Psychological assessment method, device and equipment based on common situation and storage medium

Abstract

The invention relates to the field of psychological health assessment, and discloses an estrus-sharing-based psychological assessment method, device and equipment and a storage medium. The method comprises the following steps: acquiring an emotion recognition summary result of a user; obtaining an answer record of the user to the questions in the DSM-5 psychological assessment question bank, and judging whether common-situation reply needs to be carried out or not according to the answer record; if yes, according to the answer record, the psychological and mental state of the user is obtained through analysis through a natural language processing technology of an estrus sharing model; according to the psychological and mental state, the emotion recognition summarizing result and the dialogue context of the user, generating a corresponding estrus sharing reply for the user through an estrus sharing model; and generating a psychological assessment report according to the psychological state and emotion recognition summarizing result. According to the embodiment of the invention, the psychological state of the user can be obtained through automatic analysis, the interference of human subjective factors in a traditional assessment method is reduced, the objectivity of assessment data is ensured, and thus the accuracy of psychological assessment is improved.

Description

Technical Field

[0001] The present invention relates to the field of mental health assessment, and in particular to an empathy-based psychological assessment method, device, equipment and storage medium. Background Art

[0002] Mental health refers to a continuous and positive mental state in which the subject can adapt well and fully develop his or her physical and mental potential. How to maintain people's mental health and actively intervene and guide those with psychological problems is of great importance to the entire family, society and company.

[0003] Traditional psychological assessment methods such as questionnaires and scale filling, although widely used, have obvious limitations. These methods rely on individual subjective choices and are easily affected by test motivation, test anxiety, physiological factors, etc., which can lead to inaccurate assessment results. In addition, the objectivity and errors in the scoring process can also affect the final results. Summary of the invention

[0004] The main purpose of the present invention is to solve the technical problem of low accuracy of psychological assessment.

[0005] A first aspect of the present invention provides a psychological assessment method based on empathy, the psychological assessment method based on empathy comprising:

[0006] Obtaining a summary result of emotion recognition of the user, wherein the summary result of emotion recognition is obtained by fusing various facial expression features and voice intonation features of the user through a multimodal model;

[0007] Obtaining the user's answer record to questions in the DSM-5 psychological assessment question bank, and judging whether an empathic response is needed based on the answer record;

[0008] If necessary, analyzing the user's psychological state based on the answer record through natural language processing technology of the empathy model;

[0009] Generate a corresponding empathy response for the user through the empathy model according to the psychological state, the emotion recognition summary result, and the user's conversation context;

[0010] A psychological assessment report is generated based on the psychological and mental state and the emotion recognition summary results.

[0011] Optionally, in a first implementation of the first aspect of the present invention, after generating a corresponding empathic response for the user through the empathy model according to the psychological and mental state, the emotion recognition summary result, and the user's conversation context, the method further includes:

[0012] Obtaining the real-time emotional state of the user;

[0013] According to the real-time emotional state, dynamically adjusting the question settings in the DSM-5 psychological assessment question bank through the empathy model;

[0014] Feedback the adjustment issues to the user terminal.

[0015] Optionally, in a second implementation of the first aspect of the present invention, obtaining the real-time emotional state of the user includes:

[0016] Capturing the user's facial image and answer audio in real time;

[0017] Extracting the user's real-time facial expression features and voice intonation features from the facial image and the answer audio;

[0018] The real-time emotional state of the user is obtained according to the real-time facial expression features and the voice intonation features through the multimodal model.

[0019] Optionally, in a third implementation of the first aspect of the present invention, analyzing the user's psychological state through a natural language processing technology of an empathy model based on the answer record includes:

[0020] Determining whether the answer record is an answer record corresponding to an open-ended question;

[0021] If yes, extract the keywords in the answer record through the TextRank algorithm;

[0022] According to the keywords, the user's conversation context is extracted from a database through the natural language processing technology of the empathy model, and according to the keywords, the user's intention is identified through the classification model;

[0023] The user's psychological state is analyzed based on the context, the keywords, and the intention.

[0024] Optionally, in a fourth implementation of the first aspect of the present invention, extracting keywords from the answer record by using a TextRank algorithm includes:

[0025] dividing the answer record into a plurality of parts;

[0026] Using the TextRank algorithm, a keyword graph is constructed for each part, and multiple keywords of each part are extracted;

[0027] According to the keyword graph, combined with the TF-IDF algorithm, the frequency, length and position of each keyword in each part are obtained;

[0028] According to the frequency, length and position of each keyword in each part, each keyword is scored and sorted to select the keywords in the answer record.

[0029] Optionally, in a fifth implementation of the first aspect of the present invention, the empathy model is a psychological vertical model, which is obtained by training with high-quality artificial data and synthetic data, and can provide the user with corresponding empathy responses for different psychological counseling scenarios.

[0030] Optionally, in a sixth implementation of the first aspect of the present invention, before analyzing the user's psychological state through a natural language processing technology of an empathy model according to the answer record, the method further includes:

[0031] Generate a corresponding key through an encryption algorithm, and encrypt the answer record through the key;

[0032] When the answer record is obtained, the answer record is decrypted using the key.

[0033] A second aspect of the present invention provides an empathy-based psychological assessment device, comprising: a memory and at least one processor, wherein instructions are stored in the memory, and the memory and the at least one processor are interconnected via a line; the at least one processor calls the instructions in the memory so that the empathy-based psychological assessment device executes the above-mentioned empathy-based psychological assessment method.

[0034] A third aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the above-mentioned empathy-based psychological assessment method.

[0035] In an embodiment of the present invention, the user's emotion recognition summary result is obtained, and the emotion recognition summary result is obtained by fusing the user's various facial expression features and voice intonation features through a multimodal model; the user's answer record to the questions in the DSM-5 psychological assessment question bank is obtained, and based on the answer record, it is determined whether an empathic response is required; if necessary, the user's psychological state is analyzed based on the answer record through the natural language processing technology of the empathy model; based on the psychological state, the emotion recognition summary result, and the user's conversation context, a corresponding empathic response is generated for the user through the empathy model; and a psychological assessment report is generated based on the psychological state and the emotion recognition summary result. In the present invention, the accuracy of emotion recognition is improved through a multimodal model, and the psychological state of the user is automatically analyzed through the natural language processing technology of the empathy model, which reduces the interference of human subjective factors in traditional assessment methods, ensures the objectivity of assessment data, and thus improves the accuracy of psychological assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 Schematic diagram of an embodiment of a psychological assessment method based on empathy in an embodiment of the present invention;

[0037] Figure 2 is a schematic diagram of an embodiment of a psychological assessment device based on empathy in an embodiment of the present invention;

[0038] Figure 3 Schematic diagram of an embodiment of an empathy-based psychological assessment device in an embodiment of the present invention. DETAILED DESCRIPTION

[0039] Embodiments of the present invention provide an empathy-based psychological assessment method, apparatus, device, and storage medium.

[0040] The embodiments disclosed in the present invention will be described in more detail below with reference to the accompanying drawings. Although certain embodiments disclosed in the present invention are shown in the accompanying drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as being limited to the embodiments described herein, which are instead provided for a more thorough and complete understanding of the present invention. It should be understood that the drawings and embodiments disclosed in the present invention are only for exemplary purposes and are not intended to limit the scope of protection disclosed in the present invention.

[0041] In the description of the embodiments disclosed in the present invention, the term "including" and similar terms should be understood as open inclusion, that is, "including but not limited to". The term "based on" should be understood as "based at least in part on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc. may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0042] For ease of understanding, the specific process of the embodiment of the present invention is described below. Figure 1 , an embodiment of the empathy-based psychological assessment method in an embodiment of the present invention includes:

[0043] S100, obtaining a summary result of emotion recognition of the user.

[0044] In this embodiment, a large multimodal model of vision and speech is introduced to identify the user's facial emotions and voice emotions in real time when the user answers questions, and obtain highly accurate user emotions (such as happy, sad, anxious, etc.). Specifically, a convolutional neural network (CNN) is first used to extract the features of facial expressions. CNN automatically learns and extracts high-level features in images through multiple convolutional layers, pooling layers, and fully connected layers. These features can include low-level features such as edges, textures, shapes, and more complex patterns. CNN can be used to extract local features of facial expression images, such as features in areas such as eyes, eyebrows, and mouths. And the features such as the tone, rhythm, and intensity of speech are extracted through speech signal processing technology. Then, the cross-modal attention mechanism is used to realize the sharing and fusion of multimodal features, thereby identifying the user's emotions. A multi-layer perceptron or other classifier (such as LSTM) is used to classify the fused features for emotions, and the user's emotional state is output. The emotional states identified at each time node are summarized to obtain the user's emotional recognition summary results.

[0045] S200, obtaining the user's answer record to questions in the DSM-5 psychological assessment question bank, and judging whether an empathic response is needed based on the answer record.

[0046] In this embodiment, to determine whether the user needs an empathic response, it is first necessary to obtain the user's answer record in the DSM-5 psychological assessment question bank, where DSM-5 (Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition) is a tool for diagnosing and classifying mental disorders, which includes multiple assessment scales for clinical assessment and research. According to the user's answer record in the DSM-5 psychological assessment question bank, it can be determined whether the user shows an emotional or psychological state that requires an empathic response. For example, if the user answers certain scales related to depression, anxiety or other psychological problems, such as PHQ-9 or GAD-7, these scales can assess the user's depression and anxiety levels. Once it is determined that the user may need empathic support, empathic response generation technology can be used to provide an appropriate response. Empathic response generation aims to provide psychological support by understanding the user's emotions and experiences and generating empathetic responses.

[0047] S300: If necessary, the user's psychological state is analyzed based on the answer record by using the natural language processing technology of the empathy model.

[0048] In this embodiment, if an empathic response is required, the user's answer records to questions in the DSM-5 psychological assessment question bank are first collected, and the natural language understanding ability of the empathy model is used to analyze the user's open-ended answer records to specific question bank questions, and the user's psychological condition and general mental condition are generated according to the user's answer records. Specifically, the emotional polarity is identified by analyzing the vocabulary, grammatical structure and context in the answer record. For example, a dictionary method or a machine learning model is used to identify positive, negative or neutral emotions in the answer record, and the user's psychological condition and general mental condition are generated by identifying the emotional polarity. The present invention generates the user's psychological condition and general mental condition based on the user's answer record, which can reduce the influence of human subjective factors on the evaluation results and ensure the objectivity of the data and the accuracy of the evaluation.

[0049] S400, generating a corresponding empathy response for the user through an empathy model according to the psychological state, the emotion recognition summary result, and the user's conversation context.

[0050] In this embodiment, first, at different time points, the user's facial image and answer audio are captured, facial expression features and voice intonation features are extracted from the facial image and answer audio, and the facial expression features and voice intonation features are input into the multimodal model to obtain the user's emotional state at different time points, and the emotional state at different time points is summarized to obtain the emotion recognition summary result. Secondly, the user's answer record is analyzed to generate the user's psychological condition and general mental condition. Then, the emotion recognition summary result, the user's psychological condition and general mental condition, and the user's conversation context are input into the empathy model to generate a suitable empathy response for the user.

[0051] S500, generating a psychological assessment report based on the psychological and mental state and emotion recognition summary results.

[0052] In this embodiment, a psychological assessment report is generated based on the summary results of psychological mental state and emotion recognition for diagnosis through a psychological assessment system. This system has built-in 15 main complaint modules and 23 main complaint diagnosis modules to ensure that the assessment process can cover all common psychological problems. It covers common psychological problems and their diagnostic criteria, and the diagnoses include: no recommended diagnosis, depression, bipolar disorder type I, single manic episode, bipolar disorder type II, panic disorder, agoraphobia, social phobia, obsessive-compulsive disorder, post-traumatic stress disorder, mild non-alcoholic substance use disorder, moderate non-alcoholic substance use disorder, severe non-alcoholic substance use disorder, mild alcohol use disorder, moderate alcohol use disorder, severe alcohol use disorder, affective disorder with psychotic symptoms, psychotic disorder, anorexia nervosa, mild Severe bulimia nervosa, moderate bulimia nervosa, severe bulimia nervosa, extremely severe bulimia nervosa, anorexia nervosa binge eating / purging type, anorexia nervosa restricting type, mild binge eating disorder, moderate binge eating disorder, severe binge eating disorder, extremely severe binge eating disorder, generalized anxiety disorder, antisocial personality disorder, attention deficit hyperactivity disorder, attention deficit disorder, hyperactivity disorder, conduct disorder, low suicide risk, moderate suicide and self-harm risk, high suicide and self-harm risk, insomnia symptoms, persistent insomnia symptoms, daytime functional impairment due to insomnia. Through systematic analysis and diagnosis of the main complaint, users can get a comprehensive mental health assessment in one assessment process, which greatly improves the breadth and practicality of the assessment.

[0053] In an optional implementation of the first aspect of the present invention, after generating a corresponding empathy response for the user through an empathy model according to the psychological state, the emotion recognition summary result, and the user's conversation context, the method further includes:

[0054] Obtain the user's real-time emotional state; dynamically adjust the question settings in the DSM-5 psychological assessment question bank based on the real-time emotional state through the empathy model; and feed back the adjusted questions to the user terminal.

[0055] In this embodiment, the real-time emotional state of the user's feedback is obtained, and the evaluation process and question settings are dynamically adjusted according to the user's real-time emotional state feedback, so that the evaluation process is more personalized and humane, and the user's sense of participation and the accuracy of the evaluation are improved.

[0056] In an optional implementation of the first aspect of the present invention, obtaining the real-time emotional state of the user includes:

[0057] Capture the user's facial image and answer audio in real time; extract the user's real-time facial expression features and voice intonation features from the facial image and answer audio; and obtain the user's real-time emotional state through a multimodal model based on the real-time facial expression features and voice intonation features.

[0058] In this embodiment, real-time capture of the user's facial image and answer audio can be achieved through a variety of technologies. Facial expression recognition usually includes image acquisition, facial detection, image preprocessing and expression classification. Common facial feature extraction methods include Gabor filter, directional histogram HOG, discrete cosine transform (DCT) and scale invariant feature transform (SIFT), and then SVM or other methods are used to classify expressions. In terms of voice, voice intonation features are extracted, and commonly used methods include spectrogram images, etc. By capturing the user's facial image and answer audio in real time, and using a multimodal model to fuse facial expression features and voice intonation features, the user's real-time emotional state can be effectively identified.

[0059] In an optional implementation of the first aspect of the present invention, according to the answer record, the user's psychological state is analyzed by natural language processing technology of the empathy model, including:

[0060] Determine whether the answer record is the answer record corresponding to the open-ended question; if so, extract the keywords in the answer record through the TextRank algorithm; based on the keywords, extract the user's conversation context from the database through the natural language processing technology of the empathy model, and identify the user's intention based on the keywords through the classification model; analyze the user's psychological and mental state based on the context, keywords, and intention.

[0061] In this embodiment, it is first determined whether the answer record is an answer record corresponding to an open-ended question. If so, the keywords in the answer record are extracted by the TextRank algorithm. Keyword extraction is an important task in natural language processing, which aims to automatically extract representative, general and important words or phrases from the text. The TextRank algorithm is an automatic keyword extraction method based on statistical information. It does not need to rely on domain knowledge or grammatical rules, and only needs to be based on the statistical information of the text itself to obtain keywords. Then, based on the keywords, the user's conversation context is extracted from the database through the natural language processing technology of the empathy model. Specifically, the structure and semantic relationship of the sentences in the answer record are analyzed in combination with the keywords, and the attention mechanism is used to capture the important parts of the answer record to help understand the context. And based on the keywords, the intention or purpose expressed in the user's answer record is determined. Commonly used intention recognition methods include rule-based methods, statistical methods, deep learning-based methods, etc. Then, the user's psychological state is analyzed in combination with context information, keywords, and intentions.

[0062] In an optional implementation of the first aspect of the present invention, extracting keywords from the answer record by using the TextRank algorithm includes:

[0063] The answer record is divided into multiple parts; a keyword graph is constructed for each part through the TextRank algorithm, and multiple keywords of each part are extracted; based on the keyword graph, the frequency, length and position of each keyword in each part are obtained in combination with the TF-IDF algorithm; based on the frequency, length and position of each keyword in each part, each keyword is scored and ranked to select the keywords in the answer record.

[0064] In this embodiment, before keyword extraction, the answer record text is preprocessed, including cleaning text data, removing irrelevant characters, punctuation marks, stop words, etc., dividing the answer record text into several parts, constructing a keyword graph for each part, and extracting several keywords in each part. Finally, the final keywords are selected based on comprehensive factors such as word frequency, length, position and part of speech. This method can avoid the phenomenon that the connection between sentences is severed when the traditional TextRank algorithm divides the text into sentences, thereby improving the accuracy. Among them, combined with the TF-IDF algorithm, the word frequency, length and position of each keyword in each part can be obtained, and the importance of words and text structure can be considered more comprehensively, thereby improving the accuracy of keyword extraction. Optionally, keyword weights can be assigned during the iteration process to achieve a more accurate calculation of the influence of nodes. In addition, Latent Dirichlet Allocation (LDA) can be used to capture word similarity information, thereby improving the effect of keyword extraction.

[0065] In an optional implementation of the first aspect of the present invention, the empathy model is a psychological vertical model, which is obtained by training with high-quality artificial data and synthetic data, and can provide users with corresponding empathy responses for different psychological counseling scenarios.

[0066] In this embodiment, the empathy model is an application of a psychological vertical model. The psychological vertical model refers to a large-scale language model designed and optimized specifically for the field of mental health, which aims to provide psychological support and treatment services. This model generates empathic responses by understanding and perceiving the user's emotional state and combining psychological knowledge. During training, the data set for empathy model training uses synthetic data to supplement the deficiencies of real data, improve data quality and quantity, and reduce the cost of data collection and processing. An important role of synthetic data in the training of psychological vertical models is to improve the alignment ability of the model and the vertical application effect. By using synthetic data, ordinary optimization data can be replaced to improve the performance of the model. This method is particularly suitable for scenarios that require highly customized features, such as hoping that the language model has a specific bias when answering certain questions. Since the quality of the acquired data is often uneven, it is necessary to manually collect high-quality data.

[0067] In an optional implementation of the first aspect of the present invention, before analyzing and obtaining the user's psychological state through the natural language processing technology of the empathy model according to the answer record, the method further includes:

[0068] A corresponding key is generated through an encryption algorithm, and the answer record is encrypted with the key; when the answer record is obtained, the answer record is decrypted with the key.

[0069] In this embodiment, the user's answer record needs to be encrypted to ensure the user's privacy and security. Specifically, a corresponding key is generated through an encryption algorithm, and the answer record is encrypted with the key. For authorized users, the key is obtained, and when the answer record is obtained, the same key is used to decrypt the answer record.

[0070] See also Figure 2 According to a second aspect of the present invention, there is provided a psychological assessment device based on empathy, the psychological assessment device based on empathy comprising:

[0071] The emotion result acquisition module 10 is used to acquire the emotion recognition summary result of the user, where the emotion recognition summary result is obtained by fusing the facial expression features and voice intonation features of the user through a multimodal model;

[0072] The empathic response judgment module 20 is used to obtain the user's answer record to the questions in the DSM-5 psychological assessment question bank, and judge whether an empathic response is needed based on the answer record;

[0073] The psychological state analysis module 30 is used to analyze the psychological state of the user according to the answer record by using the natural language processing technology of the empathy model if necessary;

[0074] An empathy response generation module 40 is used to generate a corresponding empathy response for the user through an empathy model according to the psychological state, the emotion recognition summary result, and the user's conversation context;

[0075] The assessment report generation module 50 is used to generate a psychological assessment report based on the psychological and mental state and emotion recognition summary results.

[0076] In an optional implementation of the second aspect of the present invention, the empathy-based psychological assessment device further includes:

[0077] The question adjustment module is used to obtain the user's real-time emotional state; according to the real-time emotional state, the question settings in the DSM-5 psychological assessment question bank are dynamically adjusted through the empathy model; and the adjusted questions are fed back to the user terminal.

[0078] In an optional implementation of the second aspect of the present invention, the question adjustment module is also used to capture the user's facial image and answer audio in real time; extract the user's real-time facial expression features and voice intonation features from the facial image and answer audio; and obtain the user's real-time emotional state through a multimodal model based on the real-time facial expression features and voice intonation features.

[0079] In an optional implementation of the second aspect of the present invention, the psychological state analysis module 30 is also used to determine whether the answer record is an answer record corresponding to an open-ended question; if so, the keywords in the answer record are extracted through the TextRank algorithm; based on the keywords, the user's conversation context is extracted from the database through the natural language processing technology of the empathy model, and based on the keywords, the user's intention is identified through the classification model; based on the context, keywords, and intention, the user's psychological state is analyzed.

[0080] In an optional implementation of the second aspect of the present invention, the mental state analysis module 30 is also used to divide the answer record into multiple parts; through the TextRank algorithm, a keyword graph is constructed for each part, and multiple keywords of each part are extracted; based on the keyword graph, combined with the TF-IDF algorithm, the word frequency, length and position of each keyword in each part are obtained; according to the word frequency, length and position of each keyword in each part, each keyword is scored and sorted to select the keywords in the answer record.

[0081] In an optional implementation of the second aspect of the present invention, the empathy-based psychological assessment device further includes:

[0082] The encryption and decryption module is used to generate the corresponding key through the encryption algorithm, and encrypt the answer record through the key; when obtaining the answer record, the answer record is decrypted through the key.

[0083] Figure 3 is a schematic diagram of the structure of an empathy-based psychological assessment device provided by an embodiment of the present invention. The empathy-based psychological assessment device 500 may have relatively large differences due to different configurations or performances, and may include one or more processors (central processing units, CPU) 510 (for example, one or more processors) and a memory 520, and one or more storage media 530 (for example, one or more mass storage devices) storing application programs 533 or data 532. Among them, the memory 520 and the storage medium 530 may be temporary storage or permanent storage. The program stored in the storage medium 530 may include one or more modules (not shown in the figure), and each module may include a series of instruction operations in the empathy-based psychological assessment device 500. Furthermore, the processor 510 may be configured to communicate with the storage medium 530 to execute a series of instruction operations in the storage medium 530 on the empathy-based psychological assessment device 500.

[0084] The empathy-based psychological assessment device 500 may also include one or more power supplies 540, one or more wired or wireless network interfaces 550, one or more input and output interfaces 560, and / or one or more operating systems 531, such as Windows Serve, Mac OS X, Unix, Linux, FreeBSD, etc. It will be appreciated by those skilled in the art that Figure 3 The illustrated structure of the empathy-based psychological assessment device does not constitute a limitation on the empathy-based psychological assessment device, and may include more or less components than illustrated, or may combine certain components, or may have different component arrangements.

[0085] The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium. The computer-readable storage medium stores instructions, which, when executed on a computer, enable the computer to execute the steps of the empathy-based psychological assessment method.

[0086] In the context of the present disclosure, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0087] In addition, although each operation is described in a specific order, this should be understood as requiring such operation to be performed in the specific order shown or in a sequential order, or requiring that all illustrated operations should be performed to obtain desired results. Under certain circumstances, multitasking and parallel processing may be advantageous. Similarly, although some specific implementation details are included in the above discussion, these should not be interpreted as limiting the scope of the present disclosure. Some features described in the context of a separate embodiment can also be implemented in a single implementation in combination. On the contrary, the various features described in the context of a single implementation can also be implemented in multiple implementations individually or in any suitable sub-combination mode.

[0088] Although the subject matter has been described in language specific to structural features and / or methodological logical actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. On the contrary, the specific features and actions described above are merely example forms of implementing the claims.

Claims

1. A psychological assessment method based on empathy, characterized in that: The empathy-based psychological assessment method includes: Obtaining a summary result of emotion recognition of the user, wherein the summary result of emotion recognition is obtained by fusing various facial expression features and voice intonation features of the user through a multimodal model; Obtaining the user's answer record to questions in the DSM-5 psychological assessment question bank, and judging whether an empathic response is needed based on the answer record; If necessary, analyzing the user's psychological state based on the answer record through natural language processing technology of the empathy model; Generate a corresponding empathy response for the user through the empathy model according to the psychological state, the emotion recognition summary result, and the user's conversation context; A psychological assessment report is generated based on the psychological and mental state and the emotion recognition summary results.

2. The empathy-based psychological assessment method according to claim 1, characterized in that: After generating a corresponding empathy response for the user through the empathy model according to the psychological state, the emotion recognition summary result, and the user's conversation context, the method further includes: Obtaining the real-time emotional state of the user; According to the real-time emotional state, dynamically adjusting the question settings in the DSM-5 psychological assessment question bank through the empathy model; Feedback the adjustment issues to the user terminal.

3. The empathy-based psychological assessment method according to claim 2, characterized in that: The obtaining of the real-time emotional state of the user comprises: Capturing the user's facial image and answer audio in real time; Extracting the user's real-time facial expression features and voice intonation features from the facial image and the answer audio; The real-time emotional state of the user is obtained according to the real-time facial expression features and the voice intonation features through the multimodal model.

4. The empathy-based psychological assessment method according to claim 1, characterized in that: The psychological state of the user analyzed based on the answer record by using the natural language processing technology of the empathy model includes: Determining whether the answer record is an answer record corresponding to an open-ended question; If yes, extract the keywords in the answer record through the TextRank algorithm; According to the keywords, the user's conversation context is extracted from a database through the natural language processing technology of the empathy model, and according to the keywords, the user's intention is identified through the classification model; The user's psychological state is analyzed based on the context, the keywords, and the intention.

5. The empathy-based psychological assessment method according to claim 4, characterized in that: The extracting of keywords from the answer record by using the TextRank algorithm includes: dividing the answer record into a plurality of parts; Using the TextRank algorithm, a keyword graph is constructed for each part, and multiple keywords of each part are extracted; According to the keyword graph, combined with the TF-IDF algorithm, the frequency, length and position of each keyword in each part are obtained; According to the frequency, length and position of each keyword in each part, each keyword is scored and sorted to select the keywords in the answer record.

6. The empathy-based psychological assessment method according to claim 1, characterized in that: The empathy model is a large psychological vertical model, which is trained by using high-quality artificial data and synthetic data, and can provide the user with corresponding empathy responses for different psychological counseling scenarios.

7. The empathy-based psychological assessment method according to claim 1, characterized in that: Before analyzing the user's psychological state based on the answer record by using the natural language processing technology of the empathy model, the method further includes: Generate a corresponding key through an encryption algorithm, and encrypt the answer record through the key; When the answer record is obtained, the answer record is decrypted using the key.

8. A psychological assessment device based on empathy, characterized in that: The empathy-based psychological assessment device comprises: An emotion result acquisition module is used to acquire a summary result of emotion recognition of a user, wherein the summary result of emotion recognition is obtained by fusing various facial expression features and voice intonation features of the user through a multimodal model; An empathic response judgment module is used to obtain the user's answer record to the questions in the DSM-5 psychological assessment question bank, and judge whether an empathic response is needed based on the answer record; A psychological state analysis module, for analyzing the psychological state of the user according to the answer record by using the natural language processing technology of the empathy model if necessary; An empathy response generation module, configured to generate a corresponding empathy response for the user through the empathy model according to the psychological state, the emotion recognition summary result, and the user's conversation context; An assessment report generating module is used to generate a psychological assessment report based on the psychological and mental state and the emotion recognition summary result.

9. A psychological assessment device based on empathy, characterized in that: The empathy-based psychological assessment device comprises: a memory and at least one processor, the memory storing instructions, the memory and the at least one processor being interconnected via a line; The at least one processor calls the instructions in the memory to enable the empathy-based psychological assessment device to perform the empathy-based psychological assessment method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the empathy-based psychological assessment method according to any one of claims 1 to 7 is implemented.

Images