Criminal unconscious risk psychological assessment device and assessment system thereof
Through the combination of multi-source information collection and convolutional neural network model, the complex and single problem of information collection of existing psychological assessment systems is solved, more accurate psychological assessment and risky behavior intervention are achieved, and management effect is improved.
Patent Information
- Application Number
- CN202411177647.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-26
- Publication Date
- 2024-12-13
AI Technical Summary
The existing psychological assessment system mainly collects information through question and answer and test questionnaires. The work is complex and the source is single, and it is easily affected by the lies and disagreements of the assessees, resulting in inaccurate test results.
Multi-source information collection method is adopted, combining cameras to collect behaviors and expressions in real time, and recording equipment to collect sounds, and through convolutional neural network model and COPA-PI test, combined with online big data acquisition and offline face-to-face inquiry, a psychological evaluation model is generated and report, recommendations for interventional treatment options.
It improves the accuracy and reliability of psychological assessment, reduces the impact of intentional concealment and lying by the assessee, and provides more accurate results of unconscious psychological assessment, which facilitates intervention and management of dangerous behaviors.
Smart Images
Figure CN119112185A_ABST
Abstract
Description
Technical Field
[0001] The invention relates to the technical field of evaluation systems, in particular to a criminal unconscious dangerousness psychological evaluation device and an evaluation system thereof. Background Art
[0002] Relevant studies have shown that some criminals are prone to unconscious dangerous behaviors and psychology due to their poor cultural level, poor self-control ability, and some subjective and objective reasons. In order to improve the management of criminals, it is necessary to evaluate the psychological condition of criminals. After searching, it is found that the psychological assessment system in the prior art is typically such as a psychological assessment system and method with publication number CN112182339A. The system obtains network behavior data that does not require subjective cooperation from the person to be assessed, or data that is less affected by the subjective factors of the person to be assessed as analytical sample data for psychological assessment. Its main feature is that it can improve the accuracy of sample data and ensure the accuracy of test results, so that the psychological assessment score results calculated by the psychological assessment model can objectively and accurately reflect the current mental health status of the person to be assessed, thereby providing accurate and objective judgment basis for clinical psychological diagnosis.
[0003] In summary, the information collection of the existing psychological assessment system is mainly carried out through questions and answers and test questionnaires. The information collection work is complicated and the source is single and limited. At the same time, the test results are easily affected by the lies and non-cooperation of the person being assessed. In view of the above problems, the existing equipment needs to be improved. Summary of the invention
[0004] The purpose of the present invention is to provide a criminal unconscious dangerousness psychological assessment device and an assessment system thereof, so as to solve the problem that the information collection of the existing psychological assessment system proposed in the above background technology is mainly carried out in the form of questions and answers and test questionnaires, the information collection work is complicated and the source is single and limited, and the test results are easily affected by the lies and non-cooperation of the person being assessed.
[0005] To achieve the above-mentioned purpose, the present invention provides the following technical solutions: a criminal unconscious dangerous psychological assessment device, comprising a camera and a computer component, characterized in that the camera collects the criminal's behavior and expression in real time, and the camera is electrically connected to the computer component, and the computer component includes a recording device, a display and a host.
[0006] Preferably, a criminal unconscious dangerousness psychological assessment system is installed in a criminal unconscious dangerousness psychological assessment device as described in claim 1, characterized in that it comprises:
[0007] A collection module, which is used to collect personal information of criminals, including question-answering and big data retrieval;
[0008] A training module, wherein the training module uses the personal information of criminals collected by the collection module as sample information, inputs the sample information into a convolutional neural network model for training, and obtains a criminal psychological assessment model;
[0009] A test module, wherein the test items of the test module include a convolutional neural network test and a COPA-PI test;
[0010] An analysis module, which analyzes the test results and generates a text report output;
[0011] The recommendation module analyzes the obtained test results and actively generates the best interventional treatment plan.
[0012] Preferably, the content collected by the collection module includes the criminal's cultural level, values, personality traits, family relationships, mental health and social relationship cognition.
[0013] Preferably, the collection module collects information in the form of face-to-face question and answer, test papers, daily behavior monitoring, interactive games and big data collection. The big data collection form is mainly manifested in using Internet big data as a basis to query and concentrate its preferences. At the same time, the information extraction process on the Internet big data is: first, use a crawler to crawl the web pages that it has browsed, and then use an extraction algorithm to extract key information on the web pages.
[0014] Preferably, the convolutional neural network in the training module is a LeNet architecture, and the processing flow of the LeNet architecture is convolution, nonlinear processing, i.e. ReLU, pooling or subsampling, and classification, i.e. a fully connected layer.
[0015] Preferably, the test questions of the convolutional neural network test in the test module are preset, and the test questions of the convolutional neural network test involve aspects such as life, emotions, interpersonal relationships and career.
[0016] Preferably, the COPA-PI test item indicators in the test module include questions mainly composed of extroversion, intelligence, sympathy, subordination, volatility, impulsivity, vigilance, inferiority, anxiety, violent tendencies, abnormal psychology and criminal thinking.
[0017] Preferably, the analysis module analyzes the causes of the test results through an analysis algorithm, and the analysis results are synchronously stored in the database. The analysis module uses a linear regression algorithm for analysis, and the formula of the linear regression algorithm is:
[0018] y=wx+b,
[0019] Where y is the predicted value. In order to construct this functional relationship, known data points are used to solve the unknowns w and b in the equation.
[0020] Preferably, the recommendation module recommends a treatment plan based on the recommendation method stored in the database, and the recommendation module calculates with the plan stored in the database through a similarity matching algorithm.
[0021] Preferably, the similarity matching algorithm adopts the KMP algorithm, and the calculation process of the KMP algorithm is that the KMP algorithm first calculates the prefix function array next of the pattern string, where next represents the longest common prefix and suffix length in the substring ending with i in the pattern string, and then during matching, if a mismatch is found, the pattern string is moved to the right by several characters according to the next array, and the comparison is restarted. The time complexity of the KMP algorithm is 0(m+n), where m and n are the lengths of the main string and the pattern string respectively.
[0022] Compared with the prior art, the beneficial effects of the present invention are: the criminal unconscious dangerous psychological assessment device and the assessment system thereof,
[0023] The present invention can effectively solve the problem that the information collection of the existing psychological assessment system is mainly carried out in the form of questions and answers and test questionnaires, the information collection work is complicated and the source is single and limited, and the test results are easily affected by the lies and non-cooperation of the person being assessed. In order to prevent the person being assessed from deliberately concealing and lying in front of the collector, the relevant information of the person being assessed is collected through online and offline methods. The online collection is manifested in using a crawler to crawl the content that the person being assessed browsed before being monitored, that is, when he was not restricted in the past, and then using a keyword extraction algorithm to extract key words and sentences appearing in the browsed content. The offline performance is that the staff synchronously records the facial expressions and expectations of the person being assessed when answering through face-to-face inquiries, so as to ensure the accuracy of the collected information, and then more accurately conduct unconscious psychological assessments on him, and at the same time facilitate further intervention and treatment of his dangerous action psychology, so as to optimize his management and avoid the occurrence of criminal behavior. BRIEF DESCRIPTION OF THE DRAWINGS
[0024] Figure 1 This is a schematic diagram of the evaluation process of the present invention;
[0025] Figure 2 It is a content composition diagram collected by the collection module of the present invention;
[0026] Figure 3 It is a COPA-PI test item composition diagram of the test module of the present invention;
[0027] Figure 4The figure is a flow chart of the interactive game in the acquisition module of the present invention.
[0028] In the figure: 1. Acquisition module; 2. Training module; 3. Testing module; 4. Analysis module; 5. Recommendation module. DETAILED DESCRIPTION
[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0030] See also Figure 1-4 The present invention provides a technical solution: a criminal unconscious dangerous psychological assessment device and an assessment system thereof,
[0031] Embodiment 1
[0032] A criminal unconscious dangerousness psychological assessment device includes a camera and a computer component, wherein the camera collects the criminal's behavior and expression in real time, and the camera is electrically connected to the computer component, and the computer component includes a recording device, a display and a host.
[0033] Specifically, the camera and recording equipment are used to record the behavior and voice of the subject being evaluated when the subject is being questioned, i.e., when collecting information, so as to facilitate subsequent judgment on whether the answer is true based on the fluctuations of his or her body behavior and voice.
[0034] Embodiment 2
[0035] This embodiment is a further description of the above embodiment. It should be understood that this embodiment includes all the aforementioned technical features and is further described in detail.
[0036] A criminal unconscious dangerousness psychological assessment system, comprising:
[0037] Collection module 1, which is used to collect personal information of criminals, including question-answering and big data retrieval;
[0038] Training module 2, training module 2 uses the personal information of criminals collected by collection module 1 as sample information, inputs the sample information into the convolutional neural network model for training, and obtains a criminal psychological assessment model;
[0039] Test module 3: Test items of test module 3 include convolutional neural network test and COPA-PI test;
[0040] Analysis module 4: Analysis module 4 analyzes the test results and generates a text report output;
[0041] Recommendation module 5: Recommendation module 5 analyzes the obtained test results and actively generates the optimal interventional treatment plan.
[0042] In a further embodiment, the content collected by the collection module 1 includes the criminal's cultural level, values, personality traits, family relationships, mental health and social relationship cognition.
[0043] In a further embodiment, the collection module 1 collects information in the form of face-to-face question and answer, test papers, daily behavior monitoring, interactive games, and big data collection. The interactive game is specifically performed as follows: first, select any one of three shapes, namely, a triangle, a square, and a hexagon. Take the triangle as an example. Then, the triangular self-harm picture and the neutral picture are successively displayed to the person being evaluated to let him identify their shapes. The identification time is in milliseconds. Then, the identification time spent on the two picture shapes is recorded separately. Finally, the time spent on identifying the self-harm picture is subtracted from the time spent on identifying the neutral picture to obtain a set of difference values. When the person being evaluated has an unconscious tendency to self-harm, the longer the time spent on identifying the shape of the self-harm picture, the greater the difference value, and vice versa.
[0044] The main form of big data collection is to use Internet big data as a basis to query and condense their preferences. At the same time, the information extraction process on Internet big data is: first, use crawlers to crawl the web pages they have browsed and the texts, pictures, audios, videos, etc. downloaded by the evaluated persons from the web pages, and then use extraction algorithms to extract key information on the web pages. In addition to web pages, it also includes network media / service tools. Network media / service tools include but are not limited to QQ, WeChat, email, e-books, e-newspapers, e-magazines, radio, television, newsletters, audio-visual products, exhibition products, etc. Web page types include corporate websites, large portals, industry websites, transaction websites, classified information websites, forum websites, government websites, functional websites, entertainment websites, etc.
[0045] In a further embodiment, the convolutional neural network in the training module 2 is a LeNet architecture, and the processing flow of the LeNet architecture is convolution, nonlinear processing, i.e. ReLU, pooling or subsampling, and classification, i.e. a fully connected layer.
[0046] In a further embodiment, the test questions of the convolutional neural network test in the test module 3 are preset, and the test questions of the convolutional neural network test involve aspects such as life, emotions, interpersonal relationships and career.
[0047] In a further embodiment, the COPA-PI test item indicators in the test module 3 include questions mainly composed of extraversion, intelligence, sympathy, affiliation, volatility, impulsivity, vigilance, inferiority, anxiety, violent tendencies, abnormal psychology and criminal thinking.
[0048] In a further embodiment, the analysis module 4 analyzes the cause of the test result through an analysis algorithm, and the analysis result is synchronously stored in the database. The analysis module 4 uses a linear regression algorithm for analysis, and the formula of the linear regression algorithm is:
[0049] y=wx+b,
[0050] Where y is the predicted value. In order to construct this functional relationship, known data points x and y are used to solve the unknowns w and b in the equation.
[0051] In a further embodiment, the recommendation module 5 recommends a treatment plan based on the recommendation method stored in the database, and the recommendation module 5 calculates with the plan stored in the database through a similarity matching algorithm.
[0052] In a further embodiment, the similarity matching algorithm adopts the KMP algorithm, and the calculation process of the KMP algorithm is that the KMP algorithm first calculates the prefix function array next of the pattern string, where next represents the longest common prefix and suffix length in the substring ending with i in the pattern string, and then during matching, if a mismatch is found, the pattern string is moved to the right by several characters according to the next array, and the comparison is restarted. The time complexity of the KMP algorithm is 0(m+n), where m and n are the lengths of the main string and the pattern string, respectively. When the length of the pattern string is very large, the KMP algorithm is much more efficient than the brute force matching algorithm.
[0053] Specifically, in the actual working process, the collection module 1 collects relevant information of the person being evaluated through online and offline collection methods, and then the training module 2 uses the personal information of the criminal collected by the collection module 1 as sample information, and inputs the sample information into the convolutional neural network model for training to obtain a criminal psychological assessment model, and then the testing module 3 tests it through the convolutional neural network test and the COPA-PI test to obtain the test results, and then the analysis module 4 will analyze the results and trigger the recommendation module 5 to recommend corresponding intervention measures to the person being evaluated based on the analysis results, that is, to reduce criminal behavior through psychological monitoring and therapeutic intervention of the person being evaluated.
[0054] Although the present invention has been described in detail with reference to the aforementioned embodiments, it is still possible for those skilled in the art to modify the technical solutions described in the aforementioned embodiments, or to make equivalent substitutions for some of the technical features therein. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.
Claims
1. A criminal unconscious dangerous psychological assessment device, comprising a camera and a computer component, characterized in that: The camera collects the criminal's behavior and expression in real time, and the camera is electrically connected to the computer component, and the computer component includes a recording device, a display and a host.
2. A criminal unconscious dangerousness psychological assessment system, which is installed in a criminal unconscious dangerousness psychological assessment device according to claim 1, characterized in that: include: A collection module (1), the collection module (1) is used to collect the personal information of criminals, and the collection method includes question answering and big data retrieval; A training module (2), wherein the training module (2) uses the personal information of the criminal collected by the collection module (1) as sample information, inputs the sample information into a convolutional neural network model for training, and obtains a criminal psychological assessment model; A test module (3), wherein the test items of the test module (3) include a convolutional neural network test and a COPA-PI test; An analysis module (4), wherein the analysis module (4) analyzes the test results and generates a text report output; The recommendation module (5) analyzes the obtained test results and actively generates the best interventional treatment plan.
3. A criminal unconscious dangerousness psychological assessment system according to claim 2, characterized in that: The content collected by the collection module (1) includes the criminal's cultural level, values, personality traits, family relationships, mental health and social relationship cognition.
4. A criminal unconscious dangerousness psychological assessment system according to claim 2, characterized in that: The collection module (1) collects information in the form of face-to-face question and answer, test papers, daily behavior monitoring, interactive games, and big data collection. The big data collection form mainly uses Internet big data as a basis to query and condense preferences. At the same time, the information extraction process on the Internet big data is: first, crawl the web pages that have been browsed by the crawler, and then use the extraction algorithm to extract key information on the web pages.
5. A criminal unconscious dangerousness psychological assessment system according to claim 2, characterized in that: The convolutional neural network in the training module (2) is a LeNet architecture, and the processing flow of the LeNet architecture is convolution, nonlinear processing, i.e. ReLU, pooling or subsampling, and classification, i.e. a fully connected layer.
6. A criminal unconscious dangerousness psychological assessment system according to claim 2, characterized in that: The test questions of the convolutional neural network test in the test module (3) are preset, and the test questions of the convolutional neural network test involve aspects such as life, emotions, interpersonal relationships and career.
7. A criminal unconscious dangerousness psychological assessment system according to claim 2, characterized in that: The COPA-PI test item indicators in the test module (3) include questions mainly consisting of extroversion, intelligence, sympathy, subordination, volatility, impulsivity, vigilance, inferiority, anxiety, violent tendency, abnormal psychology and criminal thinking.
8. A criminal unconscious dangerousness psychological assessment system according to claim 2, characterized in that: The analysis module (4) analyzes the causes of the test results through an analysis algorithm, and the analysis results are synchronously stored in the database. The analysis module (4) uses a linear regression algorithm for analysis, and the formula of the linear regression algorithm is: y=wx+b, Where y is the predicted value. In order to construct this functional relationship, known data points (x, y) are used to solve the unknowns w and b in the equation.
9. A criminal unconscious dangerousness psychological assessment system according to claim 2, characterized in that: The recommendation module (5) recommends a treatment plan to the patient based on the recommendation method stored in the database. At the same time, the recommendation module (5) performs calculations with the plan stored in the database through a similarity matching algorithm.
10. A criminal unconscious dangerousness psychological assessment system according to claim 9, characterized in that: The similarity matching algorithm adopts the KMP algorithm, and the calculation process of the KMP algorithm is that the KMP algorithm first calculates the prefix function array next of the pattern string, where next represents the longest common prefix and suffix length in the substring ending with i in the pattern string, and then during matching, if a mismatch is found, the pattern string is moved to the right by several characters according to the next array, and the comparison is restarted. The time complexity of the KMP algorithm is 0(m+n), where m and n are the lengths of the main string and the pattern string respectively.