Method and system for identifying personality traits based on electroencephalogram signals

Through the recognition method based on EEG signal, LSSENet classification model and neural network technology are used to solve the subjectivity and universality of personality recognition in the existing technology, and objective and accurate assessment of personality traits are achieved, and the recognition is improved. Reliability and applicability.

CN118948282APending Publication Date: 2024-11-15DALIAN NATIONALITIES UNIVERSITY
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411172679.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-26
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

The lack of objective and automated personality recognition methods in the prior art leads to the fact that personality measurement results are susceptible to subjective falsehood, and the testing materials lack universality, resulting in identification bias caused by factors such as cultural background and educational level.

Method used

Using personality trait recognition method based on EEG signals, the EEG response signal was collected by playing personality test materials, preprocessing and standardizing, and inputting it into the pre-trained LSSENet classification model to calculate the subject's personality trait score, combined with Full convolutional neural network, LSTM network and attention mechanism are classified.

Benefits of technology

It realizes an objective, accurate and authentic assessment of personality traits, avoids subjective influence, improves the reliability and effectiveness of identification, and has the applicability of cross-cultural background and educational level.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118948282A_ABST
    Figure CN118948282A_ABST
Patent Text Reader

Abstract

The invention discloses a personality characteristic identification method and system based on electroencephalogram signals, and belongs to the technical field of electroencephalogram signal analysis. The method comprises the following steps: playing a personality test material to a subject to induce a personality trait related cognitive state of the subject, and collecting an electroencephalogram response signal of the subject; the collected electroencephalogram response signals are preprocessed to filter artifact noise and standardized signals, and a training sample set is constructed; and inputting the electroencephalogram response signal into a pre-trained personality trait classification model, calculating the input electroencephalogram signal, and calculating the personality trait score of the subject. According to the method, automatic measurement of personality characteristics is realized, and based on objective electroencephalogram data, the personality characteristics can be evaluated more accurately and truly, so that the reliability and effectiveness of personality characteristic identification are improved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of electroencephalogram (EEG) signal analysis and engineering psychology, and in particular to a method and system for identifying personality traits based on EEG signals. Technical Background

[0002] Personality is a characteristic set of human behavior and emotional patterns that evolve from biological and environmental factors. The most common way to describe personality is the Big Five personality. In the Big Five personality theory, personality is divided into openness, which indicates the openness of personality; conscientiousness, which indicates the level of prudence of personality; extraversion, which indicates the extroversion of personality; agreeableness, which indicates the affinity of personality; and neuroticism, which indicates the stability of emotions. The Big Five personality theory not only promotes a deep insight into individual psychology, but also provides a scientific basis for the study of academic achievement, social status, career success, and online behavior in real life.

[0003] Although the study of personality traits has always been a key focus in some academic fields, the identification of personality traits has always faced many challenges. In the existing technology, the mainstream methods are mainly based on self-evaluation questionnaires and other people's evaluation methods based on short videos. However, in the self-evaluation questionnaire method, the subjects are asked to answer questions related to personality traits, and their answers are evaluated by the examiner based on professional experience to obtain the final measurement results. The results of this method are completely determined by the personal reports of the subjects and are easily interfered by subjective falsification. The evaluation method of others based on short videos also has similar problems. This identification method also lacks objective evaluation criteria and requires a lot of time and manpower costs.

[0004] In the prior art, there is no method that can achieve objective and automated personality recognition so as to measure the personality of the subject directly and quickly. At the same time, in the existing personality recognition technology, the test materials often lack universality, and the recognition results may also have certain deviations due to the different cultural backgrounds of the subjects. Summary of the invention

[0005] The present invention proposes a personality recognition method and system based on EEG signals to solve the problem of lack of objective and automated personality recognition methods in the prior art, thereby achieving intuitive and rapid personality measurement of subjects.

[0006] The technical solution adopted by the present invention is: a personality trait recognition method based on EEG signals, comprising the following steps:

[0007] S1. Playing personality test materials to the subjects to induce the cognitive state related to the personality traits of the subjects, and collecting the EEG response signals of the subjects;

[0008] S2. preprocessing the collected EEG response signals to filter out artifact noise and standardize signals, and constructing a training sample set;

[0009] S3. Input the EEG response signal into a pre-trained personality trait classification model, calculate the input EEG signal, calculate the personality trait score of the subject, and complete the classification;

[0010] The classification model is a LSSENet classification model;

[0011] In the spatial dimension, the electrode channels that contribute to personality classification are identified by integrating the fully convolutional neural network and the compressed excitation network. The input of the structure is the EEG data that has been de-aliased and standardized, and the shape is (N×C×S), where N represents the batch size, C is the number of channels, and each channel has S time steps. There are three temporal convolution blocks in the structure, and the convolution block contains a convolution layer, a batch normalization layer, and a ReLU activation layer. After the first two convolution blocks, the SE Block is inserted to extract channel features of the EEG signal, and its weight matrix represents the contribution of each electrode channel to feature extraction. After the third convolution block, the global average pooling layer is connected.

[0012] In the time dimension, a 2D convolution filter is used to extract the time dimension information of the EEG signal. The 2D convolution filter is composed of a 2D convolution block, which includes a 2D convolution layer, a batch normalization layer and a ReLU activation layer; the kernel size of the 2D convolution filter is set to C×S×1; batch normalization is normalized at the level, and the ReLU activation layer introduces nonlinearity in the network;

[0013] Use LSTM network to learn the short-term continuity and long-term similarity features of EEG signals, where the forget gate determines which information should be forgotten, and the input gate and output gate selectively update the current state;

[0014] Inserting the attention mechanism enables the subsequent LSTM to focus on long-term dependencies; the core formula of the attention mechanism is:

[0015]

[0016] in, is the weight associated with time step i;

[0017] The Conact layer is used to fuse the temporal and spatial feature maps according to the channel dimension, and then the softmax function is used to complete the final classification of the Big Five personality traits.

[0018] Before step S1, the following steps are also included:

[0019] For the 20 emotions in the positive and negative emotion scale, the corresponding materials are selected as personality test materials; the personality test materials in the material library are divided into 4 categories according to the valence threshold and arousal value: high valence and high arousal HVHA, high valence and low arousal HVLA, low valence and high arousal LVHA and low valence and low arousal LVLA; the type of personality test material is video, with a duration between 2 minutes and 20 minutes;

[0020] The subjects specifically include: people of different ages with different cultural backgrounds and different educational levels.

[0021] Before step 2, the method further includes:

[0022] Construct a training sample library based on the EEG data of multiple people with different personality traits and their corresponding personality scores;

[0023] The personality trait classification model is trained through the training sample library.

[0024] The EEG data in step S2 specifically includes: frequency band energy of delta, theta, alpha, beta and gamma bands in the EEG signal.

[0025] The EEG signals include at least: EEG signals obtained from electrodes AF3, F7, F3, FC5, T7, P7, O1, O2, P8, T8, FC6, F4, F8 and AF4.

[0026] When playing the personality test materials to the subject in step S1, the personality test materials are alternately sorted according to their categories and then played through the playback device.

[0027] In the step 3, a classification model is constructed with the self-assessment questionnaire score of the individual's personality traits as the target; wherein the classification model is an LSSENet classification model; a personality trait measurement method is used to perform questionnaire scores on the personality traits of the training individual; and the classification model is trained according to the questionnaire scores of the training individual to obtain a pre-trained personality trait classification model.

[0028] A personality trait recognition system based on EEG signals, the system is used to implement the above method, the system comprising:

[0029] A signal receiving module, used to play personality test materials to the subject to induce the subject's personality trait-related cognitive state, and collect the subject's electroencephalogram response signal;

[0030] A data processing module is used to pre-process the collected EEG response signals to filter out artifact noise and standardize signals, and construct a training sample set;

[0031] The calculation module is used to input the EEG response signal into a pre-trained personality trait classification model, calculate the input EEG signal, calculate the personality trait score of the subject, and complete the classification.

[0032] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the steps of the above method are implemented when the processor executes the program.

[0033] A non-transitory computer-readable storage medium stores a computer program, which implements the steps of the above method when executed by a processor.

[0034] In a first aspect, the present invention provides a personality identification method based on electroencephalogram signals, which comprises: playing personality test materials to a subject to induce a cognitive state related to the subject's personality traits, and collecting the subject's electroencephalogram response signals; preprocessing the collected electroencephalogram response signals to filter out artifact noise and standardized signals, and constructing them into a training sample set; inputting the electroencephalogram response signals into a pre-trained personality trait classification model, calculating the input electroencephalogram signals, and calculating the subject's personality trait score.

[0035] In a second aspect, the present invention provides a personality recognition system based on electroencephalogram signals, comprising: a signal receiving module, used to play personality test materials to a subject to induce the subject's personality trait-related cognitive state, and collect the subject's electroencephalogram response signal; a data processing module, used to pre-process the collected electroencephalogram response signal to filter out artifact noise and standardized signals, and construct a training sample set; a calculation module, used to input the electroencephalogram response signal into a pre-trained personality trait classification model, calculate the input electroencephalogram signal, and calculate the subject's personality trait score.

[0036] In a third aspect, the present invention provides an electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, the steps of the method for identifying personality traits based on EEG signals as provided in the first aspect above are implemented.

[0037] In a fourth aspect, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the personality trait identification method based on EEG signals as provided in the first aspect above.

[0038] The beneficial effects of the present invention are as follows: the method plays personality test materials to the subject to induce the cognitive state related to the subject's personality traits, collects the subject's electroencephalogram response signals, and then uses a pre-trained classification model to classify the subject's electroencephalogram signals into personality traits. Based on objective electroencephalogram signals, this method can avoid the influence of the subject's personal subjective factors on the traditional subjective personality measurement method, thereby making a more accurate and true assessment of personality traits. The present invention realizes the automatic measurement of personality traits, and based on objective electroencephalogram data, can make a more accurate and true assessment of personality traits, thereby improving the reliability and effectiveness of personality trait identification. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0040] Figure 1 The figure is a flowchart of a personality recognition method based on EEG signals.

[0041] Figure 2 Schematic diagram of the personality trait identification model.

[0042] Figure 3 This is a structural diagram of a personality recognition system based on EEG signals.

[0043] Figure 4 The figure is a schematic diagram of the structure of an electronic device. DETAILED DESCRIPTION

[0044] In order to more clearly demonstrate the purpose, technical solution and advantages of the embodiment of the present invention, the technical solution of the embodiment of the present invention is described in detail below in conjunction with the accompanying drawings. Obviously, the embodiment is only a part of the present invention, not all embodiments. Based on the embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the protection scope of the present invention.

[0045] refer to Figure 1 , Figure 1 A flowchart of a personality recognition method based on EEG signals provided by an embodiment of the present invention includes:

[0046] S1, playing personality test materials to the subjects to induce the cognitive state related to the personality traits of the subjects, and collecting the EEG response signals of the subjects;

[0047] S2, preprocessing the collected EEG response signals to filter out artifact noise and standardize signals, and constructing a training sample set;

[0048] S3, inputting the EEG response signal into a pre-trained personality trait classification model, calculating the input EEG signal, and calculating the personality trait score of the subject.

[0049] Specifically, in this embodiment, taking the 20 typical positive and negative emotions involved in the Positive Negative Affect Scale (PANAS) as an example, the video materials in the material library are divided into 4 categories according to the valence threshold and arousal value: high valence high arousal (HVHA), high valence low arousal (HVLA), low valence high arousal (LVHA) and low valence low arousal (LVLA). For a new subject, 4 materials are selected from the constructed material library for each category to play, and the EEG signals induced by the subject in the test are obtained. The EEG signal sampling frequency range used in the present invention is 0-128Hz, focusing on the energy of the delta (0-1Hz), theta (1-8Hz), alpha (8-12Hz), beta (12-27Hz), gamma (27-45Hz) and other frequency bands that appear after the presentation of the induced material. The event-related potential response of the subject to different induced materials is preprocessed into a vector as the input data of the personality trait classification model, and the personality trait score of the subject is calculated by the classification model.

[0050] Based on the above embodiments, the EEG signals at least include: EEG signals obtained from electrodes AF3, F7, F3, FC5, T7, P7, O1, O2, P8, T8, FC6, F4, F8 and AF4.

[0051] On the basis of the above embodiment, before the step of playing the personality test materials in the preset material library to the subject, it also includes: for each emotion in the positive and negative emotion scale, the materials in the material library are divided into 4 categories according to the valence threshold and the arousal value: high valence and high arousal (HVHA), high valence and low arousal (HVLA), low valence and high arousal (LVHA) and low valence and low arousal (LVLA), and 4 materials are selected as alternative materials for each category; wherein the type of the material is video.

[0052] Through this method, personality test materials are played to the subjects to induce cognitive states related to the subjects' personality traits, and the subjects' EEG response signals are collected. The subjects' EEG signals are then classified into personality traits using a pre-trained classification model. Based on objective EEG signals, this method can avoid the traditional subjective personality measurement method being affected by the subjects' personal subjective factors, thereby providing a more accurate and true assessment of personality traits.

[0053] In summary, the embodiment of the present invention proposes an automatic identification method for quickly and accurately measuring personality traits based on EEG signal analysis. By providing subjects with universal stimulation materials, the effectiveness of stimulation induction is ensured, thereby ensuring the reliability of the results and the wide applicability of the system. In addition, the diversity of material types ensures the diversity of stimulation, and realizes an automatic personality identification method across educational levels and cultural backgrounds. This method has important application value in the fields of personnel selection, career planning, etc.

[0054] refer to Figure 2 , Figure 2 A schematic diagram of the structure of a personality trait recognition model provided by one embodiment of the present invention.

[0055] Specifically, the classification model used in the present invention is LSSENet (Long Short-Term Memory and Squeeze-Excitation Network).

[0056] In the spatial dimension, the electrode channels that contribute to personality classification are identified by integrating Fully Convolutional Networks (FCN) and Squeeze-and-Excitation Network (SENet).

[0057] First, the input of the structure is the EEG data that has been de-aliased and standardized, with a shape of (N×C×S), where N represents the batch size, C is the number of channels, and each channel has S time steps. There are three temporal convolution blocks in the structure, and the convolution block contains a convolution layer, a batch normalization layer, and a ReLU activation layer. The convolution kernel size and number of filters in each convolution block are 8×128, 5×256, and 3×128, respectively. We use Uniform Initializer to initialize all weights in the three filters.

[0058] Then, this embodiment inserts SE Block after the first two convolution blocks to extract channel features of EEG signals, and its weight matrix represents the contribution of each electrode channel to feature extraction. Then, Global Average Pooling (GAP, global average pooling layer) is connected after the third convolution block.

[0059] In the time dimension, the present invention combines Convolutional Neural Networks (CNN), Attention Mechanism and Long Short-Term Memory Network (LSTM). These networks can simultaneously capture the short-term continuity and long-term consistency characteristics of EEG signals. In personality trait recognition, this combination can extract richer temporal information, thereby improving the accuracy and robustness of the classification model.

[0060] First, a 2D convolution filter is used to extract the time dimension information of the EEG signal. It consists of a 2D convolution block, including a 2D convolution layer, a batch normalization layer and a ReLU activation layer. The present invention sets the kernel size of the 2D convolution filter to C×S×1, which helps the classification model pay more attention to the short-term changes of the EEG signal. Then, a sliding window with a step size of 1 is used to extract features to keep the dimension of the feature map the same as the input data. Using padding instead of half padding enables the present invention to avoid using upsampling and interpolation methods on the feature map when constructing the attribution map. Then, batch normalization is normalized at the level, which can achieve faster convergence and better network generalization. In addition, the ReLU activation layer introduces nonlinearity in the network to improve the expressive power of the network, enabling it to better fit more complex functional relationships.

[0061] Next, the LSTM network is used to learn the short-term continuity and long-term similarity features of EEG signals, where the forget gate determines which information should be forgotten, helping the network capture and retain the long-term dependencies of EEG data. The input gate and output gate can selectively update the current state, which is conducive to capturing key information in the short term. Although LSTM has the ability to learn time dependencies in sequences, it has difficulties in long-term dependencies in long sequences.

[0062] The present invention helps LSTM learn this dependency by inserting an attention mechanism. The attention mechanism uses the calculation of attention weights to enable the subsequent LSTM to focus on long-term dependencies. The core formula of the attention mechanism can be expressed as:

[0063]

[0064] in, is the weight associated with time step i.

[0065] The present invention uses the Connect layer to fuse the temporal and spatial feature maps according to the channel dimension, and then uses the softmax function to complete the final classification of the Big Five personality traits. Since personality recognition is regarded as a classification task, the categorical crossentropy loss function is selected to train the performance of the classification model, and its definition is as follows:

[0066]

[0067] Where N represents the total number of categories. is the i-th element in the one-hot encoding of the true category, is the probability predicted by the model corresponding to class i.

[0068] Based on the above embodiment, the first evaluation index used in the embodiment of the present invention is accuracy, which is one of the most commonly used evaluation indicators in classification problems. For binary classification problems, accuracy can be defined as:

[0069]

[0070] Among them, TP represents the number of true positive examples, TN represents the number of true negative examples, FP represents the number of false positive examples, and FN represents the number of false negative examples. The second evaluation indicator is the average accuracy, which is used to represent the average classification accuracy of the five personality trait dimensions and is defined as:

[0071]

[0072] Among them, m = 5, representing the five dimensions of personality traits.

[0073] refer to Figure 3 , Figure 3 This is a structural diagram of a personality recognition system based on EEG signals provided by an embodiment of the present invention. The provided system includes: a signal receiving module 2_1, a data processing module 2_2 and a calculation module 2_3.

[0074] The signal receiving module 3_1 is used to play personality test materials to the subject to induce the subject's personality trait-related cognitive state and collect the subject's EEG response signal;

[0075] The data processing module 3_2 is used to pre-process the collected EEG response signals to filter out artifact noise and standardized signals, and construct a training sample set;

[0076] The calculation module 3_3 is used to input the EEG response signal into a pre-trained personality trait classification model, calculate the input EEG signal, and calculate the personality trait score of the subject.

[0077] Specifically, in this embodiment, taking the 20 typical positive and negative emotions involved in the Positive Affect Negative Affect Scale (PANAS) as an example, the video materials in the material library are divided into four categories according to valence and arousal value: high valence high arousal (HVHA), high valence low arousal (HVLA), low valence high arousal (LVHA) and low valence low arousal (LVLA). For each category, 4 materials are selected from the constructed material library for new subjects to play to obtain the EEG signals induced by the subjects during the test. In the present invention, the sampling frequency range of the EEG signal is 0-128Hz, focusing on the energy changes of frequency bands such as delta (0-1Hz), theta (1-8Hz), alpha (8-12Hz), beta (12-27Hz) and gamma (27-45Hz) that appear after the stimulus is presented. The event-related potential responses of the subjects to different induced materials are converted into vectors after preprocessing as input data for the personality trait classification model, and the personality trait scores of the subjects are calculated by the model.

[0078] Through this system, personality test materials are played to subjects to induce cognitive states related to their personality traits, and the subjects' EEG response signals are collected. Then, using a pre-trained classification model, these EEG signals are classified to identify personality traits. Based on objective EEG signals, this method can avoid the influence of the subject's personal subjective factors on traditional subjective personality measurement methods, thereby providing a more accurate and realistic assessment of personality traits.

[0079] refer to Figure 4 , Figure 4 FIG. 1 is a schematic diagram of the structure of an electronic device according to an embodiment of the present invention. Figure 4 As shown, the device provided includes: a processor (Processor) 4_1, a memory (Memory) 4_2, a communications interface (CommunicationsInterface) 4_3 and a bus 4_4. Among them, the processor 4_1, the communication interface 4_3, and the memory 4_2 complete mutual communication through the bus 4_4. The processor 4_1 can call the logic instructions in the memory 4_2 to execute the following method, for example, including: playing personality test materials to the subject to induce the cognitive state related to the personality trait of the subject, and collecting the subject's electroencephalogram response signal; preprocessing the collected electroencephalogram response signal to filter out artifact noise and standardized signals, and constructing it into a training sample set; inputting the electroencephalogram response signal into a pre-trained personality trait classification model, calculating the input electroencephalogram signal, and calculating the personality trait score of the subject.

[0080] An embodiment of the present invention discloses a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions. When the program instructions are executed by a computer, the computer can execute the methods provided by the above-mentioned method embodiments, for example, including: playing personality test materials to a subject to induce the subject's personality trait-related cognitive state, and collecting the subject's electroencephalogram response signal; preprocessing the collected electroencephalogram response signal to filter out artifact noise and standardized signals, and constructing it into a training sample set; inputting the electroencephalogram response signal into a pre-trained personality trait classification model, calculating the input electroencephalogram signal, and calculating the subject's personality trait score.

[0081] The present embodiment provides a non-transitory computer-readable storage medium, which stores computer instructions. The computer instructions enable a computer to execute the methods provided by the above-mentioned method embodiments, for example, including: playing personality test materials to a subject to induce the subject's personality trait-related cognitive state, and collecting the subject's electroencephalogram response signal; preprocessing the collected electroencephalogram response signal to filter out artifact noise and standardized signals, and constructing it into a training sample set; inputting the electroencephalogram response signal into a pre-trained personality trait classification model, calculating the input electroencephalogram signal, and calculating the subject's personality trait score.

[0082] A person skilled in the art can understand that all or part of the steps of the above method embodiment can be implemented by executing program instructions through relevant hardware. The above program can be stored in a computer-readable storage medium, and when the program is executed, each step of the above method embodiment will be executed. The above storage medium includes: ROM, RAM, magnetic disk or optical disk and other media that can store program codes.

[0083] The embodiments of the above-mentioned devices are only for illustrative purposes. In specific implementations, each unit may be physically separated as a separate component or may not be. Those components displayed as units may be physical units or may not be. In other words, these components may be centrally deployed in one location or distributed on multiple network units. In practical applications, some or all of the modules may be selected for use according to specific needs to achieve the intended functions and purposes of this embodiment. For those of ordinary skill in the art, these designs can be fully understood and implemented without the need for creative work.

[0084] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Although the above embodiments describe the present invention in detail, it should be understood that a person skilled in the art can still modify the technical solutions or replace some of the technical features with equivalents. These modifications or replacements do not change the essence of the corresponding technical solutions, and still fall within the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A personality trait recognition method based on EEG signals, characterized in that: The following steps are involved: S1. Playing personality test materials to the subjects to induce the cognitive state related to the personality traits of the subjects, and collecting the EEG response signals of the subjects; S2. preprocessing the collected EEG response signals to filter out artifact noise and standardize signals, and constructing a training sample set; S3. Input the preprocessed EEG response signal into the pre-trained personality trait classification model, calculate the input EEG signal, calculate the personality trait score of the subject, and complete the classification; The classification model is a LSSENet classification model; In the spatial dimension, the electrode channels that contribute to personality classification are identified by integrating the fully convolutional neural network and the compressed excitation network. The input of the structure is the EEG data that has been de-aliased and standardized, and the shape is (N×C×S), where N represents the batch size, C is the number of channels, and each channel has S time steps. There are three temporal convolution blocks in the structure, and the convolution block contains a convolution layer, a batch normalization layer, and a ReLU activation layer. After the first two convolution blocks, the SE Block is inserted to extract channel features of the EEG signal, and its weight matrix represents the contribution of each electrode channel to feature extraction. After the third convolution block, the global average pooling layer is connected. In the time dimension, a 2D convolution filter is used to extract the time dimension information of the EEG signal. The 2D convolution filter is composed of a 2D convolution block, which includes a 2D convolution layer, a batch normalization layer and a ReLU activation layer; the kernel size of the 2D convolution filter is set to C×S×1; batch normalization is normalized at the level, and the ReLU activation layer introduces nonlinearity in the network; Use LSTM network to learn the short-term continuity and long-term similarity features of EEG signals, where the forget gate determines which information should be forgotten, and the input gate and output gate selectively update the current state; Inserting the attention mechanism enables the subsequent LSTM to focus on long-term dependencies; The core formula of the attention mechanism is: ; in, is the weight associated with time step i; The Conact layer is used to fuse the temporal and spatial feature maps according to the channel dimension, and then the softmax function is used to complete the final classification of the Big Five personality traits.

2. The method according to claim 1, characterized in that: Before step S1, the following steps are also included: For the 20 emotions in the positive and negative emotion scale, the corresponding materials are selected as personality test materials; the personality test materials in the material library are divided into 4 categories according to the valence threshold and arousal value: high valence and high arousal HVHA, high valence and low arousal HVLA, low valence and high arousal LVHA and low valence and low arousal LVLA; the type of personality test material is video, with a duration between 2 minutes and 20 minutes; The subjects specifically include: people of different ages with different cultural backgrounds and different educational levels.

3. The method according to claim 1, characterized in that Before step 2, the method further includes: Construct a training sample library based on the EEG data of multiple people with different personality traits and their corresponding personality scores; The personality trait classification model is trained through the training sample library.

4. The method according to claim 1, characterized in that The EEG response signal in step S2 includes: frequency band energy of delta, theta, alpha, beta and gamma bands in the EEG signal.

5. The method according to claim 4, characterized in that The EEG signals include at least: EEG signals obtained from electrodes AF3, F7, F3, FC5, T7, P7, O1, O2, P8, T8, FC6, F4, F8 and AF4.

6. The method according to claim 1, characterized in that When playing the personality test materials to the subject in step S1, the personality test materials are alternately sorted according to their categories and then played through the playback device.

7. The method according to claim 1, characterized in that In the step 3, a classification model is constructed with the self-assessment questionnaire score of the individual's personality traits as the target; wherein the classification model is an LSSENet classification model; a personality trait measurement method is used to perform questionnaire scores on the personality traits of the training individual; and the classification model is trained according to the questionnaire scores of the training individual to obtain a pre-trained personality trait classification model.

8. A personality trait recognition system based on EEG signals, characterized in that: The system is used to implement the method described in any one of claims 1 to 7, and the system includes: A signal receiving module, used to play personality test materials to the subject to induce the subject's personality trait-related cognitive state, and collect the subject's electroencephalogram response signal; A data processing module is used to pre-process the collected EEG response signals to filter out artifact noise and standardize signals, and construct a training sample set; The calculation module is used to input the EEG response signal into a pre-trained personality trait classification model, calculate the input EEG signal, calculate the personality trait score of the subject, and complete the classification.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the method for identifying personality traits based on EEG signals according to any one of claims 1 to 7 are implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for identifying personality traits based on EEG signals as described in any one of claims 1 to 7 are implemented.

Citation Information

Cited By

  • Intelligent evaluation method, device and equipment for psychological safety quality of electric power practitioner based on electroencephalogram information and storage medium

    CN120458576A

  • Electroencephalogram-based classification label determination method and device, and storage medium

    CN122767878A