health-improving plans, e.g. for handling prescriptions, for steering therapy or for monitoring patient compliance relating to mental therapies, e.g. psychological therapy or autogenous training
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G06N20/10—Machine learning using kernel methods, e.g. support vector machines [SVM]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/044—Recurrent networks, e.g. Hopfield networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H80/00—ICT specially adapted for facilitating communication between medical practitioners or patients, e.g. for collaborative diagnosis, therapy or health monitoring

Definitions

  • the present disclosure relates to the technical field of psychological robotics, and in particular, to a method and device for determining a psychological counseling training program.
  • Psychological robots are products based on artificial intelligence to provide psychological treatment/counseling for users. During the use process, psychological robots can determine the psychological problems of users according to the user's choice or interaction with users, and then adopt appropriate methods to treat users. With psychotherapy or counseling, users can adjust their own psychological problems without the assistance of a third party such as a psychologist.
  • an application scenario involving interaction with a user within a certain period is often performed according to a preset fixed process or a fixed strategy to complete the interaction with the user.
  • a preset fixed process or a fixed strategy to complete the interaction with the user.
  • the process of psychotherapy/counseling by the psychological robot since the effect of the user's practice often varies from person to person, it is possible that the user did not achieve the expected practice effect at the end of the current time period, while the psychological robot starts in the next time period. At the same time, the user will still carry out the corresponding exercises in the next stage according to the originally set strategy. This situation not only fails to have the expected effect on the user, but even causes negative feedback to the user, causing the user to terminate the entire psychotherapy/counseling process in advance.
  • the user should complete at least seven exercises in the current time period, but the user actually only performs one exercise in the current time period, then the same At the end of the current time period, the expected exercise effect cannot be achieved. In this state, the corresponding exercise for the user in the next stage according to the original strategy will also affect the user's subsequent exercise effect and the overall treatment/counseling effect.
  • psychological robots provide psychological counseling to users according to a pre-set counseling process, which cannot meet the individual needs of different users, resulting in unsatisfactory counseling results. There is currently no effective solution.
  • the present disclosure provides a method and device for determining a psychological counseling training scheme, so as to at least solve the problem that psychological robots in the related art provide psychological counseling to users according to a preset counseling process, which cannot meet the individual needs of different users, resulting in counseling results. Not ideal question.
  • a method for determining a psychological counseling training program includes: acquiring training experience data of the user after each psychological counseling training of the user through interactive inquiry with the user, wherein the training The feeling data is used to indicate the user's feeling after psychological counseling; the training feeling data is input into the first classification model, and the training result of the user after each psychological counseling training is identified by the first classification model , and count the training results of the user in the current training cycle, wherein the first classification model is a model obtained by training the initial super-long text classification model; according to the training results of the user in the current training cycle A training scheme for the next training period of the user is determined.
  • a device for determining a psychological counseling training scheme including: an acquisition module, configured to acquire training experience data of the user after each psychological counseling training through an interactive inquiry with the user , wherein the training experience data is used to indicate the user's feelings after psychological counseling; a processing module is used to input the training experience data into a first classification model, and identify the user's experience in the first classification model through the first classification model The training results after each psychological counseling training are counted, and the training results of the users in the current training cycle are counted, wherein the first classification model is a model trained according to the initial super-long text classification model; determining module , which is used to determine the training scheme of the user in the next training period according to the training result of the user in the current training period.
  • a computer-readable storage medium is also provided, where a computer program is stored in the storage medium, wherein the computer program is configured to execute any one of the above method embodiments when running steps in .
  • an electronic device comprising a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute Steps in any one of the above method embodiments.
  • the user's training experience data after each psychological counseling training is obtained, and the training experience data is input into the first classification model, and the first classification model is used to identify the user's experience in each psychological counseling session.
  • the training results after training are collected, and the training results of the user in the current training period are counted; the training scheme of the user in the next training period is determined according to the training results of the user in the current training period.
  • FIG. 1 is a block diagram of the hardware structure of a mobile terminal of a method for determining a psychological counseling training scheme according to an embodiment of the present disclosure
  • FIG. 2 is a flowchart of a method for determining an optional psychological counseling training program according to an embodiment of the present disclosure
  • FIG. 3 is a schematic diagram of an optional user training result in a complete time period according to an embodiment of the present disclosure
  • FIG. 4 is a structural block diagram of a device for determining an optional psychological counseling training scheme according to an embodiment of the present disclosure
  • FIG. 5 is a schematic structural diagram of an optional electronic device according to an embodiment of the present disclosure.
  • FIG. 1 is a hardware structural block diagram of a mobile terminal for a method for determining a psychological counseling training scheme according to an embodiment of the present disclosure.
  • the mobile terminal may include one or more (only one is shown in FIG.
  • processors 102 may include, but are not limited to, processing devices such as a microprocessor MCU or a programmable logic device FPGA) and a memory 104 for storing data, wherein the above-mentioned mobile terminal may further include a transmission device 106 and an input and output device 108 for communication functions.
  • the structure shown in FIG. 1 is only a schematic diagram, which does not limit the structure of the above-mentioned mobile terminal.
  • the mobile terminal may also include more or fewer components than those shown in FIG. 1 , or have a different configuration than that shown in FIG. 1 .
  • the memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer programs corresponding to the method for determining the psychological counseling training program in the embodiments of the present disclosure.
  • the processor 102 runs the computer programs stored in the memory 104 , so as to perform various functional applications and data processing, that is, to implement the above method.
  • Memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory.
  • the memory 104 may further include memory located remotely from the processor 102, and these remote memories may be connected to the mobile terminal through a network. Examples of such networks include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.
  • Transmission means 106 are used to receive or transmit data via a network.
  • the specific example of the above-mentioned network may include a wireless network provided by a communication provider of the mobile terminal.
  • the transmission device 106 includes a network adapter (Network Interface Controller, NIC for short), which can be connected to other network devices through a base station so as to communicate with the Internet.
  • the transmission device 106 may be a radio frequency (Radio Frequency, RF for short) module, which is used to communicate with the Internet in a wireless manner.
  • RF Radio Frequency
  • FIG. 2 is a flowchart of a method for determining an optional psychological counseling training scheme according to an embodiment of the present disclosure. As shown in FIG. 2 , the method includes:
  • Step S202 through the interactive inquiry with the user, obtain the training experience data of the user after each psychological counseling training, wherein the training feeling data is used to indicate the user's feeling after the psychological counseling training;
  • Step S204 input the training experience data into the first classification model, identify the training results of the user after each psychological counseling training by the first classification model, and count the training results of the user in the current training cycle, wherein the first classification model is the model obtained by training the initial super-long text classification model;
  • Step S206 determining the training scheme of the user in the next training period according to the training result of the user in the current training period.
  • the execution subject of the method in the embodiment of the present disclosure may be a psychological robot, and the psychological robot may be implemented by any form of application program, such as an APP, a WeChat applet, or an inherent program carried in a preset terminal.
  • the execution terminal involved in the method provided by the embodiment of the present disclosure may include a user terminal and a server, and the server may be a cloud server or a local server.
  • the user terminal is used to carry the psychological robot and interact with the user, so as to realize the psychological treatment/counseling for the user.
  • User equipment includes but is not limited to mobile phones, tablet computers, PCs, wearable devices, indoor large-screen terminals, outdoor large-screen terminals, etc.
  • the server is configured to determine, according to the input data of the user, a desired psychotherapy/counseling mode or a psychotherapy/counselling mode suitable for the user, and subsequently perform psychotherapy/counselling on the user according to the selected mode.
  • the training feeling data involved in the embodiments of the present disclosure may include: the user's thoughts after training, the user's physical feeling after training, the user's emotional feeling after training, etc.
  • the physical feeling after training can be further embodied They are: head feeling, shoulder and neck feeling, limb feeling, spine feeling, feeling of the core part of waist and abdomen, etc., which are not limited in the embodiments of the present disclosure.
  • the training experience data fed back by the user after completing the training can be:
  • the first classification model can be a super-long text classification model based on BERT+LSTM, or a super-long text classification model based on BERT+LSTM+CRF.
  • a machine learning model for natural language recognition and classification is sufficient, which is not limited in this embodiment of the present disclosure.
  • step S204 may be implemented by the following steps:
  • S1 input the user's training experience data after each psychological counseling training into the first classification model, and generate an N1-dimensional vector corresponding to the training result, wherein each dimension vector in the N1-dimensional vector corresponds to a category of training experience data, respectively, Each dimension vector uses a number to indicate the user's training experience level;
  • S3 Generate an N2*M-dimensional vector from the training results of the user in the second training period, where M represents the number of first training periods included in each second training period.
  • x corresponds to the user's thoughts after training
  • y corresponds to the user's physical feeling after training
  • z corresponds to the user's emotional feeling after training.
  • the training feeling level can be represented by -1, 0, 1 to indicate bad feeling, general feeling, and good feeling respectively.
  • a corresponding 3-dimensional vector (1, 0, 1) is generated to indicate that after the user training
  • the thoughts are good, the physical feeling is average, the emotional feeling is good, and the overall feeling should be relatively positive.
  • the training feeling level can also use 1, 2, 3, 4, and 5 to indicate the degree of feeling better, for example, 1 means the worst, 5 means the best, or 1 means the best, and 5 means the worst.
  • the training experience levels are not limited to the above three or five levels, and may be set according to actual requirements, which are not limited in the embodiment of the present disclosure.
  • the co-location addition described in the embodiment of the present disclosure can be understood as: when multiple psychological counseling trainings are completed within the first training period, for example, within the same day (the first training period is one day), four psychological counseling training sessions are completed
  • the 3-dimensional vectors corresponding to the four training results are (x1, y1, z1), (x2, y2, z2), (x3, y3, z3) and (x4, y4, z4).
  • the total training results in one day are counted in the form of a 3-dimensional vector, expressed as (x1+x2+x3+x4, y1+y2+y3+y4, z1+z2+z3+z4).
  • the current 3-dimensional vector can be set as a default vector, such as (0, 0, 0).
  • the second training cycle can be understood as a treatment/counseling session, eg one week (seven days).
  • the 3-dimensional vector of the day can be set as a default vector, such as (0, 0, 0).
  • the vectors in the second training cycle are integrated together to form a 3*7-dimensional vector, that is, a 21-dimensional vector.
  • the above method before inputting the user's training experience data after each psychological counseling training into the first classification model, and before generating an N1-dimensional vector corresponding to the training result, the above method further includes:
  • first sample data where the first sample data includes at least one of the following categories: data used to describe thoughts after training, data used to describe physical feelings after training, and data used to describe emotional feelings after training The data;
  • the first classification model can collect the first sample data related to the user's thoughts, physical feelings, and emotional feelings of the user before or during the training of the model, and the above-mentioned sample data can be analyzed by psychologists. Labeling, or clustering and labeling the sample data by the computer after identifying the keywords, for example, 1 means that the tendency is better, 0 means that the tendency is no change, -1 means that the tendency is worse; Cases can also be marked with 0.
  • a classification model is trained separately.
  • the first sample data used to describe the post-training thoughts can be used to train a classification model for the user's post-training thoughts
  • the data used to describe the post-training bodily feelings can be used to train a classification model for the user's post-training bodily feelings.
  • the data describing the emotional feelings after training can be used to train a classification model for the user's emotional feelings after training.
  • the structures of the above-mentioned multiple classification models may be the same or may be different from each other.
  • the method further includes:
  • the number of training times completed by the user in the second training cycle is counted as the N2*M+1 dimensional vector.
  • the psychological robot or the server may count the number of exercises of the user, that is, count the exercise frequency of the user in the second training period.
  • the user in the second training cycle of one week, the user completes the exercise twice a day on Monday, Wednesday, Friday and Sunday, and once a day on Tuesday, Thursday and Saturday, after the end of the week, the user can record The user has practiced 11 times in total, and can use 11 as the 22nd dimension vector.
  • step S206 may be implemented by the following steps:
  • the mental robot asks the user for a summary of the exercises performed in the current second training period.
  • the user indicates the current second training period If the user expresses "I have achieved good training results this week and I feel that I have reached the initial goal", the next stage of practice will be carried out in the next time period according to the preset process.
  • the user does not expressly indicate that the exercise effect in the current cycle has reached expectations, for example, the user expresses "I feel so-so during the training this week", “I don't seem to have achieved the goal at the beginning”, etc., or, the user Failing to provide feedback to the inquiries of the psychological robot, according to the statistics of the practice situation of the user in each exercise in the current second training cycle, and the number of exercises of the user, the pre-trained second classification model is used to predict the next stage suitable for the user. Practice methods, and then adjust the user's training strategy in the next stage.
  • the second classification model can be an SVM classifier, and the training and prediction speed of the SVM classifier is more efficient than that of the classification model based on deep learning; at the same time, the data trained by the SVM classifier is the processed features, not In relation to semantic information, the feature vector output by the first classification model can be directly used as input data.
  • the use of the SVM classifier can improve the prediction efficiency of the second classification model on the premise of realizing the functions of the second classification model, and at the same time realize the lightweight of the overall model.
  • the training results of the user in the current second training cycle are input into the second classification model, and the second classification model is used to determine the training scheme of the user in the next second training cycle, which can be achieved through the following steps:
  • the L-dimensional vector includes: the N2*M-dimensional vector corresponding to the training result of the user in the current training cycle, Or, the N2*M+1-dimensional vector corresponding to the user's training results and training times in the current training cycle;
  • S3 Determine the training program of the user in the next second training period according to the result of the user's psychological tendency.
  • the feature vector input to the second classification model may be an N2*M-dimensional vector corresponding to the user's training result in the current training cycle, or may be an N2*M+1-dimensional vector including the number of training times.
  • the dimension of the vector processed by the second classification model is consistent with the dimension of the L-dimensional vector.
  • the processing dimension of the second classification model is N2*M dimension.
  • the processing dimension of the second classification model is N2*M+1 dimension.
  • the corresponding second classification model is a model for processing 21-dimensional vectors.
  • the corresponding second classification model Classification models are models that deal with 22-dimensional vectors.
  • the second classification model before determining the training scheme of the user in the next second training cycle by using the second classification model, the second classification model can be trained in the following manner:
  • the SVM classifier pre-training is required, and the 21-dimensional vector or 22-dimensional vector obtained by different users or users in different training periods is used as sample data, and the second sample data is marked by psychologists, or The computer uses a machine learning algorithm to label the second sample data by identifying keywords.
  • the way of marking can be expressed by 1 to indicate the tendency to be satisfied, -1 to indicate average or dissatisfied, or five levels from 1 to 5 to indicate the degree of satisfaction, 1 for the least satisfied, 5 for the most satisfied, etc., the more marked the category.
  • the more detailed the strategy the better the fit between the strategy adjusted by the psychological robot and the needs of the user, which is not limited in this embodiment of the present disclosure.
  • the labeled sample data is input into the SVM classifier for training, and the trained SVM classifier can be obtained.
  • the method further includes:
  • the L-dimensional vector corresponding to the statistical result can be marked as 1, indicating satisfaction, and the marked L-dimensional feature vector can be input into the SVM classifier for analysis. Training is used to update the SVM classifier, so that with the continuous use of the psychological robot by the user, the SVM classifier is continuously updated and the prediction effect is more accurate.
  • the psychological robot asks the user's training experience, and the inquiring process may include the following questions: the user's thoughts after training, the user's physical feeling after training, and the user's emotional feeling after training.
  • the method of the above query may be that after asking the user a question, the user can directly input it through text or voice, or it may be to provide the user with options such as text/rating for the user to give feedback through a click operation.
  • the following table provides feedback from multiple users:
  • the user gives feedback to the questions in S11 respectively.
  • the psychological robot identifies the above-mentioned user feedback through a pre-trained classification model (equivalent to the aforementioned first classification model) to evaluate the user's performance in this exercise. happening.
  • the psychological robot needs to record the user's exercise situation, and also needs to count the number of times of the user's exercise.
  • (1, 1, 1) can be used to express its current practice status
  • (-1, 0, -1) can be used to express its current practice status
  • (0, -1) for the above User003 , -1) Express the state of the exercise.
  • FIG. 3 is an optional schematic diagram of a user training result in a complete time period according to an embodiment of the present disclosure.
  • the exercise status of each exercise performed by the user in each day can be determined by the corresponding three-dimensional The vector is represented, and the three-dimensional vector corresponding to the practice status of the user's three exercises on the first day is accumulated in the same position, and a new three-dimensional vector used to represent the user's practice status on the first day can be obtained, that is, the above figure.
  • the practice status of the user on each day of the week can be determined by accumulating the practice status recorded after all the exercises in the day. In this way, seven three-dimensional vectors can be obtained, which are aggregated to form a 21-dimensional vector, and the 21-dimensional vector represents the user's practice status in the week.
  • the number of times the user exercises in the week can be further counted.
  • the user's practice status and practice times in a week can be expressed in the form of a 22-dimensional feature vector.
  • the user can record the user's fixed The practice status and practice times in the time period.
  • the user does not expressly indicate that the exercise effect in the current cycle has reached expectations, for example, the user expresses "I feel so-so during the training this week", "I don't seem to have achieved the goal at the beginning", etc., or, the user Failing to give feedback to the inquiry of the psychological robot, then according to the practice situation of each exercise of the user determined in S12 and the number of exercises of the user, the pre-trained SVM classifier is used to predict the exercise mode suitable for the user in the next stage, and then Adjust the user's strategy for the next stage.
  • the training and prediction speed of the SVM classifier is more efficient than the classification model based on deep learning; at the same time, the data trained by the SVM classifier is the processed feature and does not involve semantic information, so the effect is similar to that based on deep learning.
  • the classification model is equivalent.
  • the above-mentioned SVM classifier needs to be pre-trained, and the 22-dimensional vectors in S12 obtained by different users or users in different time periods are used as sample data, and the above-mentioned sample data are marked by psychologists.
  • the way of marking can be expressed by 1 for the tendency to be satisfied, and -1 for average or dissatisfied.
  • the labeled sample data is input into the SVM classifier for training, and the trained SVM classifier can be obtained.
  • the 22-dimensional feature vector used to represent the user's exercise status and number of exercises can be input into the SVM classifier, and the SVM classifier can analyze the The feature vector is identified to determine the label corresponding to the feature vector.
  • the following table shows the labels determined by the SVM classifier according to the practice status and practice times of different users:
  • the recognition result of User001 and User002 is 1, which is satisfied. Therefore, for User001 and User002, the next stage of practice can be performed according to the preset process. For User003, the recognition result is -1, which is not satisfied. For User003, the practice effect in the current time period is not ideal, so the psychological robot can re-practice the user in the previous practice method in the next time period, and can also use other practice methods to practice the user.
  • the 22-dimensional feature vector used to represent the user's exercise status and the number of exercises can be assigned to the label 1, and the labeled 22-dimensional feature vector can be assigned to the label 1.
  • the feature vector is input into the SVM classifier for training to update the SVM classifier, so that with the continuous use of the psychological robot by the user, the SVM classifier is continuously updated and the prediction effect is more accurate.
  • the psychological counseling training scheme involved in the embodiments of the present disclosure can be provided by a psychological robot, and the psychological robot can provide various treatment methods including mindfulness therapy, sleep aid therapy, CBT cognitive therapy therapy, relaxation therapy, etc.
  • the psychological robot will provide users with different mindfulness practice methods every week during the 8-week course of treatment, such as mindful meditation, mindful breathing, mindful exercise, mindful eating, etc., so that users can realize the realization through mindfulness
  • the stripping of one's own negative emotions to achieve the purpose of psychotherapy Limited by the personality of each user, the independence of living habits and the reasons for their psychological impact, the suitable treatment methods are different subjectively and objectively. Users can achieve excellent results in mindfulness therapy.
  • the next treatment cycle can be adjusted to another treatment plan that is more suitable for the user's training habits according to the number of exercises or the unsatisfactory data of a certain training experience, or Repeat the training content of the previous treatment cycle.
  • the feature application method in the judgment and identification process adopted in the embodiments of the present disclosure not only simply identifies the user's tendency, but also makes a certain tendency of the user more prominent by means of co-location accumulation, thereby increasing the number of users in the identification process. accuracy, and the corresponding strategy can be further adjusted according to actual needs.
  • the classifier can be cyclically trained and updated during the use of the user, so that the psychological robot can significantly improve the recognition or judgment effect with the increase of the number of users and the use time of the user.
  • a device for determining a psychological counseling training program for implementing the above-mentioned method for determining a psychological counseling training program.
  • the device is used to implement the above embodiments and preferred implementations, and what has been described will not be repeated.
  • the term "module" may be a combination of software and/or hardware that implements a predetermined function.
  • the apparatus described in the following embodiments is preferably implemented in software, implementations in hardware, or a combination of software and hardware, are also possible and contemplated.
  • 4 is a structural block diagram of a device for determining an optional psychological counseling training scheme according to an embodiment of the present disclosure. As shown in FIG. 4 , the device includes:
  • the obtaining module 402 is configured to obtain the training experience data of the user after each psychological counseling training through an interactive inquiry with the user, wherein the training experience data is used to indicate the user's feeling after the psychological counseling;
  • the processing module 404 is configured to input the training experience data into a first classification model, identify the training results of the user after each psychological counseling training through the first classification model, and count the user's The training result in the training period, wherein the first classification model is a model obtained by training according to the initial super-long text classification model;
  • the determining module 406 is configured to determine the training scheme of the user in the next training period according to the training result of the user in the current training period.
  • the processing module 404 includes:
  • the input unit is used to input the training experience data of the user after each psychological counseling training into the first classification model, and generate an N1-dimensional vector corresponding to the training result, wherein each of the N1-dimensional vectors
  • the dimension vectors respectively correspond to the training experience data of one category, and each dimension vector uses a number to indicate the user's training experience level;
  • the first generating unit is configured to generate the N1-dimensional vector corresponding to the multiple times of the training results of the user in the first training cycle when the user has completed multiple times of psychological counseling training in the first training cycle. performing co-position addition to generate an N2-dimensional vector corresponding to the training result of the user in the first training cycle;
  • the second generating unit is configured to generate an N2*M dimensional vector from the training results of the user in the second training cycle, wherein M represents the first training cycle included in each second training cycle. number.
  • processing module 404 further includes:
  • a first obtaining unit configured to obtain first sample data, wherein the first sample data includes at least one of the following categories: data used to describe thoughts after training, data used to describe physical feelings after training, Data used to describe emotional feelings after training;
  • a first labeling unit configured to label the training experience levels of the first sample data of each category respectively
  • a first training unit configured to use the marked first sample data to train the first classification model, wherein the first sample data of different categories corresponds to training the first classification models of different categories.
  • processing module 404 further includes:
  • a statistical unit configured to count the number of training times completed by the user in the second training period, as an N2*M+1 dimensional vector.
  • the determining module 406 includes:
  • the second obtaining unit is used to obtain the feedback result of the user after the current second training cycle is completed through the interactive inquiry with the user;
  • a first determining unit configured to determine the training scheme of the user in the next second training cycle according to a preset training process when the user's feedback result indicates that the expected effect has been achieved
  • a second determining unit configured to input the training result of the user in the current second training cycle into the second classification model when the user's feedback result indicates that the expected effect has not been achieved or the user has not explicitly given feedback , determining the training scheme of the user in the next second training cycle by using the second classification model, wherein the second classification model is a model obtained by training an initial classifier model.
  • the second determining unit includes:
  • the input subunit is used to input the L-dimensional vector corresponding to the statistical result of the user in the current second training cycle into the second classification model, wherein the L-dimensional vector includes: the user is in a current training cycle The N2*M dimensional vector corresponding to the training result in the user, or, the N2*M+1 dimensional vector corresponding to the training result and the number of training times of the user in the current training cycle;
  • an output subunit configured to output the psychological tendency result of the user through the second classification model, wherein the psychological tendency result is used to indicate whether the training of the user in the current second training cycle achieves the expected effect
  • a determination subunit configured to determine a training scheme of the user in the next second training period according to the psychological tendency result of the user.
  • the determining module 406 further includes:
  • a third acquiring unit configured to acquire second sample data consisting of L-dimensional vectors corresponding to the statistical results of the user in a second training cycle
  • the second labeling unit is configured to label the psychological tendency result corresponding to the second sample data, wherein the psychological tendency result includes at least one of the following: satisfied and dissatisfied;
  • a second training unit configured to train the second classification model by using the labeled second sample data.
  • the determining module 406 further includes:
  • a fourth obtaining unit configured to obtain the target L-dimensional vector corresponding to the statistical result of the user in the current second training period
  • a third labeling unit configured to label the psychological tendency result corresponding to the target L-dimensional vector as satisfactory
  • an updating unit configured to input the labeled target L-dimensional vector into the second classification model for training and update the second classification model.
  • an electronic device for implementing the method for determining the above-mentioned psychological counseling training scheme, and the above-mentioned electronic device can be applied to a server but is not limited to.
  • the electronic device includes a memory 502 and a processor 504, where a computer program is stored in the memory 502, and the processor 504 is configured to execute the steps in any one of the above method embodiments through the computer program.
  • the above-mentioned electronic apparatus may be located in at least one network device among multiple network devices of a computer network.
  • the above-mentioned processor may be configured to execute the following steps through a computer program:
  • S3 Determine the training scheme of the user in the next training period according to the training result of the user in the current training period.
  • FIG. 5 is for illustration only, and the electronic device may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a handheld computer, and a mobile Internet device (Mobile Internet device). Internet Devices, MID), PAD and other terminal equipment.
  • FIG. 5 does not limit the structure of the above electronic device.
  • the electronic device may also include more or less components than those shown in FIG. 5 (eg, network interfaces, etc.), or have a different configuration than that shown in FIG. 5 .
  • the memory 502 may be used to store software programs and modules, such as program instructions/modules corresponding to the method and apparatus for determining a psychological counseling training program in the embodiments of the present disclosure.
  • the processor 504 executes the software programs and modules stored in the memory 502 by running the software programs and modules. , so as to perform various functional applications and data processing, that is, to realize the method for determining the above-mentioned psychological counseling training program.
  • Memory 502 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, memory 502 may further include memory located remotely from processor 504, and these remote memories may be connected to the terminal through a network.
  • the memory 502 may specifically, but not be limited to, be used to store the program steps of the method for determining the psychological counseling training program.
  • the memory 502 may include, but is not limited to, the acquiring module 402 , the processing module 404 , the determining module 406 and the like in the apparatus for determining the psychological counseling training scheme.
  • it may also include, but is not limited to, other module units in the apparatus for determining the psychological counseling training scheme, which will not be repeated in this example.
  • the above-mentioned transmission device 506 is configured to receive or send data via a network.
  • Specific examples of the above-mentioned networks may include wired networks and wireless networks.
  • the transmission device 506 includes a network adapter (Network Interface Controller, NIC), which can be connected to other network devices and routers through a network cable so as to communicate with the Internet or a local area network.
  • the transmission device 506 is a radio frequency (RF) module, which is used for wirelessly communicating with the Internet.
  • RF radio frequency
  • the above-mentioned electronic device further includes: a display 508 for displaying an alarm push of suspicious accounts; and a connection bus 510 for connecting various module components in the above-mentioned electronic device.
  • Embodiments of the present disclosure also provide a computer-readable storage medium, where a computer program is stored in the storage medium, wherein the computer program is configured to execute the steps in any one of the above method embodiments when running.
  • the above-mentioned storage medium may be configured to store a computer program for executing the following steps:
  • S3 Determine the training scheme of the user in the next training period according to the training result of the user in the current training period.
  • the storage medium is further configured to store a computer program for executing the steps included in the method in the foregoing embodiment, which will not be repeated in this embodiment.
  • the storage medium may include: a flash disk, a read-only memory (Read-Only Memory, ROM), a random access device (Random Access Memory, RAM), a magnetic disk or an optical disk, and the like.
  • the integrated units in the above-mentioned embodiments are implemented in the form of software functional units and sold or used as independent products, they may be stored in the above-mentioned computer-readable storage medium.
  • the technical solutions of the present disclosure essentially or the parts that contribute to the prior art, or all or part of the technical solutions can be embodied in the form of software products, and the computer software products are stored in a storage medium,
  • Several instructions are included to cause one or more computer devices (which may be personal computers, servers, or network devices, etc.) to perform all or part of the steps of the methods described in various embodiments of the present disclosure.
  • the disclosed user terminal may be implemented in other manners.
  • the apparatus embodiments described above are only illustrative, for example, the division of the units is only a logical function division, and there may be other division methods in actual implementation, for example, multiple units or components may be combined or Integration into another system, or some features can be ignored, or not implemented.
  • the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of units or modules, and may be in electrical or other forms.
  • the units described as separate components may or may not be physically separated, and components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.
  • each functional unit in each embodiment of the present disclosure may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.
  • the above-mentioned integrated units may be implemented in the form of hardware, or may be implemented in the form of software functional units.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • Primary Health Care (AREA)
  • Epidemiology (AREA)
  • Biomedical Technology (AREA)
  • Mathematical Physics (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Artificial Intelligence (AREA)
  • Social Psychology (AREA)
  • Child & Adolescent Psychology (AREA)
  • Developmental Disabilities (AREA)
  • Hospice & Palliative Care (AREA)
  • Psychiatry (AREA)
  • Psychology (AREA)
  • Pathology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Molecular Biology (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Databases & Information Systems (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)
  • Electrically Operated Instructional Devices (AREA)

Abstract

一种心理辅导训练方案的确定方法及装置，所述方法包括：通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据（S202）；将训练感受数据输入第一分类模型，通过第一分类模型识别用户在每一次心理辅导训练后的训练结果，并统计用户在当前一个训练周期内的训练结果（S204）；根据用户在当前一个训练周期内的训练结果确定用户下一个训练周期的训练方案（S206）。解决了心理机器人按照预先设定好的辅导流程对用户进行心理辅导，无法满足不同用户的个性化需求，导致辅导结果不理想的问题。

Description

心理辅导训练方案的确定方法及装置
本公开要求在2021年2月2日提交中国专利局、公开号为202110141410.8、发明名称为“心理辅导训练方案的确定方法及装置”的中国专利公开的优先权，其全部内容通过引用结合在本公开中。
技术领域
本公开涉及心理机器人技术领域，具体而言，涉及一种心理辅导训练方案的确定方法及装置。
背景技术
心理机器人是基于人工智能实现的对用户进行心理治疗/辅导的产品，心理机器人可以在使用过程中，根据用户的选择或与用户的交互以确定用户存在的心理问题，进而采用合适的方式对用户进行心理治疗或辅导，用户可在无需心理医生等第三方的协助下，调节自身的心理问题。
相关技术中，涉及到在一定周期内与用户进行交互的应用场景，往往是按照预设的固定流程，或采用固定的策略与用户完成交互。这种交互方式存在一些弊端。在心理机器人进行心理治疗/辅导过程中，由于用户进行练习的效果往往因人而异，有可能用户在当前的时间周期结束时并未起到预期的练习效果，而心理机器人在下一个时间周期开始时，依然按照原先设定的策略对用户进行下一阶段的相应练习。这种情况不仅无法对用户起到预期效果，甚至还会引起对用户的负反馈，造成用户提前终止整个心理治疗/辅导的流程。另一方面，用户使用心理机器人的频率也难以固定，例如，按照预设流程，用户在当前时间周期内应完成至少七次练习，但用户在当前时间周期内，实际仅进行了一次练习，则同样在当前的时间周期结束时无法起到预期的练习效果，该状态下按照原先的策略对用户进行下一阶段的相应练习，同样会对用户后续练习效果以及整体的治疗/辅导效果造成影响。
针对相关技术中，心理机器人按照预先设定好的辅导流程对用户进行心理辅导，无法满足不同用户的个性化需求，导致辅导结果不理想的问题，目前尚未有有效的解决办法。
发明内容
本公开提供了一种心理辅导训练方案的确定方法及装置，以至少解决相关技术中心理机器人按照预先设定好的辅导流程对用户进行心理辅导，无法满足不同用户的个性化需求，导致辅导结果不理想的问题。
在本公开的一个实施例中，提出了一种心理辅导训练方案的确定方法，包括：通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据，其中，所述训练感受数据用于指示所述用户进行心理辅导后的感受；将所述训练感受数据输入第一分类模型，通过所述第一分类模型识别所述用户在每一次所述心理辅导训练后的训练结果，并统计所述用户在当前一个训练周期内的训练结果，其中，所述第一分类模型为对初始超长文本分类模型训练得到的模型；根据所述用户在当前一个训练周期内的训练结果确定所述用 户下一个训练周期的训练方案。
在本公开的一个实施例中，还提出了一种心理辅导训练方案的确定装置，包括：获取模块，用于通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据，其中，所述训练感受数据用于指示所述用户进行心理辅导后的感受；处理模块，用于将所述训练感受数据输入第一分类模型，通过所述第一分类模型识别所述用户在每一次所述心理辅导训练后的训练结果，并统计所述用户在当前一个训练周期内的训练结果，其中，所述第一分类模型为根据初始超长文本分类模型训练得到的模型；确定模块，用于根据所述用户在当前一个训练周期内的训练结果确定所述用户下一个训练周期的训练方案。
在本公开的一个实施例中，还提出了一种计算机可读的存储介质，所述存储介质中存储有计算机程序，其中，所述计算机程序被设置为运行时执行上述任一项方法实施例中的步骤。
在本公开的一个实施例中，还提出了一种电子装置，包括存储器和处理器，其特征在于，所述存储器中存储有计算机程序，所述处理器被设置为运行所述计算机程序以执行上述任一项方法实施例中的步骤。
通过本公开实施例，通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据,将训练感受数据输入第一分类模型，通过第一分类模型识别用户在每一次心理辅导训练后的训练结果，并统计所述用户在当前一个训练周期内的训练结果；根据用户在当前一个训练周期内的训练结果确定用户下一个训练周期的训练方案。解决了心理机器人按照预先设定好的辅导流程对用户进行心理辅导，无法满足不同用户的个性化需求，导致辅导结果不理想的问题。在心理机器人对用户进行心理治疗/辅导过程中，在每个时间周期结束后基于用户的练习状况判断用户的练习效果是否达到预期，根据用户的实际状态调整下一阶段的策略，进而令用户在每一阶段的练习都可以达到预期的效果，以使得用户整体的治疗/辅导效果，以及对心理机器人的使用黏性均得以改善。
附图说明
图1是本公开实施例的一种心理辅导训练方案的确定方法的移动终端的硬件结构框图；
图2是根据本公开实施例中一种可选的心理辅导训练方案的确定方法的流程图；
图3为本公开实施例的一种可选的一个完整的时间周期内用户训练结果示意图；
图4是根据本公开实施例的一种可选的心理辅导训练方案的确定装置的结构框图；
图5是根据本公开实施例的一种可选的电子装置结构示意图。
具体实施方式
下文中将参考附图并结合实施例来详细说明本公开。需要说明的是，在不冲突的情况下，本公开中的实施例及实施例中的特征可以相互组合。
需要说明的是，本公开的说明书和权利要求书及上述附图中的术语“第一”、“第二”等是用于区别类似的对象，而不必用于描述特定的顺序或先后次序。
本公开的实施例中所提供的方法实施例可以在移动终端、计算机终端或者类似的运算装置中执行。以运行在移动终端上为例，图1是本公开实施例的一种心理辅导训练方案的确定方法的移动终端的硬件结构框图。如图1所示，移动终端可以包括一个或多个(图1 中仅示出一个)处理器102(处理器102可以包括但不限于微处理器MCU或可编程逻辑器件FPGA等的处理装置)和用于存储数据的存储器104，其中，上述移动终端还可以包括用于通信功能的传输设备106以及输入输出设备108。本领域普通技术人员可以理解，图1所示的结构仅为示意，其并不对上述移动终端的结构造成限定。例如，移动终端还可包括比图1中所示更多或者更少的组件，或者具有与图1所示不同的配置。
存储器104可用于存储计算机程序，例如，应用软件的软件程序以及模块，如本公开实施例中的心理辅导训练方案的确定方法对应的计算机程序，处理器102通过运行存储在存储器104内的计算机程序，从而执行各种功能应用以及数据处理，即实现上述的方法。存储器104可包括高速随机存储器，还可包括非易失性存储器，如一个或者多个磁性存储装置、闪存、或者其他非易失性固态存储器。在一些实例中，存储器104可进一步包括相对于处理器102远程设置的存储器，这些远程存储器可以通过网络连接至移动终端。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。
传输装置106用于经由一个网络接收或者发送数据。上述的网络具体实例可包括移动终端的通信供应商提供的无线网络。在一个实例中，传输装置106包括一个网络适配器(Network Interface Controller，简称为NIC)，其可通过基站与其他网络设备相连从而可与互联网进行通讯。在一个实例中，传输装置106可以为射频(Radio Frequency，简称为RF)模块，其用于通过无线方式与互联网进行通讯。
本公开实施例提供了一种心理辅导训练方案的确定方法。图2是根据本公开实施例中一种可选的心理辅导训练方案的确定方法的流程图，如图2所示，该方法包括：
步骤S202，通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据，其中，训练感受数据用于指示用户进行心理辅导训练后的感受；
步骤S204，将训练感受数据输入第一分类模型，通过第一分类模型识别用户在每一次心理辅导训练后的训练结果，并统计用户在当前一个训练周期内的训练结果，其中，第一分类模型为对初始超长文本分类模型训练得到的模型；
步骤S206，根据用户在当前一个训练周期内的训练结果确定用户下一个训练周期的训练方案。
本公开实施例中方法的执行主体可以是心理机器人，该心理机器人可以通过任意形式的应用程序实现，例如，APP，微信小程序，或搭载在预设终端中的固有程序等方式呈现。具体而言，本公开实施例提供的方法所涉及执行终端可以包括用户终端和服务器，服务器可以为云服务器，也可以为本地服务器。
用户终端，用于搭载心理机器人，并与用户之间进行交互，以实现对用户的心理治疗/辅导。用户设备包括但不限于手机、平板电脑、PC、可穿戴设备、室内大屏终端、户外大屏终端等。
服务器，用于根据用户的输入数据识确定用户所期望的心理治疗/辅导方式或适于用户的心理治疗/辅导方式，以及后续根据选定的方式对用户进行心理治疗/辅导。
需要说明的是，本公开实施例中涉及的训练感受数据，可以包括：用户训练后的想法，用户训练后的身体感受，用户训练后的情绪感受等，可以进一步将训练后的身体感受具体化为：头部感受、肩颈感受、四肢感受、脊柱的感受、腰腹核心部位的感受等，本公开实施例对此不做限定。例如，用户完成训练后所反馈的训练感受数据可以是：
用户想法：太奇妙了，我觉得如此放松。
身体感受：我身体的一些部位如同消失了一样。
情绪感受：这种感觉太美好了，我仿佛没有了重量。
需要说明的是，第一分类模型所采用的可以是基于BERT+LSTM实现的超长文本分类的模型，也可以是基于BERT+LSTM+CRF实现的超长文本分类的模型，只要是可以实现针对自然语言识别与分类的机器学习模型即可，本公开实施例对此不做限定。
一种可选的实施方式中，上述步骤S204可以通过以下步骤实现：
S1，将用户在每一次心理辅导训练后的训练感受数据输入第一分类模型，生成对应训练结果的N1维向量，其中，N1维向量中的每一维向量分别对应一个类别的训练感受数据，每一维向量使用一个数字指示用户的训练感受等级；
S2，在用户在第一训练周期完成了多次心理辅导训练的情况下，将用户在第一训练周期的多次训练结果对应的N1维向量进行同位相加，生成用户在第一训练周期的训练结果对应的N2维向量；
S3，将用户在第二训练周期内的训练结果生成N2*M维向量，其中，M表示每一个第二训练周期内所包含的第一训练周期的个数。
例如，当训练感受数据包括用户训练后的想法，用户训练后的身体感受，用户训练后的情绪感受时，可以在用户完成一次心理辅导训练后生成对应的3维向量(x，y，z)，即N1＝3。其中，x对应用户训练后的想法，y对应用户训练后的身体感受，z对应用户训练后的情绪感受。训练感受等级可以用-1，0，1分别表示感受不好、感受一般、感受很好，例如用户完成一次心理辅导训练后生成对应的3维向量(1，0，1)表示，用户训练后的想法是好的，身体感受一般，情绪感受是好的，整体应该是比较积极的感受。训练感受等级也可以用1，2，3，4，5分别表示感受变好的程度，例如1表示最不好，5表示最好，或者1表示最好，5表示最不好。训练感受等级不限定于上述三种或五种等级，可以根据实际需求进行设置，本公开实施例对此不做限定。
本公开实施例中描述的同位相加，可以理解为：在第一训练周期内完成了多次心理辅导训练时，例如，在同一天(第一训练周期为一天)内，完成了四次心理辅导训练，这四次训练结果对应的3维向量分别是(x1，y1，z1)、(x2，y2，z2)、(x3，y3，z3)和(x4，y4，z4)，可以将这一天内总的训练结果统计为一个3维向量的形式，表示为(x1+x2+x3+x4，y1+y2+y3+y4，z1+z2+z3+z4)。当用户在某一次训练后没有进行结果反馈时，可以将当次的3维向量设置为默认的向量，例如(0，0，0)。
第二训练周期可以理解为一个治疗/辅导的疗程，例如一个星期(七天)。在整个第二训练周期内，如果用户在某一天没有进行训练，或者没有进行训练结果的反馈，可以将当天的3维向量设置为默认的向量，例如(0，0，0)。完成整个第二训练周期的训练后，将第二训练周期内的向量整合到一起，形成3*7维向量，即21维向量。
一种可选的实施方式中，在将用户在每一次心理辅导训练后的训练感受数据输入第一分类模型，生成对应训练结果的N1维向量之前，上述方法还包括：
S1，获取第一样本数据，其中，第一样本数据至少包括以下之一的类别：用于描述训练后想法的数据，用于描述训练后身体感受的数据，用于描述训练后情绪感受的数据；
S2，对每一类别的第一样本数据的训练感受等级分别进行标注；
S3，使用标注后的第一样本数据训练第一分类模型，其中，不同类别的第一样本数据对应训练不同类别的第一分类模型。
需要说明的是，第一分类模型可在训练模型之前或在训练模型的过程中，收集用户对用户想法、身体感受、情绪感受的相关第一样本数据，并且由心理学人员对上述样本数据进行标注，或者由计算机通过识别关键字后对样本数据进行聚类标注，例如，采用1表示倾向为变好，0表示倾向为无变化，-1表示倾向为变差；对于用户未进行反馈的情形，也可以通过0进行标注。
将标注后的第一样本数据输入初始超长文本分类模型中进行训练。需要注意的是，针对每一类别的训练感受数据，分别训练一个分类模型。例如，用于描述训练后想法的第一样本数据，可以训练针对用户训练后想法的分类模型，用于描述训练后身体感受的数据，可以训练针对用户训练后身体感受的分类模型，用于描述训练后情绪感受的数据，可以训练针对用户训练后情绪感受的分类模型。上述多个分类模型的结构可以是一样的，也可以各有不同。
一种可选的实施方式中，将所述用户在第二训练周期内的训练结果生成N2*M维向量之后，所述方法还包括：
统计用户在第二训练周期内完成的训练次数，作为第N2*M+1维向量。
需要说明的是，心理机器人或服务器可以对用户的练习次数进行计数，即统计用户在第二训练周期内的练习频率。在一示例中，用户在为期一周的第二训练周期内，于周一、三、五、日每天完成两次练习，于周二、四、六每天完成一次练习，则在该周结束后，可记录用户共练习有11次，可以将11作为第22维向量。
一种可选的实施方式中，上述步骤S206可以通过以下步骤实现：
S1，通过与用户的交互问询，获取用户在当前第二训练周期完成后的反馈结果；
S2，在用户的反馈结果指示达到预期效果的情况下，根据预设的训练流程确定用户在下一个第二训练周期的训练方案；
S3，在用户的反馈结果指示没有达到预期效果或用户没有明确进行反馈的情况下，将用户在当前第二训练周期内的训练结果输入第二分类模型，通过第二分类模型确定用户在下一个第二训练周期的训练方案，其中，第二分类模型为对初始分类器模型训练得到的模型。
需要说明的是，在每一个第二训练周期(例如，一周)结束后，心理机器人询问用户对当前第二训练周期内所进行练习的总结，在一示例中，如用户表示当前第二训练周期内的练习效果达到预期，如用户表达“我在本周内的训练效果很好，感觉达到了初定的目标”，则按照预设的流程在下一时间周期内进行下一阶段的练习。在另一示例中，用户没有明确表示当前周期内的练习效果达到预期，如用户表达“我在本周内的训练感觉一般”，“我好像并没有达到一开始的目标”等，或者，用户未能对心理机器人的询问进行反馈，则根据统计的当前一个第二训练周期内用户每次练习的练习情况，以及用户的练习次数，通过预先训练的第二分类模型预测适合用户下一阶段的练习方式，进而调整用户在下一阶段的训练策略。
需要说明的是，第二分类模型可以是SVM分类器，SVM分类器的训练、预测速度相比较基于深度学习的分类模型更高效；同时，SVM分类器训练的数据是经过处理后的特征， 不涉及语义信息，可以直接使用第一分类模型输出的特征向量作为输入数据。在本申请实施例中，采用SVM分类器可在实现第二分类模型的功能的前提下，提高第二分类模型预测的效率，同时实现整体模型的轻量化。
一种可选的实施方式中，将用户在当前第二训练周期内的训练结果输入第二分类模型，通过第二分类模型确定用户在下一个第二训练周期的训练方案，可以通过以下步骤实现：
S1，将用户在当前第二训练周期内的统计结果对应的L维向量输入第二分类模型，其中，L维向量包括：用户在当前一个训练周期内的训练结果对应的N2*M维向量，或，用户在当前一个训练周期内的训练结果及训练次数对应的N2*M+1维向量；
S2，通过第二分类模型输出用户的心理倾向结果，其中，心理倾向结果用于指示用户在当前第二训练周期的训练是否达到预期效果；
S3，根据用户的心理倾向结果确定用户在下一个第二训练周期的训练方案。
需要说明的是，输入第二分类模型的特征向量可以是用户在当前一个训练周期内的训练结果对应的N2*M维向量，也可以是包含训练次数的N2*M+1维向量。第二分类模型处理的向量的维度与L维向量的维度是一致的，当输入的数据不包含统计次数时，第二分类模型的处理维度是N2*M维，当输入的数据包含统计次数时，第二分类模型的处理维度是N2*M+1维。例如，当输入第二分类模型的特征向量时21维向量时，对应的第二分类模型是处理21维向量的模型，当输入第二分类模型的特征向量时22维向量时，对应的第二分类模型是处理22维向量的模型。
一种可选的实施方式中，通过第二分类模型确定用户在下一个第二训练周期的训练方案之前，可以通过以下方式训练第二分类模型：
S1，获取由,用户在一个第二训练周期内的统计结果对应的L维向量组成的第二样本数据；
S2，对第二样本数据对应的心理倾向结果进行标注，其中，心理倾向结果至少包括以下之一：满意，不满意；
S3，使用标注后的所第二样本数据训练第二分类模型。
以SVM分类器为例，需进行预先训练，将不同用户或用户在不同训练周期下所得到的21维向量或22维向量作为样本数据，由心理学人员对上述第二样本数据进行标注，或者由计算机使用机器学习算法通过识别关键字对第二样本数据进行标注。标注的方式可通过1表示倾向为感到满意，-1表示感到一般或不满意，也可以用1-5五个等级表示满意程度，1表示最不满意，5表示最满意等，标注的类别越细致，则心理机器人调整的策略与用户的需求之间越贴合，本公开实施例对此不做限定。然后将标注后的样本数据输入至SVM分类器内进行训练，即可得到训练后的SVM分类器。
一种可选的实施方式中，在用户的反馈结果指示达到预期效果的情况下，所述方法还包括：
获取用户在当前第二训练周期内的统计结果对应的目标L维向量；
将目标L维向量对应的心理倾向结果标注为满意；
将标注后的目标L维向量输入第二分类模型进行训练并更新第二分类模型。
在用户明确表示当前训练周期内的练习效果达到预期的情况下，则可将该统计结果对 应的L维向量标注为1，表示满意，将该标注后的L维特征向量输入SVM分类器内进行训练，以对SVM分类器进行更新，进而使得随着用户对心理机器人的不断使用，令SVM分类器不断更新，预测效果更加准确。
下面通过一个具体示例来说明本公开实施例一种可选的心理辅导训练方案的确定方法的实现过程。
S11，在用户每一次完成练习后，心理机器人询问用户的训练感受，询问的过程可以包括以下问题：用户训练后的想法，用户训练后的身体感受，用户训练后的情绪感受。
需要说明的是，上述询问的方式可以是向用户提出问题后，由用户通过文字或语音直接输入，也可以是向用户提供文字/评分等选项，供用户通过点选操作进行反馈。下表为多个用户进行反馈的内容：
Figure PCTCN2021105287-appb-000001
S12，用户对S11中的问题分别进行反馈，心理机器人收到用户的反馈后，通过预先训练的分类模型(相当于前述第一分类模型)对上述用户反馈进行识别，以评估用户本次练习的情况。对应的，在用户每一次完成练习后，心理机器人均需记录用户该次的练习情况，同时还需对用户的练习次数进行计数。
需要说明的是，上述分类模型所采用的是基于BERT+LSTM实现的超长文本分类的模型。
根据前述的标注规则，对上表中多个用户的反馈进行识别后，则有：
Figure PCTCN2021105287-appb-000002
以此，对于上述User001可用(1，1，1)表达其该次练习状况，对于上述User002 可用(-1，0，-1)表达其该次练习状况，对于上述User003可用(0，-1，-1)表达其该次练习状况。
当用户在当前时间周期内继续进行练习过程中，其每一次均练习均可通过上述类似的三维向量表达其练习状况。以某一用户为例，当该用户完成一个完整的时间周期时，其练习状况表达如图3所示。图3为本公开实施例的一种可选的一个完整的时间周期内用户训练结果示意图，如图3所示，该用户在每一天所进行的每一次练习的练习状况均可通过对应的三维向量进行表示，对该用户在第一天的三次练习的练习状况对应的三维向量进行同位累加，即可得到用于表示该用户在第一天的练习状况的新的三维向量，即上图中的(1，1，-1)；以此类推，该用户在一周内的每一天的练习状况，均可将该天内所有练习后记录的练习状况进行同位累加进而确定。以此，即可得到七个三维向量，将其汇总，构成一21维向量，该21维向量即表示用户在该周内的练习状况。
同时，还可进一步对该用户在该周内进行练习的次数进行统计，上图中，用户在该周内共计练习12次，即在上述21维向量后增加一维，表示用户在该周内的练习次数/练习频率。
以此，即可将用户在一周内的练习状况与练习次数以一22维特征向量的方式进行表达，对于用户在任意时间周期，或任意用户而言，均可通过上述方式以记录用户在固定的时间周期内的练习状况与练习次数。
S13，每一个时间周期(例如，一周)结束后，心理机器人询问用户对当前时间周期内所进行练习的总结，在一示例中，如用户表示当前周期内的练习效果达到预期，如用户表达“我在本周内的训练效果很好，感觉达到了初定的目标”，则按照预设的流程在下一时间周期内进行下一阶段的练习。在另一示例中，用户没有明确表示当前周期内的练习效果达到预期，如用户表达“我在本周内的训练感觉一般”，“我好像并没有达到一开始的目标”等，或者，用户未能对心理机器人的询问进行反馈，则根据S12中所确定的用户每次练习的练习情况，以及用户的练习次数，通过预先训练的SVM分类器以预测适合用户下一阶段的练习方式，进而调整用户在下一阶段的策略。
需要说明的是，SVM分类器的训练、预测速度相比较基于深度学习的分类模型更高效；同时，SVM分类器训练的数据是经过处理后的特征，不涉及语义信息，因此效果与基于深度学习的分类模型相当。
上述SVM分类器需进行预先训练，将不同用户或用户在不同时间周期下所得到的S12中的22维向量作为样本数据，由心理学人员对上述样本数据进行标注。标注的方式可通过1表示倾向为感到满意，-1表示感到一般或不满意。将标注后的样本数据输入至SVM分类器内进行训练，即可得到训练后的SVM分类器。
在上述S13中用户没有明确表示当前周期内的练习效果达到预期的情况下，即可将用于表示该用户练习状况与练习次数的22维特征向量输入至SVM分类器中，通过SVM分类器对该特征向量进行识别，以确定该特征向量对应的标签，下表为SVM分类器根据不同用户的练习状况与练习次数进行识别后所确定的标签：
Figure PCTCN2021105287-appb-000003
上表中，User001与User002的识别结果为1，即为满意，故对于User001与User002可按照预设流程进行下一阶段的练习，对于User003的识别结果为-1，即为不满意，故对User003而言，当前时间周期内的练习效果并不理想，故心理机器人可在下一时间周期内，重新以上一阶段的练习方式对用户进行练习，也可采用其它的练习方式对用户进行练习。
同时，在上述S13中用户明确表示当前周期内的练习效果达到预期的情况下，则可将用于表示该用户练习状况与练习次数的22维特征向量赋予标签1，将该标注后的22维特征向量输入SVM分类器内进行训练，以对SVM分类器进行更新，进而使得随着用户对心理机器人的不断使用，令SVM分类器不断更新，预测效果更加准确。
本公开实施例中涉及的心理辅导训练方案可以由心理机器人提供，心理机器人可提供包括正念治疗、助眠治疗、CBT认知疗法治疗、放松治疗等多种治疗方式，以正念治疗为例，当确定正念治疗方式后，心理机器人将在为期8周的疗程内，每周向用户提供不同的正念练习方式，如正念冥想、正念呼吸，正念运动，正念饮食等，以令用户通过正念的方式实现对自身负面情绪的剥离，达到心理治疗的目的。受限于每个用户的性格、生活习惯的独立性以及对其心理造成影响的缘由，其在主观与客观上所适合的治疗方式都是不同的，例如，习惯于独处放松且每天作息规律的用户可在正念治疗中起到极佳的效果，习惯于运动放松且每天缺乏固定的冥想时间的用户则不太适合采用正念治疗；又例如，用户由于职场矛盾、争执等而造成的情绪低落则宜于进行正念治疗，用户由于体力劳动过度导致的心理疲惫则正念治疗效果相对有限。
因此在用户在一个治疗周期内达到的效果不好时，可以根据练习次数、或是某一项训练感受数据不理想，将下一个治疗周期调整为更适合用户的训练习惯的其他治疗方案，或者重复上一治疗周期的训练内容。
通过本公开实施例，可在心理机器人对用户进行心理治疗/辅导过程中，在每个时间周期结束后基于用户的练习状况与练习次数，以及用户在当前时间周期结束时对本阶段练习的总结，综合判断用户的练习效果是否达到预期，以此，在用户未达预期的情形下，并不按照预设流程或预测策略对用户进行心理治疗/辅导，而是根据用户的实际状态调整下一阶段的策略，进而令用户在每一阶段的练习都可以达到预期的效果，以使得用户整体的治疗/辅导效果，以及对心理机器人的使用黏性均得以改善。
本公开实施例中所采用的判断与识别过程中的特征运用方法，不仅单纯的识别出用户的倾向，还可通过同位累加的方式，令用户的某一倾向更加显著，进而增加识别过程中的准确性，并可根据实际需求进一步调整相应的策略。
本公开实施例中，分类器可在用户使用过程中进行循环训练更新，进而令心理机器人随着用户使用人数以及使用时间的增长，可显著提高识别或判断效果。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到根据上述实施例的方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况 下前者是更佳的实施方式。基于这样的理解，本公开的技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质(如ROM/RAM、磁碟、光盘)中，包括若干指令用以使得一台终端设备(可以是手机，计算机，服务器，或者网络设备等)执行本公开各个实施例所述的方法。
根据本公开实施例的另一个方面，还提供了一种用于实施上述心理辅导训练方案的确定方法的心理辅导训练方案的确定装置。该装置用于实现上述实施例及优选实施方式，已经进行过说明的不再赘述。如以下所使用的，术语“模块”可以实现预定功能的软件和/或硬件的组合。尽管以下实施例所描述的装置较佳地以软件来实现，但是硬件，或者软件和硬件的组合的实现也是可能并被构想的。图4是根据本公开实施例的一种可选的心理辅导训练方案的确定装置的结构框图，如图4所示，该装置包括：
获取模块402，用于通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据，其中，所述训练感受数据用于指示所述用户进行心理辅导后的感受；
处理模块404，用于将所述训练感受数据输入第一分类模型，通过所述第一分类模型识别所述用户在每一次所述心理辅导训练后的训练结果，并统计所述用户在当前一个训练周期内的训练结果，其中，所述第一分类模型为根据初始超长文本分类模型训练得到的模型；
确定模块406，用于根据所述用户在当前一个训练周期内的训练结果确定所述用户下一个训练周期的训练方案。
可选的，处理模块404包括：
输入单元，用于将所述用户在每一次心理辅导训练后的训练感受数据输入所述第一分类模型，生成对应所述训练结果的N1维向量，其中，所述N1维向量中的每一维向量分别对应一个类别的训练感受数据，每一维向量使用一个数字指示所述用户的训练感受等级；
第一生成单元，用于在所述用户在第一训练周期完成了多次心理辅导训练的情况下，将所述用户在所述第一训练周期的多次所述训练结果对应的N1维向量进行同位相加，生成所述用户在所述第一训练周期的训练结果对应的N2维向量；
第二生成单元，用于将所述用户在第二训练周期内的训练结果生成N2*M维向量，其中，M表示每一个所述第二训练周期内所包含的所述第一训练周期的个数。
可选的，处理模块404还包括：
第一获取单元，用于获取第一样本数据，其中，所述第一样本数据至少包括以下之一的类别：用于描述训练后想法的数据，用于描述训练后身体感受的数据，用于描述训练后情绪感受的数据；
第一标注单元，用于对每一类别的所述第一样本数据的训练感受等级分别进行标注；
第一训练单元，用于使用标注后的所述第一样本数据训练所述第一分类模型，其中，不同类别的所述第一样本数据对应训练不同类别的所述第一分类模型。
可选的，处理模块404还包括：
统计单元，用于统计所述用户在所述第二训练周期内完成的训练次数，作为第N2*M+1维向量。
可选的，所述确定模块406包括：
第二获取单元，用于通过与所述用户的交互问询，获取所述用户在当前第二训练周期 完成后的反馈结果；
第一确定单元，用于在所述用户的反馈结果指示达到预期效果的情况下，根据预设的训练流程确定所述用户在下一个第二训练周期的训练方案；
第二确定单元，用于在所述用户的反馈结果指示没有达到预期效果或所述用户没有明确进行反馈的情况下，将所述用户在当前第二训练周期内的训练结果输入第二分类模型，通过所述第二分类模型确定所述用户在下一个第二训练周期的训练方案，其中，所述第二分类模型为对初始分类器模型训练得到的模型。
可选的，第二确定单元包括：
输入子单元，用于将所述用户在当前第二训练周期内的统计结果对应的L维向量输入所述第二分类模型，其中，所述L维向量包括：所述用户在当前一个训练周期内的训练结果对应的N2*M维向量，或，所述用户在当前一个训练周期内的训练结果及训练次数对应的N2*M+1维向量；
输出子单元，用于通过所述第二分类模型输出所述用户的心理倾向结果，其中，所述心理倾向结果用于指示所述用户在当前第二训练周期的训练是否达到预期效果；
确定子单元，用于根据所述用户的心理倾向结果确定所述用户在下一个第二训练周期的训练方案。
可选的，所述确定模块406还包括：
第三获取单元，用于获取由所述用户在一个第二训练周期内的统计结果对应的L维向量组成的第二样本数据；
第二标注单元，用于对所述第二样本数据对应的心理倾向结果进行标注，其中，所述心理倾向结果至少包括以下之一：满意，不满意；
第二训练单元，用于使用标注后的所述第二样本数据训练所述第二分类模型。
可选的，所述确定模块406还包括：
第四获取单元，用于获取所述用户在当前第二训练周期内的统计结果对应的目标L维向量；
第三标注单元，用于将所述目标L维向量对应的心理倾向结果标注为满意；
更新单元，用于将标注后的所述目标L维向量输入所述第二分类模型进行训练并更新所述第二分类模型。
根据本公开实施例的又一个方面，还提供了一种用于实施上述心理辅导训练方案的确定方法的电子装置，上述电子装置可以但不限于应用于服务器中。如图5所示，该电子装置包括存储器502和处理器504，该存储器502中存储有计算机程序，该处理器504被设置为通过计算机程序执行上述任一项方法实施例中的步骤。
可选地，在本实施例中，上述电子装置可以位于计算机网络的多个网络设备中的至少一个网络设备。
可选地，在本实施例中，上述处理器可以被设置为通过计算机程序执行以下步骤：
S1，通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据，其中，训练感受数据用于指示用户进行心理辅导后的感受；
S2，将训练感受数据输入第一分类模型，通过第一分类模型识别用户在每一次心理辅导训练后的训练结果，并统计用户在当前一个训练周期内的训练结果，其中，第一分类模 型为对初始超长文本分类模型训练得到的模型；
S3，根据用户在当前一个训练周期内的训练结果确定用户下一个训练周期的训练方案。
可选地，本领域普通技术人员可以理解，图5所示的结构仅为示意，电子装置也可以是智能手机(如Android手机、iOS手机等)、平板电脑、掌上电脑以及移动互联网设备(Mobile Internet Devices，MID)、PAD等终端设备。图5其并不对上述电子装置的结构造成限定。例如，电子装置还可包括比图5中所示更多或者更少的组件(如网络接口等)，或者具有与图5所示不同的配置。
其中，存储器502可用于存储软件程序以及模块，如本公开实施例中的心理辅导训练方案的确定方法和装置对应的程序指令/模块，处理器504通过运行存储在存储器502内的软件程序以及模块，从而执行各种功能应用以及数据处理，即实现上述的心理辅导训练方案的确定方法。存储器502可包括高速随机存储器，还可以包括非易失性存储器，如一个或者多个磁性存储装置、闪存、或者其他非易失性固态存储器。在一些实例中，存储器502可进一步包括相对于处理器504远程设置的存储器，这些远程存储器可以通过网络连接至终端。上述网络的实例包括但不限于互联网、企业内部网、局域网、移动通信网及其组合。其中，存储器502具体可以但不限于用于储存心理辅导训练方案的确定方法的程序步骤。作为一种示例，如图5所示，上述存储器502中可以但不限于包括上述心理辅导训练方案的确定装置中的获取模块402、处理模块404和确定模块406等。此外，还可以包括但不限于上述心理辅导训练方案的确定装置中的其他模块单元，本示例中不再赘述。
可选地，上述的传输装置506用于经由一个网络接收或者发送数据。上述的网络具体实例可包括有线网络及无线网络。在一个实例中，传输装置506包括一个网络适配器(Network Interface Controller，NIC)，其可通过网线与其他网络设备与路由器相连从而可与互联网或局域网进行通讯。在一个实例中，传输装置506为射频(Radio Frequency，RF)模块，其用于通过无线方式与互联网进行通讯。
此外，上述电子装置还包括：显示器508，用于显示可疑帐号的告警推送；和连接总线510，用于连接上述电子装置中的各个模块部件。
本公开的实施例还提供了一种计算机可读的存储介质，该存储介质中存储有计算机程序，其中，该计算机程序被设置为运行时执行上述任一项方法实施例中的步骤。
可选地，在本实施例中，上述存储介质可以被设置为存储用于执行以下步骤的计算机程序：
S1，通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据，其中，训练感受数据用于指示用户进行心理辅导后的感受；
S2，将训练感受数据输入第一分类模型，通过第一分类模型识别用户在每一次心理辅导训练后的训练结果，并统计用户在当前一个训练周期内的训练结果，其中，第一分类模型为对初始超长文本分类模型训练得到的模型；
S3，根据用户在当前一个训练周期内的训练结果确定用户下一个训练周期的训练方案。
可选地，存储介质还被设置为存储用于执行上述实施例中的方法中所包括的步骤的计算机程序，本实施例中对此不再赘述。
可选地，在本实施例中，本领域普通技术人员可以理解上述实施例的各种方法中的全部或部分步骤是可以通过程序来指令终端设备相关的硬件来完成，该程序可以存储于一计算机可读存储介质中，存储介质可以包括：闪存盘、只读存储器(Read-Only Memory，ROM)、随机存取器(Random Access Memory，RAM)、磁盘或光盘等。
上述本公开实施例序号仅仅为了描述，不代表实施例的优劣。
上述实施例中的集成的单元如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在上述计算机可读取的存储介质中。基于这样的理解，本公开的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的全部或部分可以以软件产品的形式体现出来，该计算机软件产品存储在存储介质中，包括若干指令用以使得一台或多台计算机设备(可为个人计算机、服务器或者网络设备等)执行本公开各个实施例所述方法的全部或部分步骤。
在本公开的上述实施例中，对各个实施例的描述都各有侧重，某个实施例中没有详述的部分，可以参见其他实施例的相关描述。
在本公开所提供的几个实施例中，应该理解到，所揭露的用户端，可通过其它的方式实现。其中，以上所描述的装置实施例仅仅是示意性的，例如所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，单元或模块的间接耦合或通信连接，可以是电性或其它的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本公开各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。上述集成的单元既可以采用硬件的形式实现，也可以采用软件功能单元的形式实现。
以上所述仅是本公开的优选实施方式，应当指出，对于本技术领域的普通技术人员来说，在不脱离本公开原理的前提下，还可以做出若干改进和润饰，这些改进和润饰也应视为本公开的保护范围。

Claims (12)

  1. 一种心理辅导训练方案的确定方法，其特征在于，包括：
    通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据，其中，所述训练感受数据用于指示所述用户进行心理辅导训练后的感受；
    将所述训练感受数据输入第一分类模型，通过所述第一分类模型识别所述用户在每一次所述心理辅导训练后的训练结果，并统计所述用户在当前一个训练周期内的训练结果，其中，所述第一分类模型为对初始超长文本分类模型训练后得到的模型；
    根据所述用户在当前一个训练周期内的训练结果确定所述用户下一个训练周期的训练方案；
    其中，将所述训练感受数据输入第一分类模型，通过所述第一分类模型识别所述用户在每一次所述心理辅导训练后的训练结果，并统计所述用户在当前一个训练周期内的训练结果包括：
    将所述用户在每一次心理辅导训练后的训练感受数据输入所述第一分类模型，生成对应所述训练结果的N1维向量，其中，所述N1维向量中的每一维向量分别对应一个类别的训练感受数据，每一维向量使用一个数字指示所述用户的训练感受等级；
    在所述用户在第一训练周期完成了多次心理辅导训练的情况下，将所述用户在所述第一训练周期的多次所述训练结果对应的N1维向量进行同位相加，生成所述用户在所述第一训练周期的训练结果对应的N2维向量；
    将所述用户在第二训练周期内的训练结果生成N2*M维向量，其中，M表示每一个所述第二训练周期内所包含的所述第一训练周期的个数。
  2. 根据权利要求1所述的方法，其特征在于，所述将所述用户在每一次心理辅导训练后的训练感受数据输入所述第一分类模型，生成对应所述训练结果的N1维向量之前，所述方法还包括：
    获取第一样本数据，其中，所述第一样本数据至少包括以下之一的类别：用于描述训练后想法的数据，用于描述训练后身体感受的数据，用于描述训练后情绪感受的数据；
    对每一类别的所述第一样本数据的训练感受等级分别进行标注；
    使用标注后的所述第一样本数据训练所述第一分类模型，其中，不同类别的所述第一样本数据对应训练不同类别的所述第一分类模型。
  3. 根据权利要求1所述的方法，其特征在于，将所述用户在第二训练周期内的训练结果生成N2*M维向量之后，所述方法还包括：
    统计所述用户在所述第二训练周期内完成的训练次数，作为第N2*M+1维向量。
  4. 根据权利要求3所述的方法，其特征在于，所述根据所述用户在当前一个训练周期内的训练结果确定所述用户在下一个训练周期的训练方案包括：
    通过与所述用户的交互问询，获取所述用户在当前第二训练周期完成后的反馈结果；
    在所述用户的反馈结果指示达到预期效果的情况下，根据预设的训练流程确定所述用户在下一个第二训练周期的训练方案；
    在所述用户的反馈结果指示没有达到预期效果或所述用户没有明确进行反馈的情 况下，将所述用户在当前第二训练周期内的训练结果输入第二分类模型，通过所述第二分类模型确定所述用户在下一个第二训练周期的训练方案，其中，所述第二分类模型为对初始分类器模型训练得到的模型。
  5. 根据权利要求4所述的方法，其特征在于，所述通过所述第二分类模型确定所述用户在下一个第二训练周期的训练方案包括：
    将所述用户在当前第二训练周期内的统计结果对应的L维向量输入所述第二分类模型，其中，所述L维向量包括：所述用户在当前一个训练周期内的训练结果对应的N2*M维向量，或，所述用户在当前一个训练周期内的训练结果及训练次数对应的N2*M+1维向量；
    通过所述第二分类模型输出所述用户的心理倾向结果，其中，所述心理倾向结果用于指示所述用户在当前第二训练周期的训练是否达到预期效果；
    根据所述用户的心理倾向结果确定所述用户在下一个第二训练周期的训练方案。
  6. 根据权利要求4或5所述的方法，其特征在于，通过所述第二分类模型确定所述用户在下一个第二训练周期的训练方案之前，所述方法还包括：
    获取由所述用户在一个第二训练周期内的统计结果对应的L维向量组成的第二样本数据；
    对所述第二样本数据对应的心理倾向结果进行标注，其中，所述心理倾向结果至少包括以下之一：满意，不满意；
    使用标注后的所述第二样本数据训练所述第二分类模型。
  7. 根据权利要求6所述的方法，其特征在于，在所述用户的反馈结果指示达到预期效果的情况下，所述方法还包括：
    获取所述用户在当前第二训练周期内的统计结果对应的目标L维向量；
    将所述目标L维向量对应的心理倾向结果标注为满意；
    将标注后的所述目标L维向量输入所述第二分类模型进行训练并更新所述第二分类模型。
  8. 根据权利要求1所述的方法，其特征在于，所述通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据包括：
    在用户每一次心理辅导训练后，向所述用户发送询问；
    接收所述用户反馈的所述训练感受数据，所述训练感受数据与所述询问对应。
  9. 根据权利要求8所述的方法，其特征在于，所述询问至少包括以下之一的问题：用户训练后的想法、用户训练后的身体感受、用户训练后的情绪感受。
  10. 一种心理辅导训练方案的确定装置，其特征在于，包括：
    获取模块，用于通过与用户的交互问询，获取用户在每一次心理辅导训练后的训练感受数据，其中，所述训练感受数据用于指示所述用户进行心理辅导训练后的感受；
    处理模块，用于将所述训练感受数据输入第一分类模型，通过所述第一分类模型识别所述用户在每一次所述心理辅导训练后的训练结果，并统计所述用户在当前一个训练周期内的训练结果，其中，所述第一分类模型为根据初始超长文本分类模型训练得到的模型；
    确定模块，用于根据所述用户在当前一个训练周期内的训练结果确定所述用户下 一个训练周期的训练方案；
    所述处理模块还用于：
    将所述用户在每一次心理辅导训练后的训练感受数据输入所述第一分类模型，生成对应所述训练结果的N1维向量，其中，所述N1维向量中的每一维向量分别对应一个类别的训练感受数据，每一维向量使用一个数字指示所述用户的训练感受等级；
    在所述用户在第一训练周期完成了多次心理辅导训练的情况下，将所述用户在所述第一训练周期的多次所述训练结果对应的N1维向量进行同位相加，生成所述用户在所述第一训练周期的训练结果对应的N2维向量；
    将所述用户在第二训练周期内的训练结果生成N2*M维向量，其中，M表示每一个所述第二训练周期内所包含的所述第一训练周期的个数。
  11. 一种计算机可读的存储介质，其特征在于，所述存储介质中存储有计算机程序，其中，所述计算机程序被设置为运行时执行所述权利要求1至9任一项中所述的方法。
  12. 一种电子装置，包括存储器和处理器，其特征在于，所述存储器中存储有计算机程序，所述处理器被设置为运行所述计算机程序以执行所述权利要求1至9任一项中所述的方法。
PCT/CN2021/105287 2021-02-02 2021-07-08 心理辅导训练方案的确定方法及装置 WO2022166110A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
US17/594,499 US11682482B2 (en) 2021-02-02 2021-07-08 Method and apparatus for determining psychological counseling training scheme
EP21773443.3A EP4064289A4 (en) 2021-02-02 2021-07-08 METHOD AND DEVICE FOR DETERMINING A PSYCHOLOGICAL COUNSELING TRAINING SCHEME

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110141410.8A CN112466434B (zh) 2021-02-02 2021-02-02 心理辅导训练方案的确定方法及装置
CN202110141410.8 2021-02-02

Publications (1)

Publication Number Publication Date
WO2022166110A1 true WO2022166110A1 (zh) 2022-08-11

Family

ID=74802482

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/105287 WO2022166110A1 (zh) 2021-02-02 2021-07-08 心理辅导训练方案的确定方法及装置

Country Status (4)

Country Link
US (1) US11682482B2 (zh)
EP (1) EP4064289A4 (zh)
CN (1) CN112466434B (zh)
WO (1) WO2022166110A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112466434B (zh) * 2021-02-02 2021-05-11 南京硅基智能科技有限公司 心理辅导训练方案的确定方法及装置
CN114944228B (zh) * 2022-04-26 2024-04-19 福建福寿康宁科技有限公司 围绕全人健康的干预措施循证与决策辅助方法

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111180043A (zh) * 2019-12-17 2020-05-19 河南水滴智能技术有限公司 一种心理建模引导的患者康复训练方法
CN111370113A (zh) * 2020-03-03 2020-07-03 淮安信息职业技术学院 一种基于物联云的远程心理辅导系统及方法
US20200342348A1 (en) * 2019-04-26 2020-10-29 Fmr Llc Automated predictive analysis and modification of user interaction features using multiple classification models
CN112086169A (zh) * 2020-09-19 2020-12-15 北京心灵力量科技有限公司 采用心理数据标签化建模的交互式心理疏导系统
CN112270973A (zh) * 2020-11-26 2021-01-26 浙江连信科技有限公司 数字智能化心理健康训练方法及装置
CN112466434A (zh) * 2021-02-02 2021-03-09 南京硅基智能科技有限公司 心理辅导训练方案的确定方法及装置

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170004260A1 (en) * 2012-08-16 2017-01-05 Ginger.io, Inc. Method for providing health therapeutic interventions to a user
WO2016071659A1 (en) * 2014-11-06 2016-05-12 Ieso Digital Health Limited Analysing text-based messages sent between patients and therapists
CN109564783A (zh) * 2017-05-11 2019-04-02 微软技术许可有限责任公司 在自动聊天中辅助心理治疗
US20190385711A1 (en) * 2018-06-19 2019-12-19 Ellipsis Health, Inc. Systems and methods for mental health assessment
US20200152330A1 (en) * 2018-11-13 2020-05-14 CurieAI, Inc. Scalable Personalized Treatment Recommendation
CN111666396B (zh) 2020-06-05 2023-10-31 北京百度网讯科技有限公司 用户意图理解满意度评估方法、装置、设备和存储介质
CN111797320B (zh) * 2020-07-02 2023-06-06 中国联合网络通信集团有限公司 数据处理方法、装置、设备及存储介质

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20200342348A1 (en) * 2019-04-26 2020-10-29 Fmr Llc Automated predictive analysis and modification of user interaction features using multiple classification models
CN111180043A (zh) * 2019-12-17 2020-05-19 河南水滴智能技术有限公司 一种心理建模引导的患者康复训练方法
CN111370113A (zh) * 2020-03-03 2020-07-03 淮安信息职业技术学院 一种基于物联云的远程心理辅导系统及方法
CN112086169A (zh) * 2020-09-19 2020-12-15 北京心灵力量科技有限公司 采用心理数据标签化建模的交互式心理疏导系统
CN112270973A (zh) * 2020-11-26 2021-01-26 浙江连信科技有限公司 数字智能化心理健康训练方法及装置
CN112466434A (zh) * 2021-02-02 2021-03-09 南京硅基智能科技有限公司 心理辅导训练方案的确定方法及装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP4064289A4

Also Published As

Publication number Publication date
US11682482B2 (en) 2023-06-20
EP4064289A1 (en) 2022-09-28
US20220277833A1 (en) 2022-09-01
CN112466434A (zh) 2021-03-09
CN112466434B (zh) 2021-05-11
EP4064289A4 (en) 2022-12-28

Similar Documents

Publication Publication Date Title
US11327556B2 (en) Information processing system, client terminal, information processing method, and recording medium
WO2022166110A1 (zh) 心理辅导训练方案的确定方法及装置
WO2018171223A1 (zh) 一种看护机器人的数据处理方法及装置
Tsiourti et al. A virtual assistive companion for older adults: design implications for a real-world application
AU2007342471B2 (en) Situated simulation for training, education, and therapy
US20210248656A1 (en) Method and system for an interface for personalization or recommendation of products
CN109074117A (zh) 用于提升记忆力与作决策的以个人情绪为基础的计算机可读取认知感觉内存储器及认知洞察力
CN110531849A (zh) 一种基于5g通信的增强现实的智能教学系统
CN108388343B (zh) 基于专注度的脑电特征反馈对vr沉浸教学管理系统
US20180375807A1 (en) Virtual assistant system enhancement
CN112289434A (zh) 基于vr的医疗培训模拟方法、装置、设备及存储介质
CN109876264A (zh) 一种神经发育障碍患者感统辅助训练方法及系统
CN112270973A (zh) 数字智能化心理健康训练方法及装置
CN112382151B (zh) 一种线上学习方法及装置、电子设备及存储介质
WO2021036954A1 (zh) 一种智能语音播放方法及设备
WO2022011448A1 (en) Method and system for an interface for personalization or recommendation of products
TW202223804A (zh) 電子資源推送方法及系統
CN109919816A (zh) 基于数据分析的排课方法及装置、存储介质及电子设备
CN113641856A (zh) 用于输出信息的方法和装置
Lim et al. Artificial intelligence concepts for mental health application development: Therapily for mental health care
Berrezueta-Guzman et al. Embedded system as a third version of a didactic transmitter of needs that provides a way of communication to children with cerebral palsy of the spastic type
Langdon et al. Designing studies for the requirements and modelling of users for an accessible set-top box
US20230237926A1 (en) Cognitive Training Using Voice Command
US20230142612A1 (en) Method and apparatus for managing tasks to promote health and wellness
Klein et al. Current trends in Gerontechnology: Recent findings from the Austrian-German-Swiss chapter

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2021773443

Country of ref document: EP

Effective date: 20210930

NENP Non-entry into the national phase

Ref country code: DE