Psychological state assessment method and device based on artificial intelligence algorithm model

Abstract

The invention discloses a psychological state assessment method and device based on an artificial intelligence algorithm model, and belongs to the technical field of psychological state assessment, and the method comprises the following steps: a tester inputs text information into an emotional feature recognition algorithm model; the emotion feature recognition algorithm model analyzes the text information of the testee and extracts the related features of the psychological state; matching the obtained psychological state related features with a psychological state feature library to obtain a matching result; generating a personalized evaluation question group according to a matching result; and after the user completes the personalized assessment question group, obtaining a psychological state assessment result. The emotion and psychological reaction of the tested person in the natural conversation are captured in real time through man-machine conversation, and inaccuracy caused by a traditional scale is avoided. The natural dialogue mode can reflect the internal state of the testee more truly. The AI algorithm can perform real-time analysis through dialogues, and dynamically adjust the evaluation content according to the response of the testee, thereby reducing the emotion influence, and enabling the evaluation result to be more comprehensive and objective.

Description

Technical Field

[0001] The present invention relates to the technical field of psychological state assessment, and in particular to a psychological state assessment method and device based on an artificial intelligence algorithm model. Background Art

[0002] The mental state scale is an important tool in the field of mental health. Commonly used psychological scales include: the State Self-Rating Scale (SCL-90), the Beck Depression Inventory (BDI), the Self-Rating Anxiety Scale (SAS), and the Healthy Lifestyle Scale. Most mental state assessment scales are based on large sample psychological data. After a standardized research and development process, they can provide highly comparable assessment results for different groups of people.

[0003] However, the scales currently widely used for psychological status assessment have some problems in practice, which may affect the accuracy and reliability of the assessment, such as: 1) Test fatigue: Long tests cause test takers to fatigue, affecting the quality of their answers to questions; 2) Social desirability bias: test takers will answer according to the answers they think society or assessors expect, rather than honestly reflecting their actual feelings or status; 3) Self-image maintenance; 4) Influence of emotional fluctuations: The subjects' immediate emotions when filling out the scale may also affect their answers; 5) Avoidance of negative experiences; 6) Deliberate exaggeration or minimization of the status; and so on.

[0004] In order to reduce the impact of the above-mentioned problems, the present invention utilizes artificial intelligence technology to optimize the assessment of psychological state, provides a more personalized and accurate assessment, and introduces tools for monitoring dynamic changes in psychology.

[0005] In the current assessment of mental status, paper-based forms or electronic versions of paper-based forms (such as spreadsheets or simple electronic forms) are still the mainstream assessment tools; 1) Insufficient evaluation capabilities for personalization and dynamic changes Paper or simple electronic scales are usually designed as standardized questionnaires, which are suitable for the general public, but lack personalized design; they cannot track changes in individual psychological states in real time, and cannot provide long-term, continuous assessment data. Psychological states are characterized by volatility and dynamic changes, and the results of a single assessment are often insufficient to reflect the true psychological state of the subject.

[0006] 2) Lack of intelligence and data integration capabilities Paper and simple electronic scales rely on traditional manual scoring and analysis, and it is difficult to use modern technologies such as artificial intelligence and machine learning for intelligent data analysis and pattern recognition. This makes them incapable of processing large-scale and diverse psychological data. It is impossible to integrate other types of data (such as psychological data, behavioral data, etc.) for comprehensive analysis. This single data source limits the comprehensiveness and accuracy of psychological assessments.

[0007] 3) Subjectivity issues Paper-based scales and simple electronic forms rely on self-reports from the test takers. Affected by emotions, motivations, and cognitive abilities, the test takers’ answers may not be accurate or true enough, leading to biased assessment results. This problem is particularly evident in static assessment tools that lack intelligent feedback and supervision. Test takers may give different answers to the same question at different times and in different situations. Traditional paper or simple electronic scales cannot identify these inconsistencies, nor can they provide intervention or corrections in real time.

[0008] 4) The evaluation is not timely and interactive enough Paper-based and simple electronic scales usually do not provide immediate feedback or results, and participants need to wait until the assessment is completed before receiving the analysis results. For mental health issues that require immediate intervention or continuous monitoring, this lag may delay the timely treatment of the problem.

[0009] These scales lack interaction with the test takers, cannot dynamically adjust the difficulty or number of questions based on the test takers' responses, and cannot provide a personalized assessment experience based on individual differences. This limits the application of scales in complex psychological assessments. Summary of the invention

[0010] The purpose of the present invention is to provide a method and device for evaluating the psychological state of an artificial intelligence algorithm model to overcome the deficiencies in the prior art.

[0011] To achieve the above object, the present invention provides the following technical solutions: This application discloses a method for evaluating a psychological state based on an artificial intelligence algorithm model, comprising the following steps: S1. The tester inputs text information into the emotion feature recognition algorithm model; S2, the emotion feature recognition algorithm model analyzes the text information of the tester and extracts the relevant features of the psychological state; S3, matching the psychological state related features obtained in step S2 with the psychological state feature library to obtain a matching result; generating a personalized assessment question set according to the matching result; S4. After the user completes the personalized assessment questions in step S3, the user obtains the psychological state assessment result.

[0012] Preferably, in step S1, the tester inputs the content that makes him feel troubled or uncomfortable as text information through text or voice.

[0013] Preferably, step S2 specifically includes the following sub-steps: S21, standardize the text information to obtain preprocessed text; input the preprocessed text into the GPT model to generate a high-dimensional semantic embedding vector; capture the semantic information and context dependency in the preprocessed text through the high-dimensional semantic embedding vector, and retain the semantic features related to emotions; S22, combining the emotion-related semantic features and emotion vocabulary obtained in step S21, identifying the words, phrases and their contextual relationships closely related to the emotion in the preprocessed text, and extracting emotion-related features; S23, inputting the emotion-related features and high-dimensional semantic embedding vector obtained in step S22 into a multi-level emotion classifier to extract psychological state-related features; the multi-level emotion classifier is used to classify emotions in the text.

[0014] Preferably, in step S21, the standardization process includes removing irrelevant characters and stop words.

[0015] Preferably, in step S23, the multi-level sentiment classifier adopts a deep learning model, including a classifier of a Transformer architecture.

[0016] Preferably, the psychological state feature library in step S3 is established as follows: S31. Based on the Self-Assessment Scale for Psychological States, according to the items describing specific psychological states contained in each dimension of the Self-Assessment Scale for Psychological States, the items are mapped into feature vectors of emotions and psychological states to form a specific state description under each dimension; S32. Map the specific state description under each dimension with the corresponding emotion category.

[0017] Preferably, in step S4, the evaluation result is outputted based on the personalized evaluation question set completed by the user in combination with the evaluation method of the psychological state self-evaluation scale.

[0018] The present invention also discloses a psychological state assessment device based on an artificial intelligence algorithm model, comprising a memory and one or more processors, wherein the memory stores executable code, and when the one or more processors execute the executable code, it is used for the above-mentioned psychological state assessment method based on the artificial intelligence algorithm model.

[0019] The present invention also discloses a computer-readable storage medium on which a program is stored. When the program is executed by a processor, the above-mentioned psychological state assessment method based on an artificial intelligence algorithm model is implemented.

[0020] Beneficial effects of the present invention: 1) Reduce subjectivity and reporting bias: Through human-computer dialogue, the emotions and psychological reactions of the subjects in natural conversations are captured in real time, avoiding the inaccuracies caused by self-report bias, social desirability effect and other problems in traditional scales. This natural dialogue method can more truly reflect the inner state of the subjects. The AI ​​algorithm can conduct real-time analysis through dialogue, dynamically adjust the evaluation content according to the reactions of the subjects, reduce the emotional impact of a single time point, and make the evaluation results more comprehensive and objective.

[0021] 2) Improve assessment efficiency: The system can automatically analyze the psychological state in the conversation, eliminating the steps of traditional scales that require the subject to answer each question, fill in and evaluate manually. The AI ​​model can predict the state in a short time and quickly match the appropriate psychological assessment questions, thereby improving the speed and efficiency of the assessment. After AI detects potential problems, it can quickly guide the subject to conduct more detailed psychological tests or recommend the next intervention measures, reducing manual operations and improving the efficiency of clinical practice.

[0022] 3) Personalized and dynamic assessment: Based on AI's analysis of the initial status, the system can dynamically adjust the type and order of questions according to the individual's specific performance. Compared with the standardized design of traditional scales, this invention can provide each subject with a personalized assessment path to capture more targeted psychological problems.

[0023] The features and advantages of the present invention will be described in detail through embodiments in conjunction with the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a flow chart of a method for evaluating a psychological state based on an artificial intelligence algorithm model of the present invention; Figure 2 is a schematic diagram of the efficiency evaluation of the present invention and conventional tests; Figure 3 is a schematic diagram of the accuracy evaluation of the present invention and the traditional state scale; Figure 4 This is a structural diagram of a psychological state assessment device based on an artificial intelligence algorithm model of the present invention. DETAILED DESCRIPTION In order to make the purpose, technical scheme and advantages of the present invention clearer, the present invention is further described in detail below through the accompanying drawings and embodiments. However, it should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the scope of the present invention. In addition, in the following description, the description of known structures and technologies is omitted to avoid unnecessary confusion of the concept of the present invention.

[0025] See also Figure 1 The present invention provides a method for evaluating a psychological state based on an artificial intelligence algorithm model, and the specific operations are as follows: At the beginning of the test, users can enter content that has recently troubled or made them feel uncomfortable through text or voice as the beginning of the conversation; the system's built-in emotional feature recognition algorithm will automatically analyze the user's input and extract features related to the psychological state. Subsequently, these features will be matched with the psychological state feature library, and the system will generate a personalized assessment question set for the user based on the matching results. After the user completes these personalized test questions in the dialog box, the system will generate the psychological state assessment results in real time to achieve an accurate assessment of the psychological state.

[0026] About the emotional feature recognition algorithm model construction process: This emotion feature recognition algorithm model combines the GPT4.0 large language model and the sentiment classifier to accurately and efficiently identify the emotion features in the input text. The specific steps are as follows: (1) Text preprocessing and embedding generation: The tester’s input text is standardized, including the removal of irrelevant characters and stop words. The processed text is input into the GPT4.0 model to generate high-dimensional semantic embedding vectors. These embedding vectors can capture the deep semantic information and contextual dependencies in the text, thereby retaining emotion-related language features.

[0027] (2) Emotional feature extraction: Based on the generated text embedding, emotion-related features are further extracted. Combining semantic features and emotional vocabulary, the words and phrases in the text that are closely related to emotions and their contextual relationships are identified.

[0028] (3) Multi-level sentiment classifier: The extracted sentiment features and language embedding vectors are input into a multi-level sentiment classifier. This classifier uses a deep learning model (a classifier with a Transformer architecture) and is able to classify the sentiment in the text after training. This classifier can not only distinguish the basic sentiment categories in the text (such as positive, negative, and neutral), but also identify compound emotions (such as the coexistence of anger and disappointment) through a multi-label learning mechanism.

[0029] (4) Model optimization and fine-tuning: During the construction process, the sentiment classifier is trained on a large number of annotated datasets (such as the public GoEmotions sentiment dataset and self-built psychological feature datasets).

[0030] About the establishment of psychological state feature library: The Self-rating Psychological Status Scale (SCL-90) is a widely used tool for assessing an individual's psychological status, covering nine main dimensions. Based on this scale, a psychological status feature library was established. The specific process is as follows: (1) State dimension definition: Each dimension contains several items that describe a specific psychological state. These items are mapped into feature vectors of emotions and psychological states to form a specific state description under each dimension. (2) Feature labeling and classification: The psychological states of the nine dimensions are managed with the corresponding emotional features. The state of each dimension is mapped with the corresponding emotion category.

[0031] About the automatic generation process of personalized assessment questions: After the tester completes the information input, the emotion feature recognition algorithm model extracts the emotion features. The extracted features are matched with the psychological state feature library to obtain the psychological state features that appear during the tester's communication process (such as the model locates the two features of interpersonal relationship sensitivity and depression). The system will extract the questions related to the two feature dimensions of interpersonal relationship sensitivity and depression in the Self-Assessment Scale for Psychological State (SCL-90) and give them to the tester for evaluation. The evaluation results are output with reference to the evaluation method of the Self-Assessment Scale for Psychological State (SCL90).

[0032] The present invention discloses a psychological state assessment method based on an artificial intelligence algorithm model. Based on the artificial intelligence algorithm model, the psychological state reflected in the dialogue of the subject is captured through human-computer dialogue, and the psychological state of the subject is pre-judged. Then, based on the state analyzed by the algorithm, the psychological assessment test questions are matched to assess and confirm the psychological state of the subject.

[0033] The implementation of the present invention can solve the problem of insufficient personalization and dynamic evaluation ability of current paper-based scales or simple electronic scales: the algorithm model automatically matches the psychological test questions related to the subject by capturing the key psychological state information in the text or language communication of the subject; and the system will generate dynamic tracking data for the test results of the subject to achieve long-term dynamic state evaluation and management of the subject. 2) Lack of intelligence and data integration capabilities: the present invention is based on an artificial intelligence algorithm model, which can realize real-time calculation and analysis of evaluation content, and has an external interface, which can be connected and integrated with data such as electroencephalogram equipment to achieve a comprehensive evaluation of the psychological state. 3) Subjectivity problem: the algorithm model of the present invention can communicate with the subject in text or voice to mine the psychological state of the subject, solving the problem that the accuracy of the results of conventional paper-based evaluation scales is affected by the subject's subjectivity. 4) The problem of insufficient immediacy and interactivity of the evaluation: the evaluation system discovered in this invention can be integrated in WeChat or used in the form of a separate app, and the subject can log in and use it at any time. At the same time, the system can automatically reply in real time based on the input content of the subject, meeting the subject's needs for immediacy and interactivity in psychological state evaluation.

[0034] See also Figure 2From the test efficiency evaluation, it can be seen that, taking SCL90 as an example, the average time required to complete the test in this system is 7.8 minutes, which is about 1 / 3 of the time required for conventional testing.

[0035] See also Figure 3 , system test accuracy evaluation, compared with traditional status scale screening, the overall accuracy of the screening results of this invention is improved by 20%.

[0036] An embodiment of a mental state assessment device based on an artificial intelligence algorithm model of the present invention can be applied to any device with data processing capabilities, and the device with data processing capabilities can be a device or apparatus such as a computer. The device embodiment can be implemented through software, or through hardware or a combination of software and hardware. Taking software implementation as an example, as a device in a logical sense, it is formed by the processor of any device with data processing capabilities in which it is located reading the corresponding computer program instructions in the non-volatile memory into the internal memory for execution. From a hardware perspective, if Figure 4 As shown in the figure, it is a hardware structure diagram of any device with data processing capability where a psychological state assessment device based on an artificial intelligence algorithm model of the present invention is located, except Figure 4 In addition to the processor, memory, network interface, and non-volatile memory shown, any device with data processing capabilities in the embodiment of the device may also include other hardware according to the actual function of the device with data processing capabilities, which will not be described in detail. The implementation process of the functions and effects of each unit in the above device is specifically detailed in the implementation process of the corresponding steps in the above method, which will not be described in detail here.

[0037] For the device embodiment, since it basically corresponds to the method embodiment, the relevant parts can refer to the partial description of the method embodiment. The device embodiment described above is only schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of the present invention. Ordinary technicians in this field can understand and implement it without paying creative work.

[0038] An embodiment of the present invention further provides a computer-readable storage medium having a program stored thereon. When the program is executed by a processor, a psychological state assessment device based on an artificial intelligence algorithm model in the above embodiment is implemented.

[0039] The computer-readable storage medium may be an internal storage unit of any device with data processing capability described in any of the aforementioned embodiments, such as a hard disk or a memory. The computer-readable storage medium may also be an external storage device of any device with data processing capability, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), an SD card, a flash card (Flash Card), etc. equipped on the device. Furthermore, the computer-readable storage medium may also include both an internal storage unit and an external storage device of any device with data processing capability. The computer-readable storage medium is used to store the computer program and other programs and data required by any device with data processing capability, and may also be used to temporarily store data that has been output or is to be output.

[0040] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modification, equivalent substitution or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A psychological state assessment method based on an artificial intelligence algorithm model, characterized in that: The steps include: S1. The tester inputs text information into the emotion feature recognition algorithm model; S2, the emotion feature recognition algorithm model analyzes the text information of the tester and extracts the relevant features of the psychological state; S3, matching the psychological state related features obtained in step S2 with the psychological state feature library to obtain a matching result; generating a personalized assessment question set according to the matching result; S4. After the user completes the personalized assessment questions in step S3, the user obtains the psychological state assessment result.

2. A method for evaluating a psychological state based on an artificial intelligence algorithm model as claimed in claim 1, characterized in that: In step S1, the tester inputs the content that makes him feel troubled or uncomfortable as text information through text or voice.

3. A psychological state assessment method based on an artificial intelligence algorithm model as claimed in claim 1, characterized in that: Step S2 specifically includes the following sub-steps: S21, standardize the text information to obtain preprocessed text; input the preprocessed text into the GPT model to generate a high-dimensional semantic embedding vector; capture the semantic information and context dependency in the preprocessed text through the high-dimensional semantic embedding vector, and retain the semantic features related to emotions; S22, combining the emotion-related semantic features and emotion vocabulary obtained in step S21, identifying the words, phrases and their contextual relationships closely related to the emotion in the preprocessed text, and extracting emotion-related features; S23, inputting the emotion-related features and high-dimensional semantic embedding vector obtained in step S22 into a multi-level emotion classifier to extract psychological state-related features; the multi-level emotion classifier is used to classify emotions in the text.

4. A method for evaluating a psychological state based on an artificial intelligence algorithm model as claimed in claim 1, characterized in that: In step S21, the standardization process includes removing irrelevant characters and stop words.

5. A method for evaluating a psychological state based on an artificial intelligence algorithm model as claimed in claim 1, characterized in that: In step S23, the multi-level sentiment classifier adopts a deep learning model, including a classifier of a Transformer architecture.

6. A method for evaluating a psychological state based on an artificial intelligence algorithm model as claimed in claim 1, characterized in that: The psychological state feature library in step S3 is established as follows: S31. Based on the Self-Assessment Scale for Psychological States, according to the items describing specific psychological states contained in each dimension of the Self-Assessment Scale for Psychological States, the items are mapped into feature vectors of emotions and psychological states to form a specific state description under each dimension; S32. Map the specific state description under each dimension with the corresponding emotion category.

7. A method for evaluating a psychological state based on an artificial intelligence algorithm model as claimed in claim 1, characterized in that: In step S4, the evaluation result is outputted based on the personalized evaluation question set completed by the user and the evaluation method of the psychological state self-evaluation scale.

8. A psychological state assessment device based on an artificial intelligence algorithm model, characterized in that: It comprises a memory and one or more processors, wherein the memory stores executable code, and when the one or more processors execute the executable code, it is used to implement a psychological state assessment method based on an artificial intelligence algorithm model as described in any one of claims 1 to 7.

9. A computer-readable storage medium, characterized in that: A program is stored thereon, and when the program is executed by a processor, a psychological state assessment method based on an artificial intelligence algorithm model as described in any one of claims 1 to 7 is implemented.

Images